System Stability of the Overhead Power Supply System used in the Electric Road System

(1)

TVE-MFE; 19008

Examensarbete 30 hp October 2019

System Stability of the Overhead Power Supply System used in the Electric Road System

Md Asif Ferdoush

Masterprogram i förnybar elgenerering

(2)

Teknisk- naturvetenskaplig fakultet UTH-enheten

Besöksadress:

Ångströmlaboratoriet Lägerhyddsvägen 1 Hus 4, Plan 0

Postadress:

Box 536 751 21 Uppsala

Telefon:

018 – 471 30 03

Telefax:

018 – 471 30 00

Hemsida:

http://www.teknat.uu.se/student

Abstract

System Stability of the Overhead Power Supply System used in the Electric Road System

Md Asif Ferdoush

This thesis presents the stability analysis of an Electric Road System (ERS), which is often abbreviated as eHighway, used for the electrification of the hybrid vehicle. The overall system modelling of the ERS, starting from the sub-station to the critical part of the Scania hybrid truck is performed in the MATLAB Simulink environment. The ERS consists of an overhead catenary line (OCL), where vehicles are electrified by using a pantograph mounted over the vehicle. The stability analysis of the power supply of the overhead line is done by taking into account several aspects of the system. The simulation results are validated with the real test track measurements and the deviations are shown. The frequency response of the system is considered to measure the stability margin. The resonance conditions are clarified and essential variable choke is proposed to damp them out. Also the harmonic components injected from the vehicle side, that are in the closer range of the resonance, are figured out and filtered. When multiple vehicles are electrified from the same catenary line, then there are interferences in between the vehicles. These disturbances both to the vehicles and the overhead power supply system are presented in the time domain. Finally, the results are shown to demonstrate the effectiveness of the variable choke to increase the stability margin in the overhead supply system. In the frequency domain results, it has shown that the resonance is shifted out of the system operating frequency. In the time domain results, it has shown that the high amplitude of the current and voltage signals are sufficiently damped out by variable choke implementation.

TVE-MFE; 19008 Examinator: Irina Temiz

Ämnesgranskare: Valeria Castellucci Handledare: Christer Roos, Kelin Jia

(3)

Sammanfattning

Detta examensarbete redovisar stabilitetsanalysen gjord på ett ”Electric Road System” (ERS) som används för elektrifiering av hybridfordon. Den övergripande systemmodelleringen av ERS, från kraftstationen till (den kritiska delen av) hybridbilen, gjordes i MATLAB Simulink. ERS ut- görs av en hängande kontaktledning (OCL) varifrån fordon kan strömförsörjas med hjälp av en strömavtagare som är monterad på fordonet. Stabilitetsanalysen av strömförsörjningssystemet görs med hänsyn till flera aspekter av systemet. Simuleringsresultaten jämförs med verkliga mät- ningar och avvikelser redovisas. Systemets frekvensrespons anses mäta (alt. vara ett mått på) dess stabilitetsmarginal. Resonansförhållanden klargörs och grunderna för en variabel induktans föreslås för att dämpa ut dessa resonanser. Även de övertoner som injiceras från fordonssidan räk- nas fram och filtreras. När flera fordon elektrifieras från samma ledning uppstår det interferens och därmed störningar mellan fordonen. Dessa störningar, både mellan fordonen och från fordon till strömförsörjningssystemet, presenteras i tidsdomänen. Slutligen visas verkan av en variabel induktans och dess förmåga att öka stabilitetsmarginalen i strömförsörjningssystemet. Från resultaten av simuleringar i frekvensdomänen visas att resonansen flyttas ut från systemfrekvensen.

Från resultaten av simuleringar i tidsdomänen visas att de höga amplituderna i ström och spänning dämpas tillräckligt genom implementering av en variabel induktans.

(4)

Acknowledgement

I would like to express my sincere appreciation to my supervisor, Christer Roos, Design Engi- neer, Scania CV AB for his guidance, encouragement, and support throughout the course of this work. I am grateful that Dr. Juan de Santiago, Associate Professor, Uppsala University, has spent a significant amount of time from his busy schedule to advice me to successfully complete this thesis.

I would like to thank my subject reader Valeria Castellucci, Associate Senior Lecturer, Uppsala University, for her guidance and help to make a better thesis.

I am thankful to my previous supervisor Dr. Kelin Jia, who found the enthusiasm in me about this ERS and gave me the opportunity to work for the thesis. I am truely thankful to Malin Andersson, Industrial PhD student, SCANIA CV AB and KTH, from whom I came to know about the ERS and got my first motivation to work for the eTruck.

It has been a great pleasure to work with Oscar Hällman, Development Engineer at Scania to understand many control aspects and gather some practical measurements from the test rig.

I am particularly thankful to Associate Professor Markus Gabrysch and PhD student Arvind Par- wal from Uppsala University, and Julian Taube from Technical University of Munich for their patience, guidance and support to my understanding on the Power Electronics and it’s application to the Electric Vehicle in a better way.

I would like to express my gratitude to the Swedish Institute, for their full support in my master study here at Sweden. I would also like to thank my friends both from Uppsala and Munich, who always inspired me during my two years of study.

Finally, I am thankful to my parents and my sisters for their unconditional support and encouragement in chasing my dreams.

(5)

List of Figures

1.1 World’s first public Electric Road System . . . 1

1.2 Real measurement of voltage and current signals from El Camino . . . . 2

2.1 Internal structure of the ERS connected eTruck . . . 4

2.2 El Camino hybrid truck . . . 5

3.1 Frequency domain model of the ERS system for one vehicle in LPM . . 8

3.2 Frequency domain model of the ERS system for one vehicle in HPM . . 9

3.3 Frequency domain model of the ERS system for two vehicles in LPM . 9 3.4 Frequency domain model of the ERS system for two vehicles in HPM . 9 3.5 Time domain model of the ERS system for one vehicle . . . 10

3.6 Time domain model of the ERS system for two vehicles . . . 10

3.7 Pantograph and DC-DC converter in the vehicle . . . 11

3.8 Traction drive and eMachine inside the vehicle . . . 11

3.9 Simplified battery model . . . 12

3.10 Traction drive controlling model . . . 12

4.1 Resonance at different distances from the substation for single vehicle at LPM . . . 16

4.2 Resonance at different distances from the substation for single vehicle at HPM . . . 17

4.3 Resonance for 1st vehicle at different distances from the substation running 2V LPM mode . . . 18

4.4 Resonance for 2nd vehicle at different distances from the substation running 2V LPM mode . . . 19

4.5 Resonance for 1st vehicle at different distances from the substation running 2V HPM mode . . . 20

4.6 Resonance for 2nd vehicle at different distances from the substation running 2V HPM mode . . . 21

4.7 THD for a single vehicle running at 30 km/h . . . 22

5.1 Performance Index for 500 m from substation (30 km/h - LPM and 36, 45 80 km/h - HPM) . . . 25

5.2 Algorithm for Performance Index . . . 26

5.3 Algorithm for Vehicle Position . . . 26

6.1 Pantograph out voltage for 80 km/h vehicle speed . . . 29

6.2 Pantograph out current for 80 km/h vehicle speed . . . 30

(8)

6.3 Torque of the eMachine at 80 km/h vehicle speed . . . 30

6.4 eMachine speed at 80 km/h vehicle speed . . . 31

6.5 Damped out Performance Index for 30 km/h vehicle speed . . . 32

6.6 Resonance shift during LPM at 150 m away from the substation . . . . 33

6.7 Pantograph out voltage with 30 km/h vehicle speed 150 m away . . . . 33

6.8 Pantograph out current with 30 km/h vehicle speed 150 m away . . . 34

6.9 Pantograph out voltage with 36 km/h vehicle speed 40 m away . . . 34

6.10 Pantograph out current with 36 km/h vehicle speed 40 m away . . . 35

6.11 Pantograph out voltage with 30 km/h vehicle speed for vehicle 1 and 2 . 35 6.12 Pantograph out current with 30 km/h vehicle speed for vehicle 1 and 2 . 36 6.13 Total harmonic distortion when vehicle running at 30 km/h (LPM) . . . 37

6.14 Total harmonic distortion when vehicle running at 36 km/h (HPM) . . . 37

6.15 Total harmonic distortion when 2 vehicles are running at 30 km/h (LPM) 38 A.1 Damped out Performance Index for 36 km/h vehicle speed . . . 43

A.2 Damped out Performance Index for 45 km/h vehicle speed . . . 43

A.3 Damped out Performance Index for 80 km/h vehicle speed . . . 44

A.4 Resonance shift during LPM at 80 m away from the substation . . . 44

A.5 Resonance shift during LPM at 120 m away from the substation . . . . 45

A.8 Pantograph out voltage with 30 km/h vehicle speed 80 m away . . . 46

A.9 Pantograph out voltage with 30 km/h vehicle speed 120 m away . . . . 47

A.12 Pantograph out current with 30 km/h vehicle speed 80 m away . . . 48

(9)

List of Tables

6.1 Choke inductance for 30 km/h vehicle speed . . . 32

A.1 Choke inductance for 36 km/h vehicle speed . . . 42

(10)

1 Introduction

The Climate Change is of great concern nowadays, for the sustainability of the society.

The emission of the greenhouse gases are mainly responsible for the global warming. A total of 65% of greenhouse gases are coming from the CO2 emissions [1]. One of the world’s major CO2 production is coming from the transport system (land, air and sea).

The emissions from the transport system will be doubled by the end of 2050, due to the high demand of the car in the developing countries and the growth of the aviation system [2]. During the last years, the transport agencies are trying to find sustainable solutions to reduce these emissions. In connection to that, the road transport system is adopting different technologies to partially or fully replace the combustion engine to reduce the CO2 emission. A novel way to overcome such challenges is the ERS that uses a branch of technologies to charge the vehicles while in motion [3]. For instance, the conductive system is more able to support the heavy duty vehicle, whereas the inductive system is suitable for the vehicle that has low power requirement. But the conductive system has higher efficiency over the inductive system [3, 4]. To meet this goal, in 2010 SIEMENS started a German government funded project ENUBA to develop a more energy-efficient system for the long haul hybrid truck (eTruck) and the infrastructure was named "eHigh- way" [5].

1.1 Background

In 2016, world’s first ERS on public road was built outside of Gävle in Sandviken, Swe- den, which is shown in figure 1.1 and since then it is in operation [6]. SCANIA CV AB is no exception in this race and some trucks are still running in that ERS. Scania hybrid

Figure 1.1. World’s first public Electric Road System

(11)

trucks are adopting already established pantograph technology from SIEMENS, to get the traction drive power from the overhead electric power supply line. In this way, the long distance transportation of the goods is possible and efficient without disrupting the environment. The application of this pantograph system in the SCANIA hybrid truck is very new and enormous research is going on to make it more endorsed by the cus- tomers. In January 2019, a hybrid truck El Camino from SCANIA has been tested at the SIEMENS test track Gross Döln outside of Berlin, Germany. The measurements were made during the test phase of the eHighway truck that was built for the German research project “Oberleitungs-LKW”, which is a cooperation project between Volkswagen AG Group Innovation and Siemens GmbH and is funded by the German Federal Ministry for the Environment, Nature Conversation and Nuclear Safety (BMU). El Camino has pantograph mounted over the roof of the truck. It feeds the power from the overhead line to the eMachine inside the truck through the Power Electronic interface. Although, this methodology of taking the power to run the train is very reliable for many years, still lot of challenges persist for the Vehicle level application. In comparison to the single overhead power supply line for the train track, the ERS for the vehicle has two overhead dc line.

1.2 Problem and the aim of study

One of the problem of the ERS is the stability in the system, when vehicles are connected to the overhead catenary and approaching towards the dc substation. As observed from the measurement shown in figure 1.2, the current out from the pantograph starts oscillat- ing. The current amplitude goes around 100 A with a frequency of 300 Hz. The oscillation was observed only when the vehicle was running less than 30 km/h speed. When the vehicle speed crosses the 30 km/h limit, then the internal electrical configuration of the vehicle changes and does not give such voltage and current oscillations. Therefore, the aim of this thesis is to explore the reason behind unexpected high oscillations in the pantograph output signals and propose the intelligent control topology to mitigate them.

Figure 1.2. Real measurement of voltage and current signals from El Camino

1.3 Scope and Tasks

The inductance and the capacitance from the system starting from the substation to the eMachine-drive plays the role to the stability problems. Understanding the behaviour of the system is quite complex, due to the continuous mobility of the vehicles. In fact, the line inductance seen from the vehicle side changes continuously when the vehicle is running. Therefore, the overhead catenary line has frequency dependent effects [7].

This primarily creates the problem in the ERS for the voltage and current fluctuations.

There is a possibility of employing a choke to compensate the change in inductance seen

(12)

from the vehicle side. The choke is an inductor which is used to filter the low or high frequency electric signals. A variable choke would be the best option to compensate the overhead catenary line (OCL) inductance. In that case, the automatic tracking of the vehicle position on a catenary line and the number of vehicles connected is very important to initiate the operation of the choke. The characteristics of the system will then be well known before the choke operation. The control system for the stability problem is easier with the proper information of these parameters. The main scope of this thesis is to design a choke according to the system performance and considering the resonance and harmonics issues, in order to avoid stability problems in the ERS.

In connection to that, initially the ERS system will be modelled in the Simulink both in frequency and time domain. The control system for power source and machine drive will be designed as closely as possible to the present technology. The electrical resonance analysis of the ERS system will be done to find out the critical points under the OCL where the oscillations are severe. The harmonics analysis will show the presence of any critical harmonic components other than the fundamental frequency component. After then, the additional variable choke will be designed based on the control topology which has inspired by the vehicle position under the OCL. Then the simulation results will be validated to the measurement results from Gross Döln. Finally, the time domain results will be shown.

1.4 Thesis Outline

Chapter 1 gives the motivation behind the ERS, the problems in the ERS and the aim of the work.

In chapter 2, the components and the operation of the ERS will be described.

In chapter 3, the model of the ERS will be presented both in the frequency and time domain considering all the critical components starting from the substation to the traction drive of the vehicle.

In chapter 4, the formation of the Resonance and harmonic disturbances will be analyzed.

In chapter 5, the control methodology of the variable choke selection will be described to avoid the system instability.

In chapter 6, the results will be demonstrated and the thesis will be discussed.

Chapter 7 will conclude the thesis.

(13)

2 ERS Components and Operation

In the eMobility concept, the OCL-constructed ERS is the proven technology and the safest way to drive the eTruck [4]. The ERS consists of several parts of the electrical components. The sub-station and the overhead catenary line are the main parts of this system. The substation converts the grid ac voltage to a 650 V dc voltage to the OCL. The ERS has several such substation which are placed in the middle of each OCL sections. In ERS, the OCL distributes electric power to the traction machine of the vehicle from the substation. The direct current (dc) supply system is fed to the hybrid vehicle using the pantograph system. When vehicles are connected with this line, then additional components from vehicle become major contributors of the disturbances in the supply system.

The purpose of the ERS is to provide the clean power to the heavy duty vehicle without violating the general operation of the highway. If needed, the ordinary vehicles can also run under the OCL line without taking the driving power from OCL. The eTruck can also run on the normal highway using the combustion engine. Figure 2.1, shows the internal structure of a hybrid truck connected to the ERS. The pantograph is taking the electric power from the OCL, which is connected to the grid power supply. Then, there is a box for the protection. The dc-dc converter is connected to the pantograph out to charge the battery. The drive inverter is connected both with the dc-dc converter out and pantograph out via switches.

Figure 2.1. Internal structure of the ERS connected eTruck

(14)

2.1 Components

2.1.1 Substation

The sub-station is placed in the middle of the 2 km overhead catenary line, which is usually fed from a 10 kV or 20 kV, 50 Hz ac power supply system. It has medium voltage and large-capacity power transformer. Then it is converted into the dc power supply by adopting a 12 pulse power electronic converter [4]. The supply power from the substation is not purely dc, due to the expensive filters. Therefore, the overhead line should have noises with a frequency of 600 Hz. But due to some impedance unbalance between the delta winding and the Y winding of the transformer, the 12 pulse converter has 300 Hz higher harmonic component at the no load state. For the simulation in the MATLAB Simulink, a very similar power supply system is modelled with a high harmonic component at 300 Hz and with a very low 600 Hz harmonic component.

2.1.2 Overhead Catenary Line (OCL)

The overhead catenary line (OCL) is the main platform that feeds power to the hybrid truck using pantograph system [4]. Currently the OCL is fragmented into 2 km long isolated section in each. The OCL consists of an upper wire (called Tragseil, TS) and the contact wire (called Fahrdraht, FD). There is a very negligible current flows in between these two wires. Therefore, the inductance between these two lines was assumed to be negligible. After then, the inductance per kilometer length of the OCL is found to be 1.04 mH/km.

2.1.3 eTruck

For the current study, the El Camino hybrid truck from Scania is considered. It combines the railway and hybrid technology. It has a 130 kW eMachine. The truck runs electrically when the ERS is available. The rest of the time it uses the combustion engine to drive the vehicle. A 18 kWh battery pack is also included inside the vehicle for the short duration electric drive power backup. Figure 2.2 shows such truck without the trailer.

Figure 2.2. El Camino hybrid truck

(15)

A list of the components inside the eTruck are briefly described below.

Pantograph

The pantograph technology from SIEMENS is integrated with the El Camino truck to receive the electric power from the OCL. This technology is very reliable for the railway system for many years. As seen in the figure 2.2, the pantograph is placed at the top of the vehicle which has two contact shoes to collect the electric power [4]. One contact shoe is connected with the positive line and another contact shoe is connected with the negative line of the OCL. Irrespective from the pantograph used in the railway system which has one contact shoe, this pantograph has two contact shoe only to make a return path for the current.

DC-DC Converter

The dc-dc converter inside the hybrid vehicle is used to isolate the truck from high voltage (HV) OCL, to avoid electrical hazard in case of low power mode. This is a safety purpose, for the human who can touch the vehicle, while running at lower speed. The isolated DC- DC converter is used to introduce double isolation so that the vehicle cannot come into contact with the OCL supply in case of a fault occurs. This is mainly crucial at lower speeds. Currently, in the El Camino truck, a full bridge 30 kW transformer isolated dc- dc converter is used.

Traction Drive

The traction drive consists of a three phase full bridge converter. The dc voltage is converted to a 3 phase ac voltages to the eMachine. The gate pulses for the six IGBTs are generated by adopting a PWM generator that follows the requirement to change the vehicle speed. The frequency of the carrier signal is taken as 6 kHz. According to the requirements, a change in the PWM is depending on the operating point, but in the study the reference signal frequency is initiated to follow the vehicle speed. A maximum vehicle speed of 80 km/h is considered for this study.

eMachine

The eMachine considered for this study is a 130 kW three phase round rotor permanent magnet synchronous machine (PMSM). It is connected with the vehicle wheel through the gearbox. Depending on the load, it takes different power from the supply system.

Battery

The current El Camino truck has 18 kWh battery pack. The battery is used for short stretches outside the ERS. When the battery is out of charge, then the vehicle uses the Internal Combustion Engine (ICE). The battery is connected to the output of the dc-dc converter irrespective to the vehicle operating mode.

2.2 Operation

The eTruck runs as a normal hybrid vehicle using the ICE when there is no ERS. In this time the pantograph is kept at lower position. When the ERS is available then the driver gets the information and raises the pantograph by pushing a switch. After the pantograph

(16)

connection, the diesel engine automatically switches off. If the eTruck is needed to be run at another section of the road while overtaking with the absence of the OCL line, then diesel engine can be automatically switched on without changing the vehicle speed.

At this moment battery energy storage can give the instant backup power when the diesel engine starts. Depending on the speed of the vehicle, there are low power mode (LPM) and high power mode (HPM). During the low power mode the speed of the vehicle is very low (<30 km/h) and in the high power mode it runs very fast (>30 km/h and <90 km/h).

During the LPM, the traction drive of the eMachine is connected via dc-dc converter and during HPM, it is directly grid connected. In the LPM, the dc-dc converter can feed a maximum of 30 kw power. But in the HPM mode high power requirement is fulfilled by the direct grid connected traction drive from the pantograph output.

In the PWM generator of the motor drive, the electric power required for the machine is taken as the reference value. This reference is compared with the real power taken by the machine which is the product of the electromagnetic torque and the angular velocity of the eMachine. Then this signal is fed to a PID controller to make it a quadrature axis signal. Then a dq0 to abc transformer is used to make three reference signals for the three phases [8]. Three PWM converters are then used to generate six gate pulses for the IGBT switches. The drive provides constant power as required by the eMachine. Depending on the input shaft torque and speed of the machine, the current taken by the PMSM varies.

It follows the fundamental principle of current and voltage.

(17)

3 System Modelling

The full system of the ERS is modelled in MATLAB Simulink starting from the substation till the eMachine inside the vehicle. The built-in components of Simulink are used to model the whole system.

3.1 Frequency domain model

In the frequency domain model of the ERS system, only the resistive, inductive, and capacitive components are considered. From this ERS system, the critical RLC components are taken into consideration. From the substation there are some resistive components.

The inductive and resistive components of the OCL, seen from the vehicle side changes over time while vehicle is running. Therefore, this part needs to be compensated without compromising high disturbances. The pantograph has its own inductive and resistive components. There are capacitive components both from the input of dc-dc converter and the inverter inside the vehicle. After considering the above components, the overall system for a single vehicle connected to the OCL, looks like the figure 3.1 when the vehicle is in low power mode. During the low power mode, the impedance from the inverter should not be considered due to the fact that the inverter is fed from the output of the dc-dc converter. On the other hand, in the high power mode the inverter is directly connected with the grid. Therefore, the impedance from the inverter is considered which is shown in the figure 3.2. When two vehicles are connected on the OCL using the pantograph, then the frequency domain model is depicted in figure 3.3 and figure 3.4 for the low power mode and the high power mode respectively. At this moment the focus of the modelling is limited up to two vehicles maximum.

Figure 3.1. Frequency domain model of the ERS system for one vehicle in LPM

(18)

Figure 3.2. Frequency domain model of the ERS system for one vehicle in HPM

Figure 3.3. Frequency domain model of the ERS system for two vehicles in LPM

Figure 3.4. Frequency domain model of the ERS system for two vehicles in HPM

(19)

3.2 Time domain model

In the time domain modelling of the system, along with the RLC components from the frequency domain modelling, additional built-in components from the Simulink are included. The time domain model for the ERS system is depicted in figure 3.5 and figure 3.6 for a single vehicle and double vehicle connections respectively.

Figure 3.5. Time domain model of the ERS system for one vehicle

Figure 3.6. Time domain model of the ERS system for two vehicles

The power supply of 10 kV, 50 Hz is represented by an ac voltage source. The transformer is a three phase three winding transformer, which has dedicated wye-wye and

(20)

wye-delta connections. For a proper 12 pulse voltage rectification, such transformer are required. Then, two 6-pulse full bridge diode rectifiers are connected together in series and the output is fed to the vehicle.

The vehicles are taking the electric power for the traction drive by using the pantograph.

Depending on the position of the vehicle on the overhead catenary line, the OCL overall inductance seen from the vehicle side changes. The internal components inside the vehicle are shown in the figures 3.7 to 3.10.

In figure 3.7, a full bridge dc-dc converter is shown, where the output of the pantograph is connected with the primary switching of the converter. The output of the switching is connected with the transformer and feeding to the secondary switching.

In figure 3.8, the traction drive is shown, which is driving the eMachine. The output of the dc-dc converter is directly connected with the inverter via some filters, when vehicle is running in the low power mode. But, when the vehicle runs at high power mode, then the inverter input is directly connected with the pantograph output parallelly with the dc-dc converter. The inverter is a 6-pulse full bridge active converter. The amount of the mechanical load is determined by the the vehicle speed and corresponding mechanical torque.

Figure 3.7. Pantograph and DC-DC converter in the vehicle

Figure 3.8. Traction drive and eMachine inside the vehicle

(21)

Figure 3.9. Simplified battery model

Figure 3.10. Traction drive controlling model

In figure 3.9, the simplified model for the battery is shown. During the low power mode, the battery is parallelly connected with the traction drive from the output of the dc-dc converter. When the vehicle is running at the high power mode, then this battery is connected alone with the output of the dc-dc converter for charging. The charging and discharging of the battery is automatic and it depends on many factors. For this study, the battery is considered as fully charged before connecting to the OCL.

In figure 3.10, the control of the traction drive is depicted. Six different switching signals are generated from this model and fed to the six switches of the inverter to control the gates. This model is capable of providing the controls for any load conditions required by the vehicle speed. During the acceleration, deceleration or constant speed, need to do is to set a reference speed which in turn will set a reference power. Then the control will follow the procedure described in the operation section of chapter 2. For the optimal control of the rotor angle, another sub-control is added with the rotor angle theta. This sub-optimal control is adopting a lookup table and comparing the rms current of the eMachine with some set points. The following equations were considered to calculate the rms current.

I_rms= I_peak

√2 (3.1)

(22)

Ipeak= q

I_d²+ I_q² (3.2)

where, Idand Iqare direct and quadrature axis current of the eMachine respectively.

(23)

4 Resonance and Harmonics Analysis

4.1 Resonance Problem

The conversion of the dc voltage from the ac grid is not purely dc. It has some ripples included with the dc voltage which is a multiplication of the power grid fundamental frequency 50 Hz, determined by the number of switches in the rectifier. When vehicle is connected on the OCL of the ERS, it takes current from the line for the propulsion.

Irrespective from the current behaviour, the OCL line has inductance which is playing a role in the resonance problem. Usually, in the ERS system, an RLC circuit is formed with the presence of resistive, inductive and capacitive components from the OCL and the hybrid vehicle. Therefore, the resonance could happen for a specific frequency range.

Within this frequency range, the system could have very high current and voltage amplitudes. If the ERS system has an operating frequency within this frequency range, then the sensitive component could be damaged from the overheating [9, 10].

In the ERS system, there could be more than a single vehicle connected to the OCL. The connection of the vehicles is in parallel from the frequency domain perspective. Depend- ing on the speed of the vehicle there are two different modes of operation HPM and LPM.

Figures 4.1 to 4.6 show the resonance for different modes of operation (Single vehicle in LPM & HPM and two vehicles in LPM & HPM). The resonance problem occurs in the test site when the vehicle is close to the substation. Therefore, only nine different points from one quarter of the overhead line starting from the substation are simulated and depicted in the figures. The points for a single vehicle are considered as 5, 10, 40, 80, 120, 150, 180, 220, and 250 m away from the substation. For the simulation, when a second vehicle is connected on the OCL then with the above positions of the first vehicle, the second vehicle has positions at 10, 20, 80, 160, 240, 300, 360, 440, and 500 m away from the substation.

When, a single vehicle is running at either low power mode or high power mode, then it is seen that the resonance is shifting from higher to lower frequency, with the position of the vehicle moving away from the substation. When two vehicles are running at LPM, then the resonance frequency also follows the same fashion as for the single vehicle. When the vehicles are running at HPM mode, then the resonance frequency does not shift very much. It stays within the frequency range of (1773-1800) Hz for the first vehicle. For the second vehicle there are two resonance formed at two different frequencies range of (1417-1487) Hz for first peak and (2150-2233) Hz for the second peak. At HPM, the resonance frequencies are very high, which do not affect to the system frequencies that much. The resonance components for a single vehicle running at LPM and HPM are around 32 dB and 26 dB respectively. When two vehicles are running at LPM mode then, the first vehicle from substation has the resonance component around 30 dB and the second vehicle has the resonance component around 25 dB. And finally, when the

(24)

two vehicles are running at HPM mode, then their corresponding resonance component are found around 24.5 dB and 18.5 dB respectively. The system studied for this thesis has a noise with the dc supply and has a operating frequency of 300 Hz instead of a pure dc supply.

4.2 Harmonics Problem

The traction drive of the hybrid truck is a major source of the harmonic components injection to the overhead power supply system [11, 12]. The other vehicles connected to the same system are also affected by these harmonics injection. Due to the non-linear operation of the Power Electronic (PE) switches, the harmonics are generated. Also the load condition changes continuously for different operating points. The feedback for the control of the PE converters have very short time delay, that makes their operation com- plicated. Although the switches participating in the traction drive application should have the same characteristics, but they usually show slightly varying characteristics which is often neglected during the consideration. This mainly drives the distortion to the power supply system. The harmonic components that are in the closer range to the resonance frequency of the system, will cause the system voltage and current to raise to a much higher value. This might cause the overheating of the other electrical components, electromagnetic interference (EMI) to the nearby communication system [10].

Also, with the variation of the traction drive power taken by the eMachine, the harmonic characteristics are affected. The frequency response is highly dependent on the traction drive power level [13].

Therefore, it is important to identify the critical harmonic components close to the resonance and damp them out by using proper filter design.

Figures 4.7 to 4.10 will show the total harmonic distortion (THD) for different scenarios of the ERS operation. The x axis represents different OCL inductances for differnt distances on the OCL in the range of (0-250) m. The y axis represents the frequency for the first three odd harmonics in the range of (0-1000) Hz. And the z axis represents the harmonic distortions.

In the closer distance from the substation, the total harmonic distortion is higher. When the vehicle is going away from the substation then the THD gets lower. From the figures, it is clearly visible that all the scenarios have very high 300 Hz frequency component which is the system operating frequency. This is considered as the fundamental frequency. Then, there are also the 2nd order and 3rd order harmonic components on 600 Hz and 900 Hz which are comparatively lower. When the vehicle speed goes above 36 km/h, then the harmonic distortion on the other low frequency grows as seen in figue 4.9 and figure 4.10. The THD is higher in the LPM and lower in HPM. When the vehicle is in HPM, then THD starts to slightly increase from the lower vehicle speed to higher vehicle speed.

(25)

Figure 4.1. Resonance at different distances from the substation for single vehicle at LPM

(26)

(27)

Figure 4.3. Resonance for 1st vehicle at different distances from the substation running 2V LPM mode

(28)

(29)

Figure 4.5. Resonance for 1st vehicle at different distances from the substation running 2V HPM mode

(30)

(31)

Figure 4.7. THD for a single vehicle running at 30 km/h

(32)

(33)

5 Control for System Stability

The stability of the overhead power supply system in the ERS highly depends on the internal parameters of the participating vehicle on the OCL. Due to the critical internal parameters of the vehicle, the OCL voltage and current going through the pantograph oscillates with high amplitudes. This phenomena is very severe when the vehicles are passing through the substation. This high oscillation could hamper the sensitive equip- ment or permanently destroy them. Therefore, the control system should be included inside the vehicle to keep the balance in between the requirement and the supply electric power. The following steps will describe the need and procedure of the control for system stability.

5.1 Performance Index

The Performance Index (PI) is a very good measurement tool to determine the amount of the disturbance. The PI based optimization is inspired by the least square method and applied in many fields to optimize the parameters for the reduction of the disturbances.

In [14], a similar application of the PI based optimization has been shown. In this study, Performance Index is applied to determine the disturbance in the pantograph out voltage and current signals, which in turn can describe the stability phenomena in the overhead power supply system. The equation for the PI is given below for the pantograph out signals, which is the sum of the disturbances for both current and voltage signals. Some weighting factors are also added to formulate the PI.

P I =Xq

w1(Vref − V_panout)²+ w2(Iref − I_panout)² (5.1) In the equation, the nominal overhead line voltage and current taken by the machine is considered as the reference and compared with the actual voltage and current taken by the vehicle. Then, PI is found by adding all the sampled results from 0 to 0.5 seconds. In the ERS, the PI is higher when the vehicle is connected to the OCL, which is very close to the substation. Going away from the substation causes lower PI. This phenomena is depicted in the figure 5.1. There are around 50 different points taken from the 500 m OCL starting from the substation. The PI is measured from each of these points for a 500 ms time span. From the figure at the LPM, it is seen that the PI is lower for the higher distance after 200 m from the substation. The PI is comparatively higher for the distance (0-200) m. At the distance between 200 m and 500 m the HPM modes show PI in the range between 50 and 60. However, it also follows the same fashion as for the LPM in between the range of (0-200) m. Therefore, starting from the 250 m distance from the substation, we would like to control the PI to reduce the instability in the overhead power supply system when vehicle is moving towards the substation.

(34)

Figure 5.1. Performance Index for 500 m from substation (30 km/h - LPM and 36, 45 80 km/h - HPM)

5.2 Motivation behind Variable Choke

There is only one parameter causing the high PI near to the substation which is the OCL inductance seen from the vehicle side. To avoid the higher PI close to the substation a variable choke is desirable. A constant choke, which is very easy to design, can also damp out the higher PI. However, this choke will cause additional losses, when the vehicle is away from the substation. In that case, a variable choke can play the essential role of increasing the overall inductance when the vehicle is close to the substation and of decreasing the choke inductance when the vehicle is away from the substation.

To properly design the choke there are some other aspects needed to be taken care of.

The resonance and harmonics are key factors that can also help the choke to be designed in a better way. Therefore, the feedback information of the position of the vehicle on the OCL is used to design such choke. The methodology of determining the position of the vehicle on the OCL is described below.

5.2.1 Vehicle Position Detection

The only varying parameter according to the position of the vehicle can be tracked to instantly determine the vehicle position.The amount of PI corresponds to an overall line inductance which in turn can tell us the vehicle position. High PI value represents a closer position of the vehicle to the substation and vice versa. The relation between the vehicle position D and the OCL overall line inductance L, seen from the vehicle side is given by equation 5.2.

D = 1000 1040

L

2 (5.2)

The following algorithms shown in figure 5.2 and figure 5.3 represent the overall methodology to determine the vehicle position.

(35)

Figure 5.2. Algorithm for Performance Index Figure 5.3. Algorithm for Vehicle Position

5.3 Selection of the Variable Choke

The line inductance in the OCL seen from the vehicle side, changes with the change in vehicle position on the OCL. As the vehicle approaches close to the sub-station then the pantograph out voltage and current starts to oscillate. This oscillation happens due to the resonance circuit formed by the RLC components. Inductance seen from the vehicle reduces when it approaches towards the sub-station. For the system operating frequency, the resonance could be formed for some positions on the OCL.

A variable choke is proposed to compensate the RLC circuit in such a way that it avoids the resonance at the system operating frequency. This choke can be used to increase the inductance when it approaches to the sub-station and to reduce the inductance when it goes away from it.

To control this variable choke, the disturbance from the pantograph output is taken as the input signal. Then a processing unit could be developed by using some control algorithms mentioned above, to initiate the actuator for changing the inductance. This would be a continuous process as long as the vehicle is running under the OCL. Therefore, in the Performance Index diagram of figure 5.1, it is possible to keep the overall PI under 60 for HPM and under 50 for LPM.

To do such task, we have the database for the vehicle speed, corresponding PI and

(36)

vehicle position from the earlier experience of the running vehicle as described in previous subsections. From the real-time measurements of the vehicle speed and computed PI from the vehicle, we can compare with the database and easily estimate the position of the vehicle. Finally the controller can select the proper choke parameter to reduce the PI in the distance range (0-200) m according to the equation 5.3.

Choke L2 = Inductance at 250 m − OCL overall inductance L1 (5.3) We would like to keep the PI for LPM less than 50 from 200 m to the substation. There- fore, when the vehicle is moving towards the substation, then the additional choke will be adjusted so that the sum of Choke inductance and OCL overall inductance seen from the vehicle side, is equal to the OCL overall inductance seen from the vehicle side at 250 m away from the substation. After then, the sum of the inductance would be similar with a corresponding adjustment in the choke, until the vehicle reaches 250 m away from the substation.

In the next chapter, the results will be discussed.

(37)

6 Results and Discussions

In the beginning, the variable choke will be validated from the resonance perspective.

Then, the simulation results will be validated with the real test site measurements. After then, the effect of the variable choke will be depicted in the Performance Index. Finally, the resonance shifting and the harmonics reduction will be illustrated.

6.1 Choke Validation from resonance

The choke selection based on the PI will be validated in this section. The resonance problem and different operation scenarios with the vehicle were described in the previous chapters.

From the Performance Index in figure 5.1, the PI for the single vehicle in the LPM mode at 250 m away from the substation is around 45. From the single vehicle LPM resonance figure in chapter 4, we can see that the resonance at 250 m away from the substation is 32.4 dB at the frequency of 232 Hz, whereas the system operating frequency (300 Hz) component is 7.5 dB. When, we are considering the resonance at 220 m away from the substation, then we find 32.4 dB at the frequency of 246 Hz. Also, the 300 Hz system frequency has the component of around 11 dB which is more than the 300 Hz frequency component at 250 m distance. To get the similar 300 Hz frequency component of 7.5 dB at 220 m distance, the choke is adjusted so that the overall inductance is equal to the OCL inductance seen from the distance 250 m from the substation. The choke is calculated as below,

The OCL inductance seen from the vehicle side at 250 m away from the substation is 520 µH, whereas 457.6 µH OCL inductance is seen at 220 m away from the substation.

Therefore, the adjusted additional choke inductance should be (520-457.6)=62.4 µH to meet the above requirements.

For the other critical operating positions 80, 120, 150, and 180 m away from the substation the above choke adjustments methodology will be followed for this single vehicle running at LPM.

When, two such vehicles are running in the LPM according to the positions described in the previous chapter, then both vehicles show the resonance in the very close frequency range. The resonance components at 250 m distance away from the substation for the 1st vehicle and 500 m distance away from the substation for the second vehicle are 30.5 dB and 24.7 dB respectively with a corresponding frequency of 232 Hz and 228 Hz. Their 300 Hz frequency components are very close to the 10 dB. For the two vehicle running at LPM we have a PI in between 50 and 60 at the distance 250 m. To keep this PI below 60 for the distance (0-200) m and the 300 Hz system frequency component around 10 dB, the choke should be adjusted as the methodology described above.

When we are considering the single vehicle running at high power mode, then the vehicle is affected by the high distortion of voltages and currents only at the distances 40, 80, and 120 m away from the substation. From the PI diagram, the PI goes over 60 from

(38)

200 m away when the vehicle is approaching towards the substation. Therefore, the previous methodology can be followed again to keep the PI less than 60 for these critical points, which in turn will keep the 300 Hz components around 5 dB. This time for the control, instead of the inductance at the point 250 m, we are considering the OCL overall inductance at 200 m away from the substation. When, both of the vehicles were in HPM, then no 300 Hz high components were observed from the vehicles. Therefore, the choke control can be neutralized when the vehicles are running in HPM.

6.2 Validation of Simulation

The system is modelled considering all the real components and parameters. The built in blocks of the Simulink are used to model the system for the simulation. This section will show the comparison between the results from such Simulink model and real test measurements.

6.2.1 Pantograph out Measurements

The following figures will show the pantograph output voltage and current for both test and simulated results. From many different operating points of the measurement data, the results will be compared only for the high vehicle speed of 80 km/h. Figure 6.1-6.2 show the comparison in the pantograph output voltage and current respectively. From the figures, it is seen that the simulated results deviated from the measured results. In figure 6.1, the measured voltage has an average value of 650 V, whereas the simulated average voltage is 654 V. In the simulation, most of the components were considered as ideal. Therefore, the power loss was neglected in some cases, which in turn results a bit higher voltage than the measured voltage in the pantograph out. As a result the simulated current is lower than the measured current, which is clearly visible in figure 6.2.

Figure 6.1. Pantograph out voltage for 80 km/h vehicle speed

(39)

Figure 6.2. Pantograph out current for 80 km/h vehicle speed

6.2.2 eMachine characteristics

The comparison between the real test results and the simulated results for the eMachine torque and eMachine speed is depicted in the figure 6.3 and figure 6.4. From figure 6.3 it is seen, that the electromagnetic torque of the eMachine follows the shaft torque. The constant shaft torque 450 Nm is applied to the eMachine to simulate vehicle speed of 80 km/h. The measured torque is very similar to 450 Nm. In figure 6.4, there is some deviation in the measured and simulated speed of the eMachine. For 80 km/h vehicle speed 1350 rpm rotor speed is required. Both of the measured and simulated rotor speeds show quite close speed to 1350 rpm. In reality the vehicle has several velocity in km/h.

Only 80 km/h results were picked for this section to validate with the simulated results.

Figure 6.3. Torque of the eMachine at 80 km/h vehicle speed

(40)

Figure 6.4. eMachine speed at 80 km/h vehicle speed

6.3 Effect of Variable Choke

6.3.1 Performance Index Damping

The effect of the variable choke in the Performance Index will be shown in this section. The Performance Index for different vehicle speed was described in chapter 5. The methodology of controlling the choke inductance according to the vehicle position was also described. The performance of the variable choke will be shown by the following results. There are four different speeds considered both from LPM and HPM. Tables 6.1 describes the results of the choke control for 30 km/h vehicle speed. In the first two rows, the actual distance from the substation is considered and their corresponding PI is picked up. Then, the third row shows the simulated distance from the controller, using the vehicle speed and PI. The distance found from the controller, deviates from the orig- inal distance. This deviation is coming from the average between two points. Then the controller is also able to deliver the current OCL inductance seen from the vehicle side and the additional inductance required from the choke to damp out the high PI. These data are given in the row 4 and row 5 in the table. The effect of the controller is depicted in the figure 6.5 for 30 km/h vehicle speed. The other three PIs damping by choke control is described by the tables A.1 to A.3 and depicted in figures A.1 to A.3.

The following figures show the Performance Index of the system when the vehicle is running at the speed of 36, 45 and 80 km/h speed. The figures show the PI without and with the choke connection. According to the methodology described in the previous chapter, the LPM PI should not exceed a limit of 50 and the HPM should not exceed the limit of 60. The PI, both in the LPM and HPM mode starts to increase rapidly when the vehicle is very close and heading towards the substation. The additional choke is proposed when the PI starts to violate the above limit. In the figures, it can be seen that the PI is damped out sufficiently to follow the limits. Theoretically, the PI remains very constant by controlling the choke. The benefits of such controlled choke will be illustrated in the next section.

(41)

Table 6.1. Choke inductance for 30 km/h vehicle speed

Distance (m) 0 20.83 41.67 62.5 72.9 104.2 145.8 187.5

PI 85.95 81.78 77.38 69.76 66.63 59.32 52.92 48.84

Sim distance (m) 0.2 15.6 46.8 57.3 78 109.4 151 192.7

OCL Inductance (µH) 0.42 32.5 97.5 120 162.5 227.5 314.2 400.8 Choke (µH) 499.5 467.5 402.5 380.8 337.5 272.5 185.8 99.2

Figure 6.5. Damped out Performance Index for 30 km/h vehicle speed

6.3.2 Resonance Shifting

The variable choke effectively works to shift the resonance such as to avoid any high resonance components in the system operating frequency. According to the description in the chapter 4 and chapter 5, we have 5 critical points for the high resonance component in the system frequency of 300 Hz, when the vehicle runs at low power mode. In chapter 5, the additional choke control was described. After applying such control in the choke inductance, the resonance diagrams for the vehicle running at 150 m away from the substation is depicted by the figure 6.6.

The resonance at 80 m, 120 m, 180 m, and 220 m away from the substation are depicted by the figures A.4 to A.7 in the appendix. The resonance are shown without the choke and with the choke control. From the figures, it can be seen that the resonance has sig- nificantly shifted from the higher 300 Hz component to lower 300 Hz component. More specifically, at 150 m away from the substation, the 300 Hz resonance component has shifted from 32 dB to 8.84 dB.

The purpose of such choke control, which in turn shifts the resonance, will be visible in the time domain simulation results in the next subsection.

(42)

Figure 6.6. Resonance shift during LPM at 150 m away from the substation

6.3.3 Time Domain Results

In the time domain model, the graphical codes were used in the MATLAB Simulink.

The simulation was performed by considering 1 MHz sampling time. The pantograph out current and voltage signals are main focus to evaluate the stability margin.

In our current study for ERS, we have 5 critical points when vehicle is running at 30 km/h during the LPM. From the resonance analysis, it was found that at the distances 80 m, 120 m, 150 m, 180 m and 220 m away from the substation, the 300 Hz system frequency component is high. Therefore, the choke control was performed within this range. Finally, the simulation results obtained from the pantograph out at 150 m away from the substation are depicted in the figure 6.7 and figure 6.8, for voltage and current signals respectively.

Figure 6.7. Pantograph out voltage with 30 km/h vehicle speed 150 m away

(43)

Figure 6.8. Pantograph out current with 30 km/h vehicle speed 150 m away

The results are shown in the figures without choke connection and after choke control.

The choke control shows better performance in the signal waveforms. In figure 6.7, the voltage upper peak is damped to 660 V from 663 V and lower peak is damped to 554 V from 552 V. In figure 6.8, the current upper peak is damped to 26.5 A from 30 A and the lower peak is damped to 12 A from 9 A. More results for other distances are shown in the appendix A by figures A.8 to A.15.

When the vehicle was on the high power mode (36 km/h, 45 km/h, 80 km/h), then from the resonance analysis, only 3 critical points (40 m, 80 m and 120 m away from the substation) are figured out. According to chapter 5, we have to consider the reference total inductance at 200 m away from the substation which is 416 µH. The controller will control the choke, such as to keep the overall inductance 416 µH. The pantograph out voltage and current signals during the high power mode at 36 km/h vehicle speed are shown in the figure 6.9 and figure 6.10 respectively at 40 m away from the substation.

Figure 6.9. Pantograph out voltage with 36 km/h vehicle speed 40 m away

(44)

Figure 6.10. Pantograph out current with 36 km/h vehicle speed 40 m away

The voltage signal is damped to 655 V from 652 V for lower peak and 658 V from 663 V for upper peak. In the same fashion, the current amplitude is damped to 40 A from 50 A for upper peak and 24 A from 15 A for lower peak. The results for two other distances are given in the appendix A by figures A.16 to A.19.

The pantograph out voltage and current signals during the high power mode at 45 km/h and 80 km/h vehicle speeds are depicted in the appendix A by figures A.20 to A.31.

As similar to the single vehicle operation at low power mode, we have same critical operating points when we considered two vehicles running at the same time. This is according to the resonance analysis. The simulation results for the pantograph output is shown in figure 6.11 and figure 6.43 for the voltage and current signals, when two vehicles were in operation.

Figure 6.11. Pantograph out voltage with 30 km/h vehicle speed for vehicle 1 and 2

(45)

Figure 6.12. Pantograph out current with 30 km/h vehicle speed for vehicle 1 and 2

One vehicle was in 150 m away and another vehicle was on 300 m away from the substation. For the first vehicle, the same control will be used as before when only one vehicle was considered. As the second vehicle is more than 250 m away from the substation, then according to the resonance analysis, no control is needed. Nevertheless, to show the effectiveness of the control system at higher distances, the choke is also included with the second vehicle. From the figures, very similar responses with and without choke were recorded for the second vehicle.

From the figures, it can be seen that the voltage and current signals are damped out for the first vehicle at 150 m away from the substation. These results are very similar with the single vehicle operation. For the second vehicle at 300 m away from the substation, very similar results are found with and without the choke control.

6.3.4 Harmonic Distortions

After doing the resonance and time domain results analysis, then the harmonics analysis is also done before and after choke implementation in the MATLAB Simulink. The total harmonic distortion (THD) in the current signals obtained from the pantograph output are depicted by figure 6.13 and figure 6.14, for LPM and HPM respectively. The THD is shown for a distance from 80 m to 220 m away from the substation. The figures show, there is improvement in the THD when choke control was implemented.

In figure 6.13, the THD without the choke at 80 m away from the substation is more than 6%. After then, it starts to decrease with the vehicle running away from the substation. At 220 m away from the substation the THD is found less than 4%. When, the choke control was implemented, the THD becomes constant and less than 3.5% for all the distances.

In figure 6.14, it is seen that the THD at HPM is comparatively lower than that for LPM.

THD starts from 3.5% at 80 m away and ends to less than 2% at 220 m away from the substation. After the choke control implementation, the THD is found constant and less than 2%.

(46)

Figure 6.13. Total harmonic distortion when vehicle running at 30 km/h (LPM)

Figure 6.14. Total harmonic distortion when vehicle running at 36 km/h (HPM)

When two vehicles were connected on the line, then the harmonic distortion obtained from the pantograph output signal is depicted in figure 6.15 for LPM. From the figure, it is visible that the harmonic distortion is comparatively lower when choke control is implemented. The second vehicle has even lower THD than the first vehicle which was closer to the substation. If we compare the figures 6.13 and 6.15, then we will see the THD in the first vehicle gets lower when two vehicles are connected on the OCL. The THD without choke connection is less than 5% at 80 m away from the substation. With the choke connection THD is found less than 2%. In the second vehicle, the THD without the choke connection is found less than 4% at 80 m away from the substation and it continues decreasing as the vehicle is moving away from the substation. With the choke control, the THD is found around 1%.

(47)

Figure 6.15. Total harmonic distortion when 2 vehicles are running at 30 km/h (LPM)

(48)

7 Conclusion

The ERS is modelled in the MATLAB Simulink and essential components are included.

The results from the simulation are validated with the real test site measurements and some deviations are found, which are clarified. It is seen from the test site measurement, the instability in the current and voltage signal occurs when vehicle is running at the low power mode.

After then, the frequency domain model is studied to find out the reasons causing the stability problem. The variable inductance seen from the vehicle side, while running, is considered as the main reason behind the stability problem. It is found that the system frequency from the substation is 300 Hz which is equal to the resonance frequency when the vehicle is around 150 m away from the substation. At that very specific point under OCL, the high fluctuation in current is seen from the pantograph out signal in the vehicle.

From the resonance point of view, a high resonance component is found at this frequency.

Therefore, it was proposed to shift the resonance from this frequency to another frequency level by employing a choke before the dc-dc converter inside the vehicle.

The PI based control of the choke shows that, the high PI at the closer distance from the substation is possible to damp out to a marginal value for the improvement of the pantograph out signal disturbances. It is found that, not only the signals of 150 m away from the substation have severe disturbances, but also some other operating points have considerable disturbances.

Therefore, a variable choke is proposed to meet the goal of avoiding the instability for all other conditions. The harmonics analysis is done to check the interruption from another vehicle, when more than one vehicle is connected to the OCL. For this study, same vehicles are considered with a maximum two in number.

The time domain results show that the high amplitude pantograph out signals are damped out sufficiently after the choke control implementation. For the vehicle distances very close to the substation, the effect of the variable choke is more significant.

It is found that the harmonic distortion for the single vehicle connected to the OCL is higher in LPM than HPM. With the choke control implementation, the THD reduces.

When, another vehicle is running under the same OCL, then the harmonic distortion at the first vehicle gets half of the previous distortion. At the second vehicle the harmonic distortion gets even lower.

Therefore, from the simulation results, it is found that the ERS has constructive interac- tion in between the vehicles from the harmonic distortion point of view. Variable choke is further facilitating to this THD reduction.

7.1 Future Work

The following recommendations are made for the future work based on the current thesis work.

• Consideration for more than two vehicles, to simulate more practical environment.

(49)

• Finding the exact position and number of the vehicles to establish a better control.

• Optimization, design and implementation of the variable choke for the upcoming hybrid truck.

System Stability of the Overhead Power Supply System used in the Electric Road System

Examensarbete 30 hp October 2019

System Stability of the Overhead Power Supply System used in the Electric Road System

Md Asif Ferdoush

Masterprogram i förnybar elgenerering

Abstract

System Stability of the Overhead Power Supply System used in the Electric Road System

Sammanfattning

Acknowledgement

Contents

List of Figures

List of Tables

1 Introduction

1.1 Background

1.2 Problem and the aim of study

1.3 Scope and Tasks

1.4 Thesis Outline

2 ERS Components and Operation

2.1 Components

2.2 Operation

3 System Modelling

3.1 Frequency domain model

3.2 Time domain model

4 Resonance and Harmonics Analysis

4.1 Resonance Problem

4.2 Harmonics Problem

5 Control for System Stability

5.1 Performance Index

5.2 Motivation behind Variable Choke

5.3 Selection of the Variable Choke

6 Results and Discussions

6.1 Choke Validation from resonance

6.2 Validation of Simulation

6.3 Effect of Variable Choke

7 Conclusion

7.1 Future Work