IN
DEGREE PROJECT VEHICLE ENGINEERING,
SECOND CYCLE, 30 CREDITS ,
STOCKHOLM SWEDEN 2018
Developing and enriching a
guidance library for the Earth
Observation Satellite MicroCarb
MAXIME THIERRY
1
Master Thesis Report
SD291X – Degree Project in Space Technology
Developing and enriching a guidance library for
the Earth Observation Satellite MicroCarb
Centre National d’Etudes Spatiales, Toulouse
Author:
M. Maxime THIERRY
921202 – T191
CNES supervisor:
Ms. Isabelle SEBBAG
KTH examiner:
M. Gunnar TIBERT
Master Thesis duration:
2
Acknowledgements
First of all, I would like to thank my Supervisor Isabelle SEBBAG. I am very grateful for her availability and her advices, for my Thesis and other subjects. Her understanding and her support helped me to gain confidence, as she gave me responsibilities. My Master Thesis has been enjoyable and formative thanks to her and it has been a privilege being her intern.
I would like to thank Nicolas THERET, who introduced me to the CNES family and recruited me for this Master Thesis. Moreover, I would like to thank him for keeping me up to date with space news, organizing visits and other activities that allowed me to assist to.
Moreover I would like to thank both Miguel MORERE and Nicolas DELAYGUE, who have been very helpful to me in my understanding and assimilation of POLARIS, the guidance laws and other tools already implemented and also during my work on the City acquisition mode.
I would like to thank the whole Guidance and Mission Planning team, who was the embodiment of the Team Spirit and with whom I had enjoyable moments. I am already missing the morning breaks.
I also would like to thank all the interns who were working at the same time as me: Rémi FIEVET, Raphaël FERME, Iñigo FERNANDEZ and Sylvain LUCAS.I really enjoyed the fun moments we had, which made me decompress. But I won’t miss the sounds of the bottles thrown in the bin.
I would like to thank KTH, and more precisely my examiner Gunnar TIBERT and the International Coordinator Elin WILJERGÅRD, who allowed me to put an end to my studies in a good way.
I would like to thank my family, my Mother, Father, Brother and Sister…all of them, whose support allowed me to bounce back after very difficult times.
3
Abstract
My Master Thesis takes place in the context of the MicroCarb mission. The goal of this mission is to identify the sinks and the sources of carbon dioxide on Earth in order to map them and to improve the knowledge of its cycle. To fulfill this mission, some particular guidance modes must be implemented in order to study their feasibility. My thesis consisted in defining and enriching the algorithms used to define the guidance laws, by implementing new tools and a new guidance law, and studying the induced performances in terms of data acquisition and with respect to the constraints related to the satellite. Alongside with this mission, the implementation of those elements support the development of the guidance library POLARIS, actual in its early phase, which is at first only dedicated to MicroCarb but which is intended to become multi-missions.
First, I describe the CNES as well as the guidance team I worked in. Then, the context of the Master Thesis is introduced. Once the context is established we will focus on the first elements I have been working on, as part of the Dazzling studies. Indeed, the spectrometer used in MicroCarb is very sensitive and has to be maintained at very low temperature. Thus the passive cooling mechanism must be protected from the Sunlight and from the light reflected by the Earth. I had to use a class of the Space mechanics library PATRIUS, called Assembly, in order to materialize the satellite and its numerous parts. Once implemented, I was able to perform some Dazzling Studies, highlighting some issues with the various strategies that were considered, and opening new perspectives. Moreover, a problem was detected on a crucial function of the guidance laws calculator. Once a new function was compiled, I had to made a cross validation using Scilab, and results were positive. This part will end with a Geometric Cape study, realized in order to quantify the influence of the satellite, and the MCV roll, over the Geometric Shifting.
In the second part, we will introduce a guidance law which was not implemented initially, and on which I had to work during the last weeks of the thesis: The City mode. Although this mode is similar to an existing calibration mode, it has its own characteristics I had to take into account. The code for this acquisition mode worked well, but the results were not satisfying, considering the Dazzling problem and the kinematic constraints. Thus new strategies had to be considered, and more particularly the 2-scans mode. This mode brought a lot of satisfactions, but there is still more work to be done.
4
Sammanfatting
Examensarbetet fokuserade på rymduppdraget MicroCarb. Målet med detta uppdrag är att identifiera koldioxidsänkor och -källor på jorden för att kartlägga dem och förbättra kunskapen om deras cykler. För att uppfylla detta uppdrag måste vissa specifika styrningsmoder implementeras för att studera uppdragets genomförbarhet. Detta bestod i att definiera och förfina de algoritmer som användes för att definiera siktningslinjer, genom att implementera nya verktyg och en ny styrning samt studera prestandan när det gäller datainsamling och utifrån begränsningar hos satelliten. Detta uppdrag stöder utvecklingen av vägledningsbiblioteket POLARIS, i dess tidiga fas, som i första hand är avsedd för MicroCarb men som också är avsett att användas i flera kommande uppdrag.
Arbetet inleds med en beskrivning av CNES, där examensarbetet utfördes, samt den grupp jag arbetade inom. Därefter presenteras motivation och sammanhanget. Sedan inriktas fokus mot de första elementen jag har arbetat med som en del av de bländande studierna. Spektrometern som används i MicroCarb är mycket temperaturkänslig och måste hållas vid mycket låg temperatur. Således måste den passiva kylmekanismen skyddas mot solljus samt från det ljus som reflekteras från jorden. En klass inom rymdmekanikbiblioteket PATRIUS, kallad Assembly, användes för att modellera satelliten och dess många delar. Därefter utfördes preliminära bländande studier, med fokus på några problem med de olika styrningsstrategier som föreslagits, vilket öppnade nya perspektiv. Dessutom upptäcktes ett problem med en avgörande funktion i siktlinjens räknare. När en korrigerad funktion sammanställts, utfördes en korsvalidering med mjukvaran Scilab, och resultaten var positiva. Denna del avslutas med en geometrisk studie för att kvantifiera påverkan av satelliten och instrumentrullningen på den geometriska skiftningen.
Den andra fasen i arbetet var implementering av en ny funktionalitet kallad Stadsläget. Även om det här läget liknar ett befintligt kalibreringsläge, har det egna egenskaper som måste tas hänsyn till. Beräkningskoden för detta läge fungerade bra, men resultaten var inte tillfredsställande utifrån bländningsproblemet och kinematiska begränsningar. Därför beaktades nya strategier, i synnerhet ett nytt skanningsläget med två avskanningar. Detta läge gav bättre resultat, men behöver utvecklas ytterligare.
5
Acronyms
CNES: Centre National d’Etudes Spatiales
PATRIUS: « Patrimoine de base SIRIUS »
VTS: Visualization Toolkit for Space data
LOF: Local Orbital Frame
MCV: « Miroir de Changement de Visée » (Sight Changing Mirror)
IFOV: Instantaneous Field Of View
FOV: Field Of View
6
Table of contents
Abstract 3
Sammanfatting 4
Acronyms 5
1. Presentation of the company 11
1.1. CNES, The French Space Agency 11
1.2. Four centers of competence 11
1.2.1. Toulouse Space Center (CST) – Toulouse 11
1.2.2. Launchers Direction Center – Paris Daumesnil 11
1.2.3. Guyana Space Center – Kourou 11
1.2.4. Head Office – Paris Les Halles 11
1.3. CNES fields of expertise 12
1.3.1. Launchers 12
1.3.2. Sciences 12
1.3.3. Earth observation 12
1.3.4. Telecommunications 12
1.3.5. Defense 12
1.4. Guidance and mission planning department 13
1.4.1. Attitude guidance 13
1.4.2. Mission planning 14
2. Context and objectives of the Master Thesis 15
2.1. The MicroCarb mission 15
2.1.1. Scientific objectives of the mission 15
2.1.2. General principle of the MicroCarb measures 15
2.1.3. Principal guidance modes 17
2.1.3.1. Nadir mode 17
2.1.3.2. Glint mode 17
2.1.3.3. Target modes 18
2.1.3.4. City mode 19
2.1.3.5. Calibration 19
2.2. Context and objectives of the Master Thesis 19
2.2.1. POLARIS library 19
7
2.2.3. Means 20
2.2.4. Methodology and organization 21
3. Enriching the POLARIS Library 22
3.1. assembly creation 22
3.1.1. the Assembly class 22
3.1.2. the thermal baffle 22
3.1.3. the Star tracker 23
3.2. Dazzling studies on glint mode 24
3.2.1. the « mcv » (sight changing mirror) 24
3.2.2. First tests on glint mode 24
3.2.3. new Series of dazzling studies 28
3.2.4. limits of the models/possible improvements 30
3.3. Geometric shifting studies 31
3.3.1. DefinitioN 32
3.3.2. cross-validation using scilab/celestlab 34 3.3.3. Study of capes: influences on the geometric shifting 35 4. A new acquisition process: The City mode 37
4.1. Theory: principles and characteristics 37
4.2. results with 3 acqusitions 39
4.2.1. first tests/fov deformation 39
4.3. results with two acqusitions 41
4.3.1. results 41
4.3.2. dazzling risks 43
4.3.3. area of improvements 43
5. Conclusion and perspectives 44
6. Bibliography 45
A. Appendix 46
A.1. Frames and attitude conventions 46
A.1.1. MicroCarb satellite frame 46
A.1.2. Local orbital frame 46
A.1.3. Attitude angles 46
A.1.4. Baffle frame and fictive sensors 47
A.2. reference angles for the dazzling studies 47
8
A.4. Thermal Baffle and star tracker’s guard angles 48
A.5. orbital parameters (𝛽min, 13h30 lhan) 48
A.6. IFOV and FOV 49
9
Table of illustrations
Figure 1: General inputs and outputs of attitude guidance activities ... 13
Figure 2: Programming loop in mission planning ... 14
Figure 3: Measurement process of MicroCarb ... 16
Figure 4: CO2 fluxes computation process ... 16
Figure 5: Nadir pointing ... 17
Figure 6: Glint pointing ... 18
Figure 7: Target pointing ... 18
Figure 8: Short-term organization of the POLARIS library ... 19
Figure 9: Long-term organization of the POLARIS library ... 20
Figure 10: Thermal baffle ... 22
Figure 11: Baffle's guard angles... 23
Figure 12: Star tracker ... 23
Figure 13: Principle of the "MCV" ... 24
Figure 14: Sun dazzling studies: Angle between the center of the Sun and the sight axis of the Thermal baffle, for different MCV angles ... 25
Figure 15: Satellite track (solid line) and Glint point track (dotted line) ... 26
Figure 16: Earth dazzling studies: Angle between the center of the Earth and the sight axis of the Thermal Baffle, for different MCV angles ... 27
Figure 17: Histogram of the possible MCV roll values, with respect to the previous graphs ... 28
Figure 18: new Sun Dazzling studies: Angle between the center of the Sun and the sight axis of the Thermal baffle, for different MCV angles ... 29
Figure 19: new Earth Dazzling studies: Angle between the center of the Earth and the sight axis of the Thermal baffle, for different MCV angles ... 30
Figure 20: Baffle's notch ... 31
Figure 21: Geometric Shifting ... 32
Figure 22: MCV effect on the Geometric Shifting ... 32
Figure 23: Roll/Pitch effects on the Geometric Shifting ... 33
Figure 24: Earth rotation's effect on Geometric Shifting ... 33
Figure 25: Geometric Shifting comparison ... 34
Figure 26: Study of Caps - GS induced by platform and MCV roll ... 36
Figure 27: Example of acquisition for a City mode ... 37
Figure 28: CalVal Target representation. The City acquisition acts in a similar way [1] ... 38
Figure 30: Other City projection, displayed with Google Earth ... 40
Figure 31: Results of a 2 acquisitions-City mode, displayed on Google Earth (Geometric Shifting compensated) ... 42
Figure 32: Baffle frame ... 47
Figure 33: Sight axis used for the Sun and Earth fictive sensors for the Assembly ... 47
Figure 34: IFOV dimensions (m) at Nadir ... 49
Figure 35: FOV dimensions (m) at Nadir ... 49
10
Table of tables
11
1.
Presentation of the company
In this part, I will introduce CNES, its organization, an overview of the different CNES sites and its principal fields of work. I will also describe the Guidance and Mission-Planning department, in which my Master Thesis has been performed.
1.1.
CNES, THE FRENCH SPACE AGENCY
CNES (Centre National d’Etudes Spatiales) was created in 1961 by the French president Charles de Gaulle. It is a state-owned industrial and commercial establishment and has a significant role in the space industry at a national, European and international level. CNES represents France at the European Space Agency (ESA) council and has a strong influence thanks to its technical expertise and its scientific research. Thus, it is a reference partner for a large number of space companies and scientific laboratories who call on CNES to participate in national and international projects.
1.2.
FOUR CENTERS OF COMPETENCE
Today, CNES employs around 2500 employees and is spread over four sites. It provides expertise in every field of the space industry and can take action anytime during a space mission. [9]
1.2.1. TOULOUSE SPACE CENTER (CST) – TOULOUSE
The Toulouse Space Center, or CST, is the main technical and operational center. This site covers the majority of technical tasks including projects management, research studies, operation centers for LEOP (Launch and Early Orbit Phase) and orbit management, IT activities, administration, logistics and communication. Activities are principally centered on orbital systems. The Toulouse Space Center was opened in 1968 and counts now more than 1700 employees. [9]
1.2.2. LAUNCHERS DIRECTION CENTER – PARIS DAUMESNIL
The Launchers Directions Center has been developing all Ariane launchers since 40 years. In collaboration with the European Space Agency, it is also in charge of the development of new generation launchers like Ariane 6. Approximately 220 people work at Paris Daumesnil’s site. [9]
1.2.3. GUYANA SPACE CENTER – KOUROU
The Guyana Space Center in Kourou was created in 1964. It is considered as the European “space harbor” where all three European launchers are launched. Its geographical position, close to the equator, confers to the Kourou’s site a safe and reliable launching base. This site employs around 270 people. [9]
1.2.4. HEAD OFFICE – PARIS LES HALLES
12
1.3.
CNES FIELDS OF EXPERTISE
With the second worldwide budget for space programs (2.44 B€ in 2018 [9], behind NASA) and successful projects, France has the resources to contribute and invest in various fields of expertise.
1.3.1. LAUNCHERS
Autonomy of access to space is appreciated for the sovereignty, guaranteed by the range of European launchers. We count three principal launchers: Ariane 5, Soyuz and Vega. The new launcher Ariane 6, currently in development, will allow re-ignition of its second stage; it will be at least half cheaper than Ariane 5 and will be a first step to compete with other “New Space” actors.
1.3.2. SCIENCES
Space systems are privileged exploration tools for astronomers, astrophysicists and planetologists. CNES played a major role in Rosetta and Philae’s success in apprehending the origin of the solar system. The French agency is also involved in many Martian missions: after having equipped Curiosity with ChemCAM and SAM instruments, CNES built the seismograph of the InSight mission, on its way to Mars. In other fields, one can also cite MICROSCOPE, whose aim is to verify one of the Einstein’s general relativity theory fundament or Euclid for the study of the universe expansion.
1.3.3. EARTH
OBSERVATION
The study of the Earth and its environment from space is at the heart of the scientific and social issues of the 21st century. CNES was very early involved in these operational observation missions. By providing reliable and continuous observation data, complementary to information from sensors on the ground or at sea, space systems designed by CNES or with its assistance are needed in all areas: operational oceanography, climate studies and meteorology, study of continental surfaces, solid Earth. MicroCarb would be part of this category.
1.3.4. TELECOMMUNICATIONS
CNES is at the heart of the rise of the digital economy. Its role: bringing out and validating the space technologies that manufacturers and operators will implement. The current challenge is the provision of very high speed satellite broadband, especially in rural areas. It also concerns satellite navigation technologies with the Galileo constellation as an example, or environmental and security data collection thanks to Argos or Cospas-Sarsat.
1.3.5. DEFENSE
13
1.4.
GUIDANCE AND MISSION PLANNING DEPARTMENT
The guidance and mission planning department is part of the Flight Dynamics sub-directorate (DSO/DV/MP) in Toulouse space center. It has the following general responsibilities:
Conduct studies on the compromise between mission planning, mission design, instrumentation, performance in terms of acquisition capacity and system operation;
Define and validate the operational and optimization strategies;
Define and develop the means to assess the effectiveness and performance of planning and attitude guidance functionalities;
Establish the specifications linked to planning and spacecraft or equipment orientation, which are followed by the ground and space segments developers.
1.4.1. ATTITUDE GUIDANCE
The figure 1 represents the general inputs and outputs of the attitude guidance activities.
Figure 1: General inputs and outputs of attitude guidance activities
Guidance modes are computed as much as possible on board. Among these modes we can find:
Waiting phases (geocentric pointing, heliocentric pointing, etc.);
Calibration modes;
Acquisition modes;
Slew maneuvers.
However, the guidance can be computed on the ground under some constraints; as an example, when there is a high need of a very accurate prediction for the guidance or when the onboard computation capacity is insufficient with respect to sophisticated guidance algorithms.
Pointing
Mission
Pointing modes Temporal
14
1.4.2. MISSION PLANNING
The mission planning is in charge of:
Taking into account the requests inputted by the system users;
Translating these requests into system-interpretable requests;
Decomposing these requests into elementary requests;
Planning each elementary request and its downloading with respect to the programming constraints: agility, electrical consumption, request priority, memory, station visibility, telemetry rate, etc.
Planning the functional activities of the spacecraft equipment involved in the request fulfillment;
Uploading the elementary request to the satellite and executing them;
Receiving the telemetry on ground;
Taking into account the feedback on the requests realization.
The mission planning works according to a Programming loop, represented in the figure 2:
Figure 2: Programming loop in mission planning
15
2.
Context and objectives of the Master Thesis
This part is dedicated to the presentation of the MicroCarb mission with a focus on the aspects that are the most related to the master thesis missions.
2.1.
THE MICROCARB MISSION
2.1.1. SCIENTIFIC OBJECTIVES OF THE MISSION
The scientific objective of the MicroCarb mission is the monitoring and characterization of CO2 fluxes on the surface, thus exchanges between the sources (natural or anthropogenic) and the wells formed by the atmosphere, the ocean, the soils and the vegetation.
The global annual flows of CO2 represent a quantity of the order of 200 Giga tons of carbon. Emissions related to human activity add about an additional 10 Giga tons, disturbing the natural balance. Half of this surplus is absorbed by vegetation, soils and oceans, while the other half causing the increase in the atmospheric concentration of greenhouse gases. This contributes to global warming, and accentuates its effects (natural disasters, etc.). [8]
A better knowledge of the CO2 flux is necessary to:
Improve knowledge of source and sink exchange mechanisms, their seasonal variability, and their evolution in response to climate change,
Identify the parameters that control carbon trading,
Validate and improve (by reducing uncertainty) the life cycle models of carbon.
The project was started by French initiative after the COP21 conference for the future preparation for a European operational project of CO2 monitoring in 2030. The objective date of the launch of MicroCarb is 2021. [8]
2.1.2. GENERAL PRINCIPLE OF THE MICROCARB MEASURES
16
Figure 3: Measurement process of MicroCarb
Each MicroCarb measurement allows acquiring a local concentration of carbon dioxide. These spectral luminance measurements will be converted to integrated column CO2 concentrations by a spectrum inversion method to yield the geophysical products. The computation process is giving in figure 4.
17
2.1.3. PRINCIPAL GUIDANCE MODES
In order to perform the measurements, MicroCarb must ensure different guidance and calibration modes, depending on the targeted area. [7]
2.1.3.1.
NADIR MODE
The nadir mode (figure 5) is the reference mode for the mission. It is not constraining for the satellite and its equipment and it is easy to realize: The Earth port is pointing towards the nadir direction. This guidance mode is interesting for measurements over land areas. A scan mechanism allows a steering of the line of sight, favoring the acquisition of uncorrelated measurements.
Figure 5: Nadir pointing
2.1.3.2.
GLINT MODE
18
Figure 6: Glint pointing
2.1.3.3.
TARGET MODES
Some grounds stations are dedicated to measure the CO2 concentration in their own area (TCCON stations); these measures may then be compared to MicroCarb measures and used to calibrate the instrument. The Earth port points then towards a TCCON station and can keep pointing at it during the entire acquisition duration or perform several sweeps over a particular area around the TCCON. In the first case, the guidance mode can be mentioned as “Fixed target” (figure 7). In the second case, it is called “CalVal Target”.
19
User request
layer
2.1.3.4.
CITY MODE
The City Mode uses the same principles as the CalVal Target mode, except that the area to be scanned covers an area of approximately 40X40 km², centered on a city. The spatial resolution along the track (ALT) must be 2 km.
2.1.3.5.
CALIBRATION
The measurement instrument needs to be regularly calibrated. Indeed, the atmosphere composition can be locally variable and must be taken into account. Then, some calibration modes are defined. These calibrations may be done pointing towards the Sun, the Moon, the deep space or even a particular area on Earth.
2.2.
CONTEXT AND OBJECTIVES OF THE MASTER THESIS
2.2.1. POLARIS LIBRARY
During my Master Thesis, I contributed to the development of a guidance library called POLARIS. The idea of POLARIS is to develop an autonomous and easy to integrate deliverable for the PLOC (PayLoad Operation Center) segment. Amongst others, it will have to handle guidance law computation, usual dazzling management or slew maneuvers and is in the first place entirely dedicated to MicroCarb.
The development of the library relies on other existing libraries:
PATRIUS: this library is dedicated to guidance and orbital mechanics. It extends the free library OREKIT
which is entirely dedicated to orbital mechanics. It has been developed by Thales Services and is coded in JAVA.
MANIAC: this library is dedicated to optimal slew maneuvers computation, which is the transition
between two guidance laws. MANIAC is coded in C language and has been developed by CS
Communication & Système.
In the long term, the library will be adapted for a multi-mission context, where MicroCarb represents the first part. The figures 8 and 9 represent, on the one hand, the current state of the library and, on the other hand, its future evolution. POLARIS will also have to be integrated to a simulation environment called ALIS (Simulation Framework for Mission Engineering), which gathers simulators for both guidance and mission planning.
20
Figure 9: Long-term organization of the POLARIS library
2.2.2. OBJECTIVES
The main objective of this work was to define and enrich the algorithms used to define the guiding laws and to study the induced performances in terms of acquisition and respect of the constraints related to the satellite. By relying on PATRIUS, I developed study tools, and new functions to the existing guidance library for the MicroCarb satellite, to launch simulations and to analyze the results. In the same time, I took part in the development of the POLARIS library as the tools and new guidance mode were implemented in it.
While implementing the tools, some unforeseen events forced me to change my priorities. For example, as one of the functions in POLARIS was defective, I had to cross validate a new version and ensure it was operational again. Thus my Thesis rhymed with the word Adaptation, as my work evolved with the new priorities for the mission.
2.2.3. MEANS
The development of the guidance and post-processing algorithms were done in JAVA within the Eclipse environment.
2.2.3.1. PATRIUS
PATRIUS is the library on which every coded algorithm relies. It contains several JAVA classes and methods which allow manipulating amongst others orbital mechanics, orbit propagation, guidance laws, etc.
2.2.3.2. VTS (VISUALIZATION TOOLKIT FOR SPACE DATA)
VTS is a visualization tool that allows importing orbit and attitude data. It is useful to verify the relevancy of the attitude guidance law or some orbital events like eclipse or dazzling. VTS does accept only a particular file format, that can be assimilated to a text file and that can be created in outputs of simulations.
User request Project X
layer
Specific project X
User request
layer
User request Project Y
20
2.2.3.3. SCILAB/CELESTLAB
Scilab is a numerically oriented programming language. The syntax of Scilab is similar to MATLAB, and Scilab includes a source code translator for assisting the conversion of code from MATLAB to Scilab. Data from text files (similar to the one used for VTS) can be used during computations. Celestlab is a toolkit based on Scilab, developed by CNES, implementing Space Mechanics and Flight Dynamics for Mission Analysis.
2.2.3.4. KML FILES, VISUALIZATION WITH GOOGLE EARTH
Results from simulations can be compiled into Excel files, in addition to text files. Then these Excel tables can be converted into KML files (Keyhole Markup Language). These last files are then used in Google Earth for 3D visualization.
2.2.4. METHODOLOGY AND ORGANIZATION
My initial approach was to familiarize with the MicroCarb mission, by gathering all the information regarding the satellite, its orbits, existing acquisition modes and the status of the POLARIS library. As the dazzling studies had not been taken in consideration before, it was a good starting point to work on the library.
In order to not interfere with the work done in POLARIS, by overwriting bad data for example, I started working in a personal workspace, where I still could import the current version of the library, test my new elements, and later commit them once they were verified and efficient.
I also had to get along with the previously cited library and software like PATRIUS and VTS. I had not been in touch with such library in my studies before, and I had to learn some key points. Also I had to find my notions of Matlab coding while working with Scilab/Celestlab.
22
3.
Enriching the POLARIS Library
In this part, I will present the work done about the implementation of new tools the Polaris library, concerning the Dazzling studies in particular. Also will be discussed the validation of a new function for the Geometric Shifting estimation. First, a presentation of the mission parameters is done, as well as the assumptions made.
3.1.
ASSEMBLY CREATION
3.1.1. THE ASSEMBLY CLASS
The PATRIUS library allows the user to define relations between specified frames. It is very interesting for us as there are multiple elements of MicroCarb, and it is implicitly possible to define a model of satellite and its components by their frames. The main concept used for this model is called the “Assembly”: it is the JAVA class from PATRIUS that allows materializing a satellite, its parts, their associated characteristics from the definition of their frames and relations between them.
In this phase are taken into account:
The spacecraft frame, lately named “Main body”;
The Thermal Baffle, as well as all its characteristics: The Sun and Earth fictive sensors, their respective sight axes, their respective fields of view and their clearance angles with respect to Sun or Earth;
The solar port, defined by a sight axis in the spacecraft frame. This port is used for the Solar Calibration mode, which is not part of this study;
The satellite’s Star Tracker.
Thanks to this class, the MicroCarb satellite and its parts can be modeled and used for orbit propagation or guidance computation, and allow us to control and analyze its performances, particularly concerning the Dazzling events on the Thermal Baffle and the Star Tracker. This function has been created in the most generic way, so that it can fit in all the current acquisition/calibration modes and these to come.
3.1.2. THE THERMAL BAFFLE
The optical instrument is very sensitive to temperature variation, as it has to be maintained at a very cold and stable temperature to be operable. In order to passively protect the instrument against heat sources [7], a thermal baffle (figure 10) is installed on MicroCarb. Its goal is to protect the instrument against both Earth thermal fluxes and solar fluxes.
23
The bottom of the baffle is a radiator. The highest side of the cone protects the radiator from the sun (this part is always directed to the sun) when the smallest side protects the radiator from the Earth. The thermal baffle has then two guard angles (see Appendix A.2 and figure 11) such as:
Sun fluxes must not enter the opening of the baffle
Earth fluxes must not enter the opening of the baffle with a certain tolerance.
The geometry of the Thermal Baffle and the guard angles are given in the Appendix A.1.4 and A.4.
3.1.3. THE STAR TRACKER
A star tracker (figure 12) is an optical device that measures the positions of stars using photocells or a camera. As the positions of many stars have been measured by astronomers to a high degree of accuracy, a star tracker on a satellite may be used to determine the orientation (or attitude) of the spacecraft with respect to the stars.
Figure 12: Star tracker
In the same way as the Thermal baffle, guard angles have been determined to protect the Star Tracker from both the Sun and the Earth fluxes.
Earth guard angle
Sun guard angle
24
3.2.
DAZZLING STUDIES ON GLINT MODE
3.2.1. THE « MCV » (SIGHT CHANGING MIRROR)
Before looking at dazzling studies themselves, let us focus on a key element of the guidance laws calculator. The Earth port, previously mentioned in this report, is actually not directly pointing towards the Nadir direction when the satellite is in Nadir pointing. The Earth port is positioned on 𝑋𝑠𝑎𝑡 and points towards the direction of the velocity (+𝑋𝑠𝑎𝑡), but the MCV brings back to the instrument all the light that comes on the 𝑍𝑠𝑎𝑡 face of the satellite, as illustrated in figure 13. [5]
Figure 13: Principle of the "MCV"
This equipment, the MCV, is commanded in roll around 𝑋𝑠𝑎𝑡. It is mounted on the satellite because some guidance modes request the possibility to point some off-track targets: the roll is then done by the MCV. The MCV is tilted to 45° in order to, in nadir pointing, have a reflected light beam parallel to the 𝑋𝑠𝑎𝑡 axis. However, when the mirror is subject to a rotation in roll, the line of sight is subject to the same rotation in roll and this tilt induces an equivalent yaw rotation around the new line of sight. This effect will be discussed later.
The MCV itself is not part of the Assembly, but has its own specific class in POLARIS, and is used directly during the computation of the satellite guidance.
3.2.2. FIRST TESTS ON GLINT MODE
25
I started working in a personal workplace, within the Eclipse software, so that I could not pollute the POLARIS library with new, non-tested, programs. However I was able to import some key classes, concerning the guidance laws and the data files generators. Once I implemented the Assembly Class, generating a model of MicroCarb and its components, I updated the data files generator to create files about the dazzling studies, returning dazzling angles from the Sun, the Earth, and a Truth table saying if the Thermal Baffle is dazzled or not.
For different roll targets of the MCV, the behavior of the satellite change. Several simulations were made, and the data combined to generate relevant graphs. The following graph (figure 14) illustrates the angle between the Sun and the sight axis of the Thermal Baffle (simulated), with respect to the roll target of the MCV. If the computed angle is lower than the reference angle (see Appendix A.2), then the Sun is entering the detection area and the Baffle is dazzled.
Figure 14: Sun dazzling studies: Angle between the center of the Sun and the sight axis of the Thermal baffle, for different MCV angles
We can observe that for a roll target for the MCV above +10°, the Sun in dazzling the Baffle around the middle of the acquisition, near the sub-solar point. Moreover, the higher the MCV roll target, the longer the Sun is entering the Thermal Baffle. This can be explained by the position of the glint point with respect to the satellite track, as shown in the figure 15 for the case of a local time at the ascending node of 1:30 pm.
0 20 40 60 80 100 120 140 160 180 200 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49
Sun Dazzling Studies - Glint with fixed Mcv
26
Figure 15: Satellite track (solid line) and Glint point track (dotted line)
As we can see, he Glint point is always situated on the left side of the Satellite track, and requires a positive roll to be targeted. If the MCV roll angle exceeds the requested angle, the satellite needs to perform a negative roll to compensate. However, due to the configuration of the Thermal baffle on the platform and its orientation, a negative roll of the satellite leads to an increased risk of Sun dazzling events.
27
Figure 16: Earth dazzling studies: Angle between the center of the Earth and the sight axis of the Thermal Baffle, for different MCV angles
This time, Earth seems to dazzle the Baffle at the beginning of the acquisition, regardless the value of the roll target. Only a +35° roll prevents the dazzling. However, the lower the MCV roll, the longer is the dazzling period at the end of the acquisition. Thus, a strategy implying a constant roll target all over the guidance mode would not be considered here, as a continuous variation of the roll target seemed mandatory to prevent any dazzling event. A new strategy that could be considered would be starting with a high MCV roll (+35°) to prevent the Earth dazzling. As we approach the sub-solar point, the roll target should be decreased to +5° to avoid the Sun fluxes and, at the end of the acquisition, it should come back to +20°.
The Histogram In figure 17 gives us a good idea of this strategy. The non-dazzling periods (when the histogram is 1) take place at different time of the acquisition and last more or less long, depending of the dazzle from the Earth or from the Sun.
0 20 40 60 80 100 120 140 160 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49
Earth Dazzling Studies - Glint with fixed MCV
28
Figure 17: Histogram of the possible MCV roll values, with respect to the previous graphs
As the Assembly implementation and the additions to the files generator worked well, the first dazzling studies offered interesting results. A constant roll target strategy seems not to work, although a continuous variation of the roll would imply other perturbations like vibrations, and thus interfere with the collection of data. An issue concerning one of the calculation functions of POLARIS led me to stop this first series of dazzling studies. More information about this issue is available in sections 3.3.1 and 3.3.2. Once the problem was solved, a new series of studies were conducted and the results are given in the next part, 3.2.3.
3.2.3. NEW SERIES OF DAZZLING STUDIES
Once the problem mentioned above was corrected, the new Dazzling studies took over, using the same process despite some minor patches and adjustments.
I analyzed numerous orbital contexts, that differ according to the date of the acquisition (when the angle 𝛽 is minimum or maximum) and the local time at the ascending node (1:30 pm or 10:30 pm). I kept this theoretical hypothesis to have guidance with constant MCV roll target. At that point we considered two possible strategies:
At the beginning of the guidance, when the Solar Zenith Angle is at maximum (SZA max) with respect to the satellite, we managed to fix the MCV roll so that the satellite roll would be 0. This initial roll angle would be kept constant during the acquisition, as the satellite would correct the attitude to keep the Glint point in sight.
At the sub-solar point, when the Solar Zenith Angle is at minimum (SZA min) with respect to the satellite, we proceeded in the same way. This gave us a second possible value for the MCV roll to be tested. Mcv = -20° Mcv = 10° 0 0,2 0,4 0,6 0,8 1 4378,15082 4558,15082 4738,15082 4918,15082 5098,15082 5278,15082 5458,15082 5638,15082 5818,15082 5998,15082 6178,15082 6358,15082 6538,15082 6718,15082 6898,15082 7078,15082 7258,15082
29
One of the important points to emphasize during these new analyzes is the similarity of behavior between the previous attitudes and the new ones. Indeed, Sun dazzling events still occurred in the middle of the acquisition while Earth dazzling events occurred at the beginning or at the end.
However, the major changes concern the MCV angles for which the dazzle occurred. While the previous satellite rolls were around 30/40°, the rolls turned around 10° after the Geometric Shifting function changed, highlighting the upset this change brought. This evolution regarding the satellite roll allowed us a greater flexibility with respect to the MCV roll target.
The following graphs illustrate this statement. On the one hand, the first graph (figure 18) seems to be similar to the previous one, where an important MCV roll angle leads to dazzling near the sub-solar point. We can already say that the strategies mentioned previously are compromised.
Figure 18: new Sun Dazzling studies: Angle between the center of the Sun and the sight axis of the Thermal baffle, for different MCV angles
The most important elements are concerning the Earth dazzling events. Due to the change in the satellite attitude, all the curves have been raised so that the dazzling events are nearly non-existent for the previous MCV roll targets. In order to make the dazzle occurred; the roll had to go below -5°, as represented on the graph below (figure 19):
0 20 40 60 80 100 120 140 4378,15082 4468,15082 4558,15082 4648,15082 4738,15082 4828,15082 4918,15082 5008,15082 5098,15082 5188,15082 5278,15082 5368,15082 5458,15082 5548,15082 5638,15082 5728,15082 5818,15082 5908,15082 5998,15082 6088 ,1508 2 61 78 ,1508 2 6268,15082 6358,15082 6448,15082 6538,15082 6628,15082 6718,15082 6808,15082 6898,15082 6988,15082 7078,15082 7168,15082 7258,15082
Sun Dazzling Studies - Glint with fixed MCV
Angle Référence Mcv SZA min = 13,410493755242559°
30
Figure 19: new Earth Dazzling studies: Angle between the center of the Earth and the sight axis of the Thermal baffle, for different MCV angles
These new results opened new opportunities we did not obtained at the end of the first studies. Instead of a strategy implying a continuous change in the MCV roll, we could finally consider a fixed MCV strategy. However, fixing the MCV roll at SZA min or SZA max seemed not possible as they induced Sun dazzle. Instead, a range of possible MCV roll is possible, more or less large depending on the orbital context considered. Table 1 gives us the possible ranges.
Table 1: Possible values for the MCV roll
The minimal angle is rather determined by the Earth dazzling events, and the maximum one by the Sun dazzling events. As the Sun is farther away from the orbital plane at 𝛽max than at 𝛽min, the range is larger in the first case. Thus, a Glint acquisition mode can be considered at any period of the year.
3.2.4. LIMITS OF THE MODELS/POSSIBLE IMPROVEMENTS
These previous results were obtained using a simplified version of the Thermal baffle, used in the Assembly. Some modification and improvements, discussed during frequent meetings, have to be considered in order to obtain data closer to the reality:
0 20 40 60 80 100 120 140 4378,15082 4468,15082 4558,15082 4648,15082 4738 ,1508 2 4828,15082 4918,15082 5008,15082 5098,15082 5188,15082 5278,15082 5368,15082 5458,15082 5548,15082 5638,15082 5728,15082 5818,15082 5908,15082 5998,15082 6088,15082 6178,15082 6268,15082 6358,15082 6448,15082 6538,15082 6628,15082 6718,15082 6808 ,1508 2 6898,15082 6988,15082 7078,15082 7168,15082 7258,15082
Earth Dazzling Studies - Glint with fixed MCV
Angle Référence Mcv SZA min = 13,410493755242559°
Mcv SZA max = 15,25781159667203° Mcv = 10° Mcv = 5° Mcv = 9° Mcv = 0° Mcv = -5° MCV SZA max = 15,25781159667203° MCV = 5° MCV = 0°
Reference angle MCV SZA min = 13.410493755242559°
31
The sun guard angle was considered too large. The lower the guard angle, the lower the Sun Dazzling risk will be.
The Baffle is represented principally as a hollow cylinder. However a notch, illustrated in the figure 20, is missing. The model of such notch was considered difficult in the first studies, as some dimension data from the Baffle are missing. These data would be provided later by Airbus Defense and Space.
Figure 20: Baffle's notch
Other dazzling studies should also be considered. The first one is the dazzling of the Earth port, which could be directly exposed to the Sun at high pitch configuration (beginning of the Glint mode for example).
As we considered the reflected fluxes from the Earth in the study, the reflected fluxes from the Moon could also be a possibility, even if the influence of such fluxes would be largely limited compared to the Sun and the Earth.
3.3.
GEOMETRIC SHIFTING STUDIES
While the Dazzling studies were in progress, a problem occurred with a key function of the guidance laws. We discussed earlier that for the Glint mode, an effect called Geometric Shifting (GS) should be minimized, or even canceled if possible. A function implemented in POLARIS was supposed to compute the GS and adapt the satellite attitude in order to compensate it.
However, this function was malfunctioning, and the attitudes generated were not corresponding to GS-compensated ones. This initial function was based on an analytical formula, implying several frame changes, and we suspect an error coming from these changes. This problem was not detected earlier due to the fact that the function seemed to work well around the Nadir attitude and the first guidance modes were using this particular attitude.
Thus a new function had to be developed, and had to be tested. I was asked to validate the new function using another software, Scilab and its Space Mechanics toolbox Celestlab. The idea was to validate the function with different tools, independent from POLARIS.
32
3.3.1. DEFINITION
The Geometric Shifting at time 𝑡, illustrated with figure 21, is defined as the angle between [4]:
The normal to the ground projection of the IFOV (see Appendix A.6) length at time 𝑡;
The ground velocity of the targeted point
Figure 21: Geometric Shifting
The geometric shifting induces errors on the measurement. Therefore, it has to be compensated during ground pointing modes as much as possible. There are three main sources of geometric shifting, as illustrated below:
Figure 22: the MCV roll
Figure 23: The MicroCarb attitude, particularly Roll and Pitch
Figure 24: The Earth rotation while scanning the ground
Figure 22: MCV effect on the Geometric Shifting LOF
LOF
33
Figure 23: Roll/Pitch effects on the Geometric Shifting
For a given roll, given to the MCV or the satellite itself, we can observe that the Geometric Shiftings induced are in opposite directions.
Figure 24: Earth rotation's effect on Geometric Shifting LOF
LOF
LOF
34
3.3.2. CROSS-VALIDATION USING SCILAB/CELESTLAB
The validation protocol I used required the use of some specific data files from a guidance simulation from POLARIS: the files containing the position, velocity and the attitude during the acquisition. These data were supposed to arise from a GS-compensated acquisition, it only remained to analyze and confirm (or not) the GS compensation.
Thanks to the mentioned files, I was able to compute on Scilab some precise sight vectors from the Earth port, and then the targeted points, as the intersection of these vectors with a model of the Earth, available with Celestlab. A Time step of 0,01 s between two sets of points was required to obtain sufficient precision. In order to reduce the calculation time, the acquisition period had been reduced to 1 s.
The targeted points, required for the study, are those so that we can obtain:
The corners of the aperture bigger side for each time step
The center of the aperture. Once these points are computed, each separated by 0,01 sec, they gave us a good estimation of the ground speed of the aperture.
Also the Earth rotation had been taken into account, as it was also a main cause of the Geometric Shifting. The graph in figure 25 represents both Geometric Shiftings, from the POLARIS simulation and from the Scilab protocol.
Figure 25: Geometric Shifting comparison
We can observe the two curves are relatively similar on average. Some iteration process implemented in POLARIS made its curve smoother. Nevertheless the difference between the curves is about 0,001 deg in order of magnitude, maybe due to differences in the models used on POLARIS and Scilab. The tolerance that was allowed to POLARIS simulation explains the POLARIS GS value (about 0,0135 deg instead of 0 deg). With respect to this tolerance, the difference is negligible.
0,0115 0,012 0,0125 0,013 0,0135 0,014 0,0145 0,015 0,0155 5060,933452 5060,963453 5060,993452 5061,023453 5061,053452 5061,083452 5061,113452 5061,143452 5061,173453 5061,203452 5061,233453 5061 ,2634 53 5061,293452 5061,323452 5061,353452 5061,383453 5061,413452 5061 ,4434 53 5061,473453 5061,503452 5061,533452 5061,563452 5061,593453 5061,623452 5061,653453 5061,683453 5061,713452 5061,743452 5061,773453 5061,803453 5061,833452 5061,863452 5061,893453
GS Test on Scilab - Validation
35
Several simulations of that kind were released, at different part of the Glint acquisition process (1 s duration each, 0,01 s time step), and we obtained similar results.
The validation process is thus successful. The new Geometric Shifting function is now fulfilling its role.
3.3.3. STUDY OF CAPES: INFLUENCES ON THE GEOMETRIC SHIFTING
We discussed earlier that there are three main sources of Geometric Shifting. As we can estimate the shifting induced by the Earth rotation (which reaches a maximum of 3° at the equator, in Nadir configuration), we do not know what the contribution of the other two is: the MCV roll target and The MicroCarb attitude, i.e. its roll.
A better understanding of these two contributions is important, as we can choose more adequately the MCV roll target, among the validity range we discussed just above as an example.
First Study:
Using Scilab, I proceeded to a first preliminary study. For a given orbit with a Nadir attitude as a starting point, I modified the attitude in Pitch (0° to +50°) and Roll (0° to +30°) after giving the MCV a fixed roll target (at 0°, +10° and +30°). The Yaw is fixed at 0°. The objective was to target the same points using different roll target for both MCV and satellite.
For a given targeted point the larger the MCV roll, the larger the induced Geometric Shifting. The contribution of the MCV roll is far more important than the one of an equivalent roll from the satellite.
This first study brought some interesting results, and has to be confirmed with another study. Using Scilab again, and thanks to some data files from a Glint simulation, I performed a study of geometric capes which would give us the influence on the Geometric Shifting for an entire acquisition period.
Study of geometric capes:
The protocol was as follows:
Realization of Two Glint Simulations, with a MCV roll fixed at 0° and generation of data files. These simulations are shifted in time by 0,01 s. Among the generated files, the attitude file was in the “TRL” convention (Pitch, Roll and Yaw) in the Local Orbital Frame.
Thanks to these shifted-in-time simulations, I was able to calculate the geometric cape of the Glint point over the acquisition period.
As the MCV roll is fixed at 0°, the satellite Yaw is used to compensate the total Geometric Shifting. On Scilab, I fixed the Yaw at 0° and conserved the other attitude values. Then I calculated the corners of the bigger side of the aperture, and the cape of the orthogonal direction of this side over the period.
The gap between the two previous calculated capes led me to the Geometric Shifting due to satellite pitch and roll angles during the total period of acquisition (Purple curve on the graph below, figure 26).
36
Figure 26: Study of Caps - GS induced by platform and MCV roll
This study of capes confirmed the results of the previous study, as a MCV roll induces more GS than an equivalent satellite roll and is the main contributor to the total GS. This induced GS, exceeding 30° at the beginning and the end, is far from the acceptable GS for a Glint acquisition mode.
Thus a MCV roll target has to be chosen wisely, regarding the possible MCV roll we obtained in the last dazzling studies for Glint as an example.
-20 -10 0 10 20 30 40 50 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97
Study of capes - BetaMin 2021 - 13h30
37
4.
A new acquisition process: The City mode
The City mode is an important acquisition mode for MicroCarb. It will make possible the measurement of the CO2 concentration above the world greatest metropolises, and give an idea of the influence of the human activities on greenhouse gas and global warming. The City mode was not implemented in POLARIS at the beginning of my work, and my work consisted in laying the first stones of this new mode.
4.1.
THEORY: PRINCIPLES AND CHARACTERISTICS
The City mode [6] meets the need to cover an area of about 40 km ALT (ALong Track) X 40 km ACT (ACross Track) centered on a point of interest (a city or a large CO2 production area). This global acquisition may be achieved in several elementary acquisitions. As far as possible, the position of the MCV is maintained for the duration of one elementary acquisition, in order to minimize the disturbance induced on the line of sight. Thus the satellite has to realize the compensation manoeuvers calculated by the guidance program, as long as they are compatible with the constraints, in particular kinematic and thermal.
The constraints to be respected during a basic acquisition include:
The minimal ground length of one elementary acquisition (40 km here)
The compensation of the Geometric Shifting
A spatial resolution constraint (Field Of View ALT length, 2 km here) that drives the slow motion In the City mode, the user must be able to modify the duration of the integration time. The figure 27 gives us an example of acquisition for a City mode over Paris:
38
The City guidance mode, on an elementary acquisition, is similar to an acquisition in CalVal Target mode, which is already implemented in POLARIS. Indeed, in this mode, MicroCarb is supposed to perform several acquisitions, on a straight line, on both sides of a TCCON as illustrated in figure 28:
Figure 28: CalVal Target representation. The City acquisition acts in a similar way [1]
39
Typically, the number of acquisitions has been determined to be three (the ACT FOV in Nadir pointing is close to 13 km wide).
4.2.
RESULTS WITH 3 ACQUSITIONS
4.2.1. FIRST TESTS/FOV DEFORMATION
I started working on the City mode by having a first look to the CalVal target code which was available at that moment. However several modifications and additions were applied in order to correspond to the needs mentioned above.
Typically, here is the calculation procedure for the City mode.
At a given starting date, the guidance has to calculate the date at which that mode can be executed, regarding visibility (with respect to the targeted city). Once executed, the date at which the satellite passes the closest to the city is computed: It is called “Middle Guidance Date”.
After determining the middle guidance date, a central acquisition over the city is performed. The city center is supposed to be the center of this acquisition. Then the starting date and ending date of this acquisition are computed. Indeed, the minimal ground length of one elementary acquisition and the spatial resolution constraint induce a fixed duration of one elementary acquisition.
The cape of the satellite, during this acquisition, is thus known. The intermediate points are then computed from the city center, orthogonal to this heading and separated by a distance of 13 km (the ACT FOV in Nadir pointing).
To calculate the first acquisition, which would take place before the central one, we proceed by steps. First, by using a function called BoundDateAttitudeLaw, we calculate the attitude the satellite would have at the end of an acquisition over the time, before the Middle Guidance Date. Each date has its own “end” attitude. Once done, we call MANIAC in order to perform a search for linking in a minimum time: For each date, a slew is intended in order to rally the corresponding attitude to the initial attitude of the central acquisition. MANIAC returns the closest date, from the Middle Guidance Date, which allow us to respect the kinematic constraints during the rallying slew. Once the closest date is known, we can determine the date the satellite will pass over the first intermediate point, knowing the acquisition duration, and compute the first elementary acquisition.
For the third and last acquisition, we proceed in a similar way, but instead we are using
BoundDateAttitudeLaw to determine the closest date the satellite may start the acquisition after the
central one. This acquisition goes over the other intermediate point.
40
Figure 29: Results of a 3 acquisitions-City mode displayed on Excel (Geometric Shifting not compensated)
When looking at this first figure, the difference between each acquisition, and their superposition, are problematic. The figure 30 is more relevant, as the projection is over a given city (Darwin, Australia)
Figure 30: Other City projection, displayed with Google Earth
41
In order to respect the City mode constraints, the acquisitions were supposed to occur close to a Nadir attitude, in a very short time interval centered on the Middle Guidance Date. In this attitude, the IFOV dimensions are approximately 580 m ALT X 13500 m ACT. Using the appropriate slow motion and appending the three acquisitions with no superposition at all, all the City mode constraints would have been respected.
Nevertheless, due to a lack of agility, the first and last acquisitions are not performed close to Nadir, but at a high-pitch angle of 60°. This high-pitch attitude results in a non-negligible deformation for the IFOV (see Appendix A.7): around 4800 m ALT X 39800 m ACT (largest dimensions). Actually only the second acquisition is close to Nadir, and is literally covered by the other two.
Among all the possible reasons, the ALT dimension of the IFOV itself, exceeding the spatial resolution of the City mode, makes us unable to proceed using 3 elementary acquisitions.
Later, more studies were intended, compensating for the Geometric Shifting and also trying to change the allocation of the four reactions wheels used in MicroCarb. However, as the reaction wheels could not be modified on the Myriad platform used for this satellite, a significant change of the allocation (and possibly a change in the reaction wheels themselves) cannot be considered.
4.3.
RESULTS WITH TWO ACQUSITIONS
4.3.1. RESULTS
The three acquisitions City mode is not a viable option so; we had to consider other alternatives. Due to the lack of agility, and the ACT deformation due to high pitch acquisition, the first idea was to decrease the number of acquisitions from three to two. By decreasing the pitch angle, we hoped to decrease also the ALT dimension of the IFOV, in order to fit the spatial resolution criteria. Also, the ACT deformation would allow us to cover 40 km just with two acquisitions.
This change of strategy implies a change in the calculation protocol:
The calculation of the Middle Guidance Date remains unchanged.
The two intermediate points are computed using a simulation of the central acquisition. The process is similar to the one used previously. However, this central simulation will not be part of the final guidance slew.
The first acquisition is computed so that its middle date (the date when the line of sight passes over the first intermediary point) is separated from the Middle Guidance Date by an initial interval, given in parameter as an input.
In the same way we computed the third acquisition earlier, we compute the acquisition passing over the other intermediary point, as well as the rallying slew thanks to MANIAC.
42
These changes in the code created some difficulties. Some kinematic issues led to computation errors, with the iterations not converging most of the time. In the best cases, when converging, the two acquisitions were not symmetrical, not solving the high-pitch problem. Eventually, the code worked. On the figure below (figure 31), we can see the projection of the 2 acquisitions City mode (using KML data files and Google Earth)
Figure 31: Results of a 2 acquisitions-City mode, displayed on Google Earth (Geometric Shifting compensated)
On the last day of the Thesis, as the lasts tests were compiled, the two acquisitions occurred at a pitch around 40°. Figure 32 illustrates one of the last projections I recorded. At this pitch, the IFOV dimensions are: 1100 m ALT X 18500 m ACT.
With respect to the City mode constraints, the ALT distance covered turned around 40 km as requested. The IFOV dimensions allowed us to meet the spatial resolution constraint, as the FOV length can now be brought to the requested 2 km.
43
4.3.2. DAZZLING RISKS
The new strategy, using two acquisitions, brought a lot of satisfactions concerning the performances during the elementary acquisitions, as well as the kinematic constraints. We could afford to switch this mode under the spectrum of the dazzling tests.
All the work done on the Assembly and the previous dazzling tests were available to be performed to other guidance modes, and the City mode is one of them. By using the guidance data in the same way I did for the Glint acquisition mode, I could determine if the Thermal Baffle was dazzled or not.
First, once the mode was working properly, I spotted the Sun fluxes were dazzling the Baffle at the beginning of the acquisition period. By changing the MCV roll target, I intended to cancel this dazzle. Unfortunately, as I finally canceled the Sun dazzle, the Earth fluxes entered the Baffle at the end of the acquisition period this time.
Seeing that my changes in the MCV roll targets were not enough, I tried to change the minimal elevation in which the satellite has to pass over the city to allow the City mode activation. This modification led not only to a closer approach of the trace above the city, but also to a modification of the Middle Guidance Date. What I was looking for was a limitation of the platform’s roll, so that a particular value of the MCV roll could leads to a cancelation of all dazzling events.
In my last simulations, this last attempt seemed to have good perspectives, as the dazzle was effectively canceled.
4.3.3. AREA OF IMPROVEMENTS
The implementation of this new acquisition mode is not yet complete. So, some improvements could be considered for future studies.
First of all, some “gross” values were used all over the program and should be removed. These values strain the user and do not allow him or her to proceed to more personalized acquisitions, or more suitable ones.
Another possible improvement would be implementing a small iteration to make the 2 acquisitions at the right distance, preventing gap between and superposition. As we can see in the figure representing the 2 acquisition projection (fig 30), the acquisitions are superposing a little. During other tests, it is possible to have a gap instead.
