GNCDE: AN INTEGRATED GNC DEVELOPMENT ENVIRONMENT FOR ATTITUDE AND ORBIT CONTROL SYSTEMS

Transcription

1 GNCDE: AN INTEGRATED GNC DEVELOPMENT ENVIRONMENT FOR ATTITUDE AND ORBIT CONTROL SYSTEMS Fernando Gandía (1), Luigi Strippoli (1), Valentín Barrena (1) (1) GMV, Isaac Newton 11, P.T.M. Tres Cantos, E Madrid Spain, ABSTRACT The design of Guidance Navigation and Control (GNC) systems for several kinds of Spacecraft s is a cooperative and iterative process in its own nature. GNC engineers and System engineers need to cooperate and support themselves in their respective responsibility areas. Software tools must also contribute to it by providing the required functionality but also by implementing the inner mechanisms that allow data coherency and easy integration among the different tools. GNCDE (GNC Development Environment) was conceived to fulfil the need for this kind of computer engineering tools. It is able to support the user at every step of the GNC design process: from analysis/consolidation of system requirements to control plant modelling and synthesis of control laws to finally reach the detailed performance evaluation phase and the generation of on-board code. The present paper shows in more detail how GNCDE is able to cover all these steps and the implementation approach behind each functionality. 1. INTEGRATED ENVIRONMENTS FOR THE GNC DESIGN PROCESS The GNC design process usually comprises a set of different disciplines and requires involvement of a team composed of people with different background knowledge and formation. The set of knowledge areas usually involved to a certain extent in the GNC design process includes: - Mission design and planning - Spacecraft systems knowledge - Trajectory design - Control design - Sensor technology - Navigation strategy and Navigation filters design - Onboard SW coding - SW verification - System verification (including HW in the loop) The GNC design loop is, in general, an iterative process, where the designer moves forward and backward through the several steps of the sequence, in order to refine the functions being designed. Indeed, initial requirements and assumptions can be usually reconsidered in view of the results obtained from either preliminary analyses or detailed performances evaluation. Due to this iterative nature of the work, the handling of the design data, including inputs and outputs to the process (requirements, synthesis models, parameterisation of the models, mathematical representation of the navigation and control functions, etc...) needs to be managed in a coherent way and made available in a suitable shape to every support tool being used in the process. All these facts point to the need for integrated GNC development environments that not only provide the tools able to support the analysis, synthesis and evaluation activities required but that also manage in an integrated way the data being used within the full process. GNCDE is such a kind of integrated GNC development environment, able to provide these tools and the data coherence mechanisms among them. 2. GNCDE DESIGN GUIDELINES Some high-level considerations were taken into account for the selection of an adequate GNCDE implementation approach. The first one was related to the need for easy operation at both novel level and expert level. So, it was decided that the tool had to offer predefined cases, classified inside reference scenarios, which did not need deep knowledge of the tool. Moreover, the predefined cases had to be able to be extensively customised and modified if needed (expert level). In addition, GNCDE needed to be flexible enough to allow designing/analysing very different cases, belonging to different types of mission, while maintaining powerful design and analysis capabilities (additionally these features needed to maintain the same operational procedure for all the possible scenarios). From the point of view of the user-tool interaction, the tool needed to be highly intuitive, providing graphical resources (dialog boxes, graphical manipulation of models, interactive modification of figures) while keeping in parallel the powerful use of ASCII files for

2 all parameters susceptible to be modified and reported for an expert use. In this way, all tasks performed through the GNCDE could be easily automated and customised by means of the use of scripts developed by an expert user. Another important need identified was that GNCDE had to be easily maintainable and expandable, allowing continuous update of models and utilities. It assures that the tool maintains usable and useful along time. Considering all these requirements, the first decision adopted was to build GNCDE on top of the Matlab/Simulink environment. This environment combines adequate characteristics to cope with the previous requirements, assuring high flexibility and significant availability of already developed functionality (commercial toolboxes, Simulink basic functionality blocks, in-house Simulink Space libraries). Other implementation options are explained in the following paragraphs, relating them to each of the GNC design processes. Template Management Project Management Tools area Tool Settings and Exit button Library access icon Plot area Templates access icons Logos area Message Box Figure 1. View of the main GNCDE Graphical User Interface window. 3. GNCDE APROACH TO THE DESIGN AND ANALYSIS PROBLEM As already highlighted, GNCDE supports the user at the different GNC design cycle steps. These steps are mainly: - Requirements analysis/consolidation and equipment selection and sizing - System modelling - Guidance profiles design and analysis - Control laws synthesis and analysis - Estimation filters synthesis and analysis - Detailed performance evaluation - GNC coding and verification At each of these steps, the user is allowed to iterate and decide whether the design performed is good enough or need to be refined or initiated from a new consolidated set of requirements or assumptions. Each of these steps and the way GNCDE supports them are described in the following paragraphs. 4. REQUIREMENTS ANALYSIS, CONSOLIDATION AND EQUIPMENT SIZING The first phases of the GNC design process involve the analysis and completion of the requirements and a set of preliminary analyses devoted to: - Supporting the sensors and actuators trade-off and selection. - Defining a preliminary mission profile, that, at the same time, allows having a first estimation of consumed propellant and timeline for completion of the mission. - Assessing the feasibility of the requirements imposed, given the preliminary selection of Spacecraft equipment and the first design of the mission profile. In view of its results, it could lead to the review of the equipment trade-off, the preliminary mission profile or indeed the revision/completion of the requirements imposed in the case that they cannot be fulfilled or they are not properly specified under the current technology limitations. GNCDE provides a set of tools (generally based on analytical formulations) which support these tasks. They allow quick setting of the problem to be analysed and fast evaluation of the solution. In this way the user has the possibility of performing complete trade-offs or checking preliminary mission assumptions without the need for spending a large amount of time and effort. These tools are: - Guidance Analysis Tool (GATO) - Covariance Analysis Tool Each tool can be used in a variety of ways that are explained better later. In addition to these analytical formulation-based tools, GNCDE also provides a numerical computation tool for assessing the effects of the perturbation environment on the Spacecraft. This tool allows using a geometrical definition of the vehicle for computing and obtaining a very reliable estimation of the expected perturbation environment. This estimation is required for sizing actuators in some cases or as input for consolidation of the GNC requirements and mission profiles in other ones. The Covariance Analysis Tool, already mentioned, is

3 aimed to perform quick analytical covariance propagation analyses. This tool is able to work with LTI (Linear Time Invariant) models or LTV (Linear Time Variant) user provided models that can represent the dynamics of the Spacecraft. The number of interesting applications of this tool is large. For instance, it can be used as a support tool for a preliminary analysis of requirements since it allows quickly propagating an initial state and its associated initial covariance, given the dynamics of our vehicle. In the case illustrated by Fig.2, the tool helps to determine if the requirement on the final hopping position accuracy for a Rendezvous mission can be fulfilled given a set of initial conditions or if we would need some additional action (e.g. such as scheduling an intermediate correction manoeuvre or improving the accuracy of the thrust by selecting a different thrusters model) in order to fulfil it. Tool and allows improving the efficiency of the CAD model definition process for complex vehicle shapes. In addition to the CAD model specification, a visualization feature that allows a user easily assessing that the model being generated fits to the expected shape of the vehicle, is available. After the importation or definition of the vehicle model, the model file is passed to routines for meshing and perturbation computation. The perturbations computation is performed on basic surface elements, which need to be computed from the user defined CAD model geometry (created by the meshing routine). See Fig 3. After the meshing process, the perturbation computation loop can run on each of the elemental surface elements to provide a final estimation of the overall Spacecraft perturbation environment at each time step during the mission profile. The perturbations computed on this basis are: - Air drag forces and torques. - Solar radiation pressure forces and torques. Figure 2. Relative position covariance propagation represented through dispersion ellipses in a hopping arc. There are also a second type of perturbations whose values are computed based on the overall physical properties defined for the spacecraft (magnetic disturbance, based on the equivalent dipole) or on the physical properties defined for each of the primitives from which is composed the vehicle model (gravity gradient torques). By assessing the perturbation level in the selected orbit it is possible to have a first estimation of the control actions required and thus preliminary sizing the actuators required. The problem of actuators sizing, considering the control requirements imposed and the expected perturbation environment in the selected orbit, is also supported in GNCDE through a numerical tool. This tool allows the computation of perturbation effects on a given Spacecraft by defining vehicle geometry through a CAD model. The GNCDE CAD Tool provides both: - CAD Model specification/importation and visualization. - Perturbation computation on the CAD model and output generation. The CAD Tool contains a CAD model specification module. The routines of this module are able to generate models from a set of basic primitives (spheres, cylinders, prisms etc...) Additionally, a function for importation of.dxf files (limited to basic.dxf features) is available within the CAD Tool. It makes easier the workflow within the Figure 3. Spacecraft CAD model after the meshing process. 5. SYSTEM MODELLING The GNC design process requires from a set of models that represent the Spacecraft dynamics in its real environment. The fidelity and shape of the models depend on its use purpose, which can be:

4 - Control and state estimator synthesis and analysis. - Detailed performance evaluation of each GNC mode. The first of the two above mentioned activities requires models that directly fit to the synthesis and analysis technique selected, while the second task requires from higher fidelity models able to represent with high accuracy the real vehicle properties and the real dynamic effects to which it is subject in orbit. Linear model representations are widely used in control synthesis and analysis while, in general, non-linear models allow describing with a higher fidelity level Spacecraft dynamics and environment effects on them. The solution adopted by GNCDE to cope with both kinds of requirements is relying on two different model environments that, however, are coherent in terms of Spacecraft and scenario parameterisation data in order to make easier and smoother the designer s work. These model environments are called Templates in GNCDE terminology: - Analysis and Design Template (composed of linear models and devoted to synthesis and analysis activities). - Mission Template (composed of high fidelity nonlinear models and devoted to detailed performance assessment). Both, Analysis and Design Template and Mission Template, implemented through Simulink blocks and a set of associated ASCII files (for parameterisation and initialisation purposes), contain representations of the Spacecraft dynamics and equipment (sensors, actuators) and on-board software (basically GNC and mission management functions). These model environments can be created from single or individual models that represent each of the elements (vehicle dynamics, orbital environment, sensors actuators, GNC etc...) and that are joined to shape the full mission representation (Mission Template) or an specific GNC mode representation (Analysis and Design Template). The single models that, placed together, conform each of these environments, can be reused from the GNCDE libraries (based on the ESA Simulink libraries standard SPACELAB) when available. If no available, they can also be implemented or incorporated by the user, so that the same reuse approach can later be applied for any further activity. Additionally, GNCDE supports this environment modelling activity by providing a framework that allows automating many activities required to prepare and later use these model environments or Templates. Among them: - Automates the process of initialising and running a Template. - Automates the process of generating a unified user input data file that contains the parameters required for each individual model present in the Template. - Automates the process of collecting output signal data from the Template and specifying which signals from those available, need to be collected. - Automates the process of generating usable outputs (plots and output data files) from output signals of the Template. - Automates the process of updating the associated files when a modification in terms of models is implemented in the Template. - Automates the process of extracting complete plant linear models by appropriately combining the elemental linear models included inside an Analysis and Design Template. The linear model representation used by default in GNCDE is the State-Space form, which is widely extended and convenient for most synthesis and analysis techniques. The linear model collected from an Analysis and Design Template (as a Matlab object) can be later used by the GNCDE tools (Covariance Analysis Tool, Control and Estimator Design Tool). 6. GUIDANCE PROFILES DESIGN AND ANALYSIS Guidance design is normally the first (non preparatory) activity in the GNC development process. The task of designing the Mission guidance profiles is also an iterative activity, which needs to account for V budgets while fulfilling a set of different requirements related to sensor operation ranges, safety constraints etc... It is clear, that the use of a tool able to support both the profiles generation and the related analysis tasks (allowing the checking of requirements and constraints fulfilment) is a very valuable resource. GNCDE provides a very complete tool for that purpose, which is able to support the guidance design problem in the elliptic orbit scenario (also in other scenarios). The GNCDE Guidance Analysis TOol (GATO) is able to perform preliminary analysis of guidance features (in terms of required impulsive V, continuous thrust history and reference trajectory) through a simplified dynamics model. This model makes available analytical solutions for the guidance features and the reference trajectories computation allowing rapid analysis of a broad range of different missions. GATO provides user interfacing through formatted ASCII files reporting the necessary input information,

5 as well as the results corresponding to the guidance analysis of the proposed scenario, and a Graphical User Interface that allows the user to manage projects, define input files and select output data to plot. The tool addresses the relative motion in elliptical orbits; it means computing the state of a chaser spacecraft that manoeuvres with respect to a target spacecraft in non-actuated orbital motion, both with respect to the same central body. It manages the relative motion in two reference frames: the Local Vertical Local Horizontal (LVLH) and the Local Orbital (LO), which coincide in circular orbits. The dynamic model implemented for computations corresponding to the elliptical orbit scenario is based on the Yamanaka-Ankersen transition matrix (see [1]). In the case of forced manoeuvres, the evolution of the system can be described through the homogeneous solution for the equations of motion plus a particular solution for a constant thrust. A wide range of predefined manoeuvres are implemented within GATO. The manoeuvre parameters are computed in function of user defined requirements and finally the manoeuvre profile is generated including also the analysis of V and thrust history. The single manoeuvres can be connected among them (final/initial state coherence among consecutive manoeuvres is automatically implemented in the tool) to shape a full mission profile. In addition to the Rendezvous scenario for circular an elliptic orbits, GATO also implements solutions for a Formation Flying scenario with a configurable number of Spacecrafts flying around L2 or in Earth orbit. It also includes attitude manoeuvre computation for 3-axis stabilised vehicles as well as a visualizer for launcher ascent trajectories. 7. CONTROL LAWS SYNTHESIS AND ANALYSIS GNCDE supports the synthesis and analysis of controllers via two major resources: - Analysis and Design Template. As already introduced, an Analysis and Design Template is itself a model environment that can be used for synthesis and analysis purposes. It is composed of linear models for Dynamics Kinematics and Environment (DKE), sensors and actuators, which represent our plant model for control synthesis and analysis. It allows, also, closing the loop with the synthesized GNC functions so that fast analyses can be run in this linear-model environment before going to detailed performance assessment. - Automatic Control and Estimator Design Tool (ACEDTool). It is a tool devoted to support a user in the process of synthesizing and analysing a compensator and/or state observer for a control problem. It is able to work with the plant model representations obtained from the Analysis and Design Template or with plant models imported from the Matlab workspace, Matlab binary files or ASCII data files in a given format. ACDETool provides both Single Input Single Output (SISO) and Multiple Input Multiple Output (MIMO) control synthesis and analysis capabilities, including: - SISO time-domain or frequency-domain analysis and synthesis features (lag, lead and lag-lead compensation nets, manual root locus manipulation etc...). - MIMO control analysis and synthesis techniques, such as Pole Placement, LQG and robust control techniques such as H and µ-synthesis. The methodology usually followed when synthesizing and analysing controllers with GNCDE involves the above mentioned resources and comprises the following steps: 1.1. Getting a synthesis model that represents the plant dynamics. As mentioned before, these specific mission dynamics (e.g. relative position dynamics and absolute or relative attitude dynamics) are already described by the models implemented inside the Analysis and Design Template. These models are parameterisable by the user and GNCDE fully supports the customisation process so that the Analysis and Design Template represents each of the GNC modes to be designed (it is done through different configuration files that are loaded when required). The ACEDTool is able to import this plant model (including spacecraft orbit and attitude dynamics as well as actuator and sensor dynamics), representing it in state-space form. It is done directly from the Analysis and Design Template, in a process where all the information present in the Template (such as inputs, outputs and state names) is preserved in order to make easier further manipulations of the plant model. Alternatively, plant models can also be imported from the Matlab workspace, Matlab binary files or ASCII files allowing also working with plant representations obtained from other sources Analysing the plant dynamics. Once the linear formulation representing the plant to be controlled is available inside ACEDTool, a first analysis of the dynamics can be performed. It allows deciding whether for the GNC mode currently being designed,

6 certain states can be considered as decoupled or whether certain perturbation inputs or secondary dynamic effects are not going to be considered in the design. All these decisions have an implication on the sub-plant selection and plant model preparation for synthesis. ACEDTool supports stability analyses as well as time-domain and frequency-domain analyses of several responses of the different transfer functions in the control loop (open loop, sensitivity transfer function, complementary sensitivity transfer function etc...) Selecting a synthesis method. Taking into account the nature of the dynamics to be controlled as well as the requirements imposed to the design, the user selects a control technique. Currently, GNCDE offers SISO and MIMO, Pole placement, LQG H and µ-synthesis methods but is flexible enough to allow the use of other external synthesis tools while taking advantage of the analysis, visualisation and manipulation features implemented within ACEDTool Preparing the plant model for the synthesis. Previous to the control synthesis computation, several preparation tasks can be required: - Subplant selection. Many times, we will not use the full plant description for the controller synthesis. For example, in many cases, position and attitude compensators are synthesised independently or we do not want to cope with secondary effects (e.g. sloshing or flexible effects) in an explicit way for the synthesis. ACEDTool supports the selection of a sub-plant from the overall plant model, which contains only the states, inputs and outputs required. - Discretisation of the plant model (if required). - Simplification (if required). Sometimes, the original plant model already contains dynamics whose effects can be neglected when comparing to the main ones. ACEDTool supports the order reduction or simplification of plant models. In many cases it helps to reduce the complexity of the synthesised compensators and to simplify the design process Synthesis requirement specification. Depending on the synthesis method, a set of design requirements can be imposed to the synthesized controller (e.g. signal transient response and steady state related requirements for SISO time-domain design, stability margins for SISO frequency-domain design, closed-loop poles position for MIMO Pole-placement or shape of the weighting functions (see Fig 4.) included in the transfer function to be minimised when coping with robust synthesis techniques such as H or µ-synthesis). Figure 4. Interactive synthesis framework for definition of weights and scaling factors in robust control methods Synthesis of the compensator. Once the design requirements have been set and the synthesis model is prepared, ACEDTool will perform the compensator computation for us. The compensator is provided in state-space form Stability and performance analysis of the closed loop system. ACEDTool places the synthesized compensator in the control loop and allows analysing the dynamics of the compensated plant in several ways. As a first step, it allows selecting any of the possible transfer functions in the closed loop system (from any input or perturbation input signal to any output signal), including frequently used transfer functions such as the sensitivity transfer function and the complementary sensitivity transfer function. After that, the user can select from a set of analysis graphs (Bode, Nyquist, Nichols, Pole-zero maps, Step response) of the selected input-output channel or ask for specific analyses of the system such as: - Analysis of feedback loop gain and phase margins (including worst case analysis). - Analysis of feedback loop poles. - Robust performance margin. - Robust stability margins. - Bounds on worst-case gain of the uncertain system Exporting the compensator for analysis outside ACEDTool. ACEDTool includes an export functionality that allows having the synthesized compensator representation available in the Matlab workspace as a state-space object or in a Matlab binary file (.mat) or ASCII file. In the same line, Analysis and Design Templates are also able to use these controller files, so that the process of synthesizing a compensator and having it available for analysis in a GNCDE Template is smooth and direct. It is important to note here, that in the same way that a compensator representation can be exported from

7 ACEDTool, it can also import compensator objects coming from other sources (e.g. alternative controller design packages) while these compensators are available as Matlab state-space objects. In this way, it is possible to take advantage of the GNCDE analysis capabilities while using other tools for synthesis. Once the synthesized controller is considered to be satisfactory it can be evaluated within a higher fidelity model environment ( known as Mission Template). It is, in fact, a Functional Engineering Simulator, that allows full Mission simulation. Controllers can be imported to GNCDE Mission Templates, in the same way as is done for Analysis and Design Templates (the same statespace object contained in a file can be used). 8. ESTIMATION FILTERS SYNTHESIS AND ANALYSIS Most of the times, the full internal state of our control system cannot be directly measured. Nevertheless, many control techniques are based on state feedback and thus it has to be estimated from available sensor measurements and knowledge of the input to the system (generally forces and torques in our Spacecraft control system). In state feedback control, the compensator and the observer can be independently designed since the closed-loop poles from the observer and the closed-loop poles coming from the compensator are independent of each other. In fact, the poles of the observer are usually selected so that the observer response is considerably faster than the system response (the only limitation to observer speed normally arises from noise sensitivity problems). GNCDE also supports the synthesis of state observers by two methods: - MIMO pole placement - Kalman filter design As well, as in the control synthesis problem, the support tools provided can use plant models imported from the linear model environment (Analysis and Design Template) that represents our problem. Via MIMO pole placement technique, GNCDE can compute a state observer whose dynamics are given by the user specification of the closed-loop poles position. In the same way, the Kalman filter synthesis technique is able to compute an optimal state observer given a state-space model of the plant and the process and measurement noise covariance data. 9. DETAILED PERFORMANCE EVALUATION Once, the first stability and performance analyses have been performed and the compensator fulfils the imposed requirements (in a linear-model environment such as is a GNCDE Analysis and Design Template), it is needed to test and analyse the behaviour of the control function in a higher fidelity model environment, where other effects, not taken into account till that moment, are considered. The same happens for state observers. A high-fidelity Functional Engineering Simulator is normally used for that purpose. This kind of resource allows accounting for detailed orbital dynamics (considering all applicable orbital perturbations), detailed Spacecraft dynamic effects (e.g. fuel sloshing, flexibility), detailed actuators configuration and modelling (including several sources of thrust errors), detailed sensors models (including also several sources of measurement degradation or error sources) as well as the possibility for reproducing transitions between the different GNC modes and the high level vehicle management functions acting. GNCDE also supports the preparation, data management, execution and results exploitation of Functional Engineering simulators that are known in GNCDE terminology as Mission Templates. As already highlighted, one of the main advantages of GNCDE is that all resources available can work in a coordinated way. It means, for example, that Control and Navigation functions being developed with the Automatic Control and Estimator Design Tool or imported from other external tool can be directly placed inside an Analysis & Design Template or a Mission Template and evaluated there. Additionally, the GNCDE infrastructure automates the parameterisation and initialisation process for the models contained inside each Template, assuring coherence of the data being used (e.g. among a Mission Template and an Analysis and Design Template corresponding to the same Mission case). Summarising, the GNCDE framework is able to make easier the work with Templates or Functional Engineering Simulators by providing these features: - Ability to save the current working environment in project files that can be later recovered. These files can be loaded later or indeed be executed in command mode (without the need for a Graphical User Interface). - Automation of the output selection mechanism. It allows a user specifying the set of required output data from the simulator, avoiding collection of unnecessary data while making simulation runs faster. - Ability to modify Templates by replacing models or adding new ones (automatic regeneration of input data files, output selection mechanism etc...). - Simplifies the results exploitation process. It allows creating plots easily (2D and 3D) with any output variable and exporting them to Matlab binary or ASCII column data files with selected subsets of

8 variables. - Support for statistical analyses of the output data collected. It is done in GNCDE through the Statistical Analysis Tool. This tool is able to use data from output files from Templates but also is able to use data imported from external files or from the Matlab workspace. It is very useful for the detailed assessment of performances and for the sensitivity evaluation of performance criteria with respect to certain GNC or mission parameters. - Support for sensitivity analyses and worst case analyses. GNCDE provides a MonteCarlo Tool that is able to run both parametric and statistical MonteCarlo simulations with GNCDE Templates. This tool supports the selection of the deflected input variables and the associated parametric variation or statistical distribution for these variables. - Increased situational awareness for the designer. A 3D Visualisation Tool (see Fig. 5) is provided within GNCDE. It can be fed with orbit data, attitude data or any other dynamic data variable coming from any Tool of the GNCDE Tool suite or indeed from other external tools. - Input/Output variable selection and compilation options feature. - Model Template import feature: it is in charge of importing the Template model specified by the user, as well as, the configuration options applicable to the model. - Compilation feature: it manages the data and uses appropriately the different external tools (Real- Time Workshop, Embedded Coder, and external compilers). - Real-Time Workshop and Embedded Coder. They are properly components of the Matlab family devoted to the coding of Simulink blocks. They are driven from the model Template import utility in order to generate the C code and applying the user production options. - External compilers. Once the Real-Time Workshop or Embedded coder have generated the C code corresponding to the Simulink blocks, it must be compiled and linked for obtaining a stand-alone executable on the target platform. Once, the executable code is generated, it must be verified and tested with specific tools available for debugging and analysing SW on the target platform (e.g. LEON2 or LEON3 processor board). 11. CONCLUSION Integrated environments are able to support a wide range of users with different background and roles in the several steps of the GNC design process. To achieve it with success, it is required that these tools provide not only the functionality but the data coherence and interaction mechanisms among the provided features. Figure 5. 3D dynamic view of the ATV docking to ISS port. 10. GNC CODING AND EVALUATION If a GNC design has been positively evaluated in a Functional Engineering Simulator, it can be considered validated only from a functional point of view. The next development step is intended to face real hardware architecture implementation constraints and requirements. The logical approach is taking advantage of the already existing implementation and migrating it to the final coding language (e.g. C language) and target hardware platform, where it needs to be validated. For that purpose, GNCDE implements an Autocoding tool that is built on top of the Matlab tools Real Time Workshop and Embedded Coder. Some additional automation capabilities were added to obtain a userfriendly tool able to exploit the features of GNCDE Templates. These elements are: GNCDE supports each of the GNC process steps in a flexible way so that a number of different space scenarios can be covered. Furthermore, different users working with such a tool usually find different ways of taking advantage of every tool provided when applied to an specific design or analysis activity. The selection of a commercial environment such as Matlab/Simulink, characterised by the high availability of commercial toolboxes and pluggable options as well as its high degree of flexibility demonstrated to be also very effective. 12. REFERENCES 1. Yamanaka K., Ankersen F., New State Transition Matrix for Relative Motion on an Arbitrary Elliptical Orbit, Journal of Guidance Control and Dynamics, Vol 25, No. 1