Design of Multidimensional Space Motion Simulation System For Spacecraft Attitude and Orbit Guidance and Control Based on Radar RF Environment

Transcription

1 2016 Sixth International Conerence on Instrumentation & Measurement, Computer, Communication and Control Design o Multidimensional Space Motion Simulation System For Spacecrat Attitude and Orbit Guidance and Control Based on Radar RF Environment Lu Zhijun 1,3, Xia Hongwei 2, Ma Guangcheng 2, Wang Changhong 2, Yang Shengqing 3 1.Key Laboratory o Radar Imaging and Microwave Photonics (Nanjing Univ. Aeronaut. Astronaut.), Ministry o Education, Nanjing University o Aeronautics and Astronautics, Nanjing, , China luzhijun2004@sina.com 2. Department o control science and engineering, Harbin Institute o Technology, Harbin , China xiahongwei@hit.edu.cn 3.Shanghai key laboratory o space intelligent control technology, Shanghai , China. Abstract The hardware-in-the-loop (HIL) simulation system o spacecrat attitude and orbit control has been studied. In this paper, the composing o the HIL system and the main unctions o sub-systems are given. The method to determine the scaling ratio o a simulation system is introduced. A novel communication scheme is designed based on optical relective iber memory. The details to design a GNC/dynamic simulator including their working low is developed. Finally, the ground monitor sub-system is discussed. The design and develop o HIL simulation system provides a basic platorm or ground HIL simulation o spacecrat. Keywords simulation; spacecrat; scaling ratio; monitor system I. INTRODUCTION The ground simulation system o spacecrat plays an important role in improving the reliability and saving the costs in space industry [1,2]. Hardware-in-the-loop (HIL) reers to a kind o simulation in which the dynamic o orbit control and attitude control is built on the mathematical model, while the sensors and controllers are the real equipment [1-4]. Since the HIL has the superiority o high precision, it has been used as the main method o the ground simulation. The HIL system is used to be motion simulators, such as or attitude and or. By applying the electricity-driven simulators, the six degrees o reedom (6-DOF) motion can be simulated with high precision sensing and controlling [3-6]. Recently, a series ground HIL system simulation systems are developed by distributed structure [2] and multiple HIL systems are developed or speciic space tasks such as rendezvous and docking [4-5], lexible satellite [7], the navigation system o spacecrat [8]. However the HIL systems designed or speciic space task are short in extensibility. In this paper, the structure o the HIL system, the design o subsystems and the data interaction are studied in order to achieve This work was supported by the National Natural Science Foundation o China ( , ). a good scalability and reusability o the HIL simulation. Dierent dynamics simulators can be selected or dierent space tasks. The design o 12-DOF HIL system can be applied to multiple situations o navigation, guidance and control o space tasks. II. SYSTEM DESIGNING The HIL system is consisted by our sub-systems: GNC/dynamic simulator, motion simulator, sensing subsystem and monitor sub-system. The composing o HIL system is described as igure 1. Fig.1 Composing diagram o HIL simulation system The GNC simulator computes the navigation, guidance and control (GNC) which are applied into the dynamic simulator. The dynamic simulator computes the motion integration based on the control moments given by GNC simulator and drives the motion simulator. The motion simulator includes two sub-systems as translation motion and rotation motion sub-system. The translation motion sub-system simulates the 3-dimension relative motion o spacecrat. The rotation sub-systems use a three-axis turntable or simulation along pitch, yaw and roll directions. The motion simulator receives motion inormation rom dynamic motion simulation or monitor sub-system which decide the eature o motions. The translation and rotation /16 $ IEEE DOI /IMCCC

2 simulators have separate motion controllers which can track the inputs and meet the required accuracy. The sensing systems use visual-aids inertial navigation equipment or measurement and sent the inormation to GNC system orming a closed loop control system. Also, the data are sent to the monitor system which displays the measurement inormation on the monitor screen. The monitor sub-system can be used to display the motion inormation and operate other sub-systems by command. Also, the monitor sub-system supplies error handling unction which can protect the operational o HIL system. III. SCALING RATIO OF MOTION SIMULATOR The motion simulators can simulate the 3-DOF o translation motion and the rotation motion along the pitch, yaw and roll direction. System diagram is shown in Figure 2. The motion simulator runs on the bed-1. The translation motion is simulated on the beam-2, vertical beams-3 and guide rails. The 6-DOF rotation motion is simulated on a three-axis turntable-7, beside which lay the control system-5, 6. When HIL simulation system works, because o the limitation o the real scale o motion simulator, it is impossible to simulate the orbit motion in real scale. Thereore, a scaling ratio method o simulation system is necessary. The determination o scaling ratio is the core technology o the simulation system. Fig. 2 The physical map and entity o motion simulator Deine the parameters real motions o satellites x s, y s, z s and parameters o motion simulator x, y, z. The maximal scaling ratio o motion simulator K max should satisy xs Kmax x x ys Kmax y (1) y zs Kmax z z Kmin max( Kmax x, Kmax y, Kmax z) (2) In considering o the sensing system, l s and l are deined as the range o the sensor and simulator, ls and l are deined as error o the real sensor and simulated sensor. The scaling ratio o simulation should satisy Kl min lsmin lsmax lsmin K (3) Kl max lsmax l max l min Kl ls (4) According to the ormulations (1)-(4), the scaling ratio o the motion simulator is determined. IV. DESIGN OF DISTRIBUTED COMMUNICATION SYSTEM Ground simulation system is a comprehensive distributed system with various systems command and data to be interacted. Traditional ways o communication includes RS422/485 networks. However, due to the real-time requirements o simulation, such methods cannot meet the requirement o real-time communication o a large amount o data. Optical relective iber memory card as a production o GE Company, orms a relective memory network, in which the nodes on the network "share" a virtual memory. The underlying communication protocol can achieve a ast and eicient real-time communication o large amounts o data. Thus, simulation system using optical iber relective network built on the optical relective iber memory (RFM) cards can be used or data and commands exchanging between various sub-systems. The structure o communication scheme is shown in igure 3. Sub-systems constitute a iber optic relective network based on iber optic memory relective memory card. Data communication between the s is located in dierent areas o the relective memory cards. Typical storage iber optic relective memory card o type VMI5565 is 128M. It is suicient or usage. Motion simulator is also controlled by the ground monitor system. When the system is not in simulation, motion simulator is manually controlled by ground monitor system, receiving operation commands including system reset, system homing, moving to the speciied location. When the motion simulator is in simulation, the motion model and operating command are rom dynamics simulation system. 29

3 Fig.3 The coniguration diagram o communication scheme Sensing system only works during the simulation period. When the HIL system is in simulation, the sensing system measure the position and attitude inormation is written to a speciic location in RFM which is read by the dynamics simulation sub-system and ground monitoring sub-system. The dynamic simulation systems only work during system simulation period. When the dynamic simulation system works, it writes control data to a particular location RFM, including running position, attitude, speed inormation, the mode o operation and the start and stop inormation. On the other hand, the dynamic simulation system rom RFM another position reads the position and attitude inormation measured by sensing system, and eedback the inormation to GNC simulator orming a closed-loop system or navigations. The ground monitor sub-system gives a start-stop control and communicates with other sub-systems when the system is not in simulation situation. The ground monitor system sends a variety o operating commands and data and it read data rom the motion simulator and sensing sub-system. The ground monitor sub-system displays the inormation in on a screen and save it or later analysis. Furthermore, at all stages o system operation (including system test and operational phase), each sub-system needs ground monitor system to reports sub-systems running status. Once a sub-system is ound abnormal, it should stop the running system and display the corresponding message. In a word, a large amount o data is required in communication between the sub-systems. Thereor the division o RFM network is necessary, in which dierent segments represent the dierent communication content. V. DESIGN OF GNC/DYNAMIC SIMULATOR The GNC/dynamic simulator is the kernel o HIL system. It computes the control outputs based on the sensors and determines the position and attitude o motion. The close loop o GNC/dynamic simulator is shown in the igure 4. As shown in the igure 4, the period o simulation is 10ms. It s shorter that the control period o GNC system. The dynamic simulator system simulates the real-time motion o spacecrat, including the environments, the sensors and executors o spacecrat. The dynamic simulator should satisy: 1) excellence real-time perormance, 2) extendable s. Fig.4 The close-loop o GNC/dynamic simulator and Control Cycle Schematic Diagram According to the requirements, the dynamic simulator is built on a high-power PC, o which the external interace is the optical relective iber memory. The dynamic simulator includes the ollowing. Table 1 Composing o the dynamic simulator communication and complete the communication control and control between s actuator simulation simulation actuator outputs, such as jet control agencies dynamic simulator complete the simulation o space environment simulation motion o spacecrat complete the simulation o space environment, such as disturbance torque A. Communication and control The communication and control response to the coordination and interaction o sub-systems. The working low o this is shown in igure 5. Fig.5 The working low o communication and control In order to ensure a real-time communication, communication and control inquire the command receiving area every 10ms. I the inquired data is updated (a lag o receiving area), then updates the data o actuator simulation and space environment simulation. The dynamic simulation will complete the update 30

4 every 10ms. The sensor then outputs its data to communication receiving area in where the GNC system can receive real-time navigation data o satellite. B. Actuator simulation Usually, the spacecrat is controlled by the jets, which can be seen as a constant thrust switch control. The inputoutput characteristic is approximately determined as a delay as shown in igure 6. a and initial conditions x 0, y 0, z 0, vx 0, vy 0, vz 0. The outputs o dynamic simulation include position and velocity x, yzvxvyvz,,,, o next time step. According to the igure 4, the calculation period o dynamic calculation is 10ms, while the period o close-loop control is 100ms. Thereore, period o the dynamics simulation is 10ms. The working low o dynamic simulation is shown in igure 7. Fig.6 The input-output characteristics o jet engine Acceleration caused by the jet engine is determined by the ollowing ormula: Fi m I g (5) Fi ai m where m mass o the spacecrat I speciic impulse o the jet g gravitational constant Fi thrust caused by the i-th jet When simulating the actuators, the quality o spacecrat is known as a constant. The input o GNC control algorithm is the control torque, the output is the acceleration which is used as inputs o dynamics simulation system. When the actuator receives a new control torque rom relective memory card, it can update the "Data Update" lag o dynamics simulation sub-system. The dynamic simulation will inquire the lag every 10ms, and update the initial condition o dynamic calculations. C. Dynamic simulator Dynamic simulation is generated by the Matlab/Simulink and translated to C code under Matlab RTW. Through modiications o the C code, it can meet the requirements o the system input and output, which can be compiled into a dynamic link library. Labview can call the dynamic link library, which embeds the dynamic calculations into dynamic simulation. It only needs to set up the dynamic model in Matlab/Simulink, the dynamic model can be accessed into dynamics simulation. This enhances the portability and reduces the development time. In relative orbit motion developed on the C-W equations, the inputs o dynamic simulation include accelerations Fig. 7 The working low o motion simulation D. Space environment simulation The situation o disturbance torque in the space is complex, which can be accurately simulated by adding generally sinusoidal perturbation. The working low o space environment simulation is similar to the situation o agencies analog s, in which the disturbance torque is set according to the theory. When the operating system sends command o disturbance or simulation, the space environment simulation can send the corresponding disturbance torque to dynamics simulation. Suppose the disturbance torque is in the orm o a sinusoidal signal, its output period should be the same as the control loop cycle, which is 100ms. In order to guarantee the authenticity o the space environment simulation, nonspherical perturbation, solar pressure and other space disturbance models can be considered in the space environment simulation. The output o disturbance torque is in the orm o acceleration. In order to improve development eiciency and shorten the development cycle, the disturbance torque models are constructed in Matlab/Simulink. It can be automatically translated into C code under RTW with a dynamic link library (DLL) which can be direct called by Labview. VI. GROUND MONITOR SUB-SYSTEM Ground monitor sub-system includes computers and corresponding control sotware, which are used to complete the display and monitor/control sub-systems. The ground monitor sub-system may have the ollowing eatures(1) control the sub-systems(2) monitor the sub-systems, react to simple ailures(3) display the real-time data(4) store the data, replay the data according the requirements. 31

5 The ground monitor sub-system is developed in Labview environment, including the ollowing s Command creation and control, data receiving and managing, data display and interaction, data storing and replay. A. Command creation and control The command creation and control is designed to create the operation commands, which are sent by communication network to realize the control o sub-systems. This should create commands according to multiple inputs and set alarm when ailures exist. B. Data receiving and managing The data receiving and managing reads the data rom the optical relective iber memory card and translates the position and attitude inormation. The data receiving and managing period is 10ms. C. Data display and interaction The data display and interaction is developed in Labview environment, which supplies display o virtual equipment and graphs/curves o data. The display inormation includes: position, velocity, working status and control inormation o sub-systems. D. Data storing and replay In order to improve the eiciency, the data should be written to cache memory irst and written to disk while the amount o data is enough. This can supply the curve o data according to dierent scaling ratio. The interace o ground monitor sub-system is shown in igure 8. Fig.8 Interace o integrated ground monitor sub-system VII. CONCLUSION In this paper, the problem o multidimensional space motion simulation system design or spacecrat attitude and orbit guidance and control is investigated. The composing o the HIL system, the system scaling ratio, the distributed communication schemes, the dynamic simulator and monitor sub-system are introduced in detail. A computational method o the system scaling ratio is investigated based on a speciic space task. Memory division is considered or optical iber relective network which is consisting o optical relective iber memory card. Solving method and working low o each in the dynamic simulation are designed. Finally, the requirements o ground monitor sub-system unctional are discussed. The HIL simulation system designed in this paper can provide a platorm or the urther simulation experiments. REFERENCES [1] S. Xiong, R. Zhang, T. Bai. Research on satellite attitude & orbit control semi-physical simulation system architecture, Journal o system simulation. vol.26, no.7, pp , [2] S. Li, T. Bai, S. Xiong. Satellite light Decoupling Method simulation based on new three-axis turntable, Electronic design engineering, vol.3, pp.72-75, [3] K. Xu, Y. Lv, G. Jin. Design o Semi-Physical Simulation or Small Satellite by Virtual Display, Applied Mechanics & Materials, pp , 2011 [4] Y. Yu, Z. Yang. A semi-physical simulation platorm o attitude determination and control system or satellite, Advances in Mechanical Engineering, vol.8, no.5, pp.1-5, [5] L. Christian, M. Eric. Towards Docking Emulation using Hardware-inthe-Loop Simulation with Parallel Platorms, Proceedings o the workshop on undamental issues and uture research directions or parallel mechanisms and manipulators, vol.10, pp.1-4, 2002 [6] Z. Chu, S. Ren. Semi-Physical Simulation on Vibration Control o Small Satellite with Flexible Manipulators, Journal o Astronautics, vol.34, no.6, pp , [7] Z. Wang, H. Yang, X. Feng. A design method o high-idelity simulation system based on ault diagnosis on small satellite platorm, IEEE Control Conerence, pp , [8] H. Hu, Q. Tang, C. Wei, et al. The experiment with dierential GNSS relative navigation o ractionated modular spacecrat, Chinese Control Conerence, pp ,