| Laser communication between satellites has many advantages such as large communication capacity,strong anti-interference,light weight and good confidentiality.Therefore,it has been widely studied and applied.The establishment and maintenance of satellite laser communication link is determined by Acquisition,Tracking,Pointing(ATP)system.The vibration of satellite platform is one of the main reasons affecting the tracking and aiming accuracy of ATP.In order to suppress the error caused by platform broadband jitter,a precise tracking system with high-speed processing performance is needed.The main processors of traditional precise tracking control system are DSP,PC104,Power PC and other serial processors,which are easy to develop and implement floating point control algorithm.However,this kind of processor is limited in controlling the real-time processing ability of algorithm.FPGA has the characteristics of high frame rate and parallel processing.However,it is found that the traditional FPGA development has some problems,such as cumbersome design process,difficult debugging,low precision of fixed-point operation and difficult to realize floating point operation.On the other hand,various vibrations in space increase the nonlinearity of the control system,and space is not convenient for manual maintenance.A traditional PID(Proportion Integration Differentiation)controller with fixed parameters cannot achieve good control effect.Aiming at the above problems,this paper mainly studies how to implement the parallel floating point control algorithm on FPGA.Compound traditional PI algorithm and fuzzy control algorithm,and strive to achieve high speed and high precision processing effect and high precision control.Floating point fuzzy PI algorithm is built by System Generator(SG).Finally,it is applied to FSM control system to complete the hardware implementation of FPGA.Specifically involved:1.In view of the problems of traditional hardware programming and design methods,such as cumbersome design process,long development cycle,low design efficiency,difficult algorithm translation,and difficult floating-point operation.Reference SG modeling design method applied in the field of digital signal processing,combined with the requirements of control algorithm design.This paper presents the design framework of floating-point control algorithm based on SG.In the face of abstract and complex fuzzy algorithm design,in-depth understanding of the essence of the algorithm,using SG basic mathematical module to build complex intelligent algorithm.The experimental results show that the floating-point fuzzy PI control algorithm implemented by the modeling method is successfully applied to the FSM system and achieves good control effect.It proves the feasibility of the SG model design method.2.Aiming at the demand of high speed and high precision processing for satellite laser communication.In this paper,the parallel floating-point control algorithm is realized by modeling.In the face of the problem of timing violation in algorithm construction,this paper proposes a large combinatorial logic of inserting registers and splitting to improve the throughput rate of algorithm processing.The experimental results show that compared with the traditional fixed-point PI control algorithm,the floating-point control algorithm designed in this paper has high precision operation,faster convergence of the control system,and smaller steady-state error.The fastest fuzzy PI control algorithm realized by FPGA can reach 47.828 MHz operation frequency.Compared with DSP and other serial processors,the control algorithm based on FPGA implemented in this paper has high processing performance.3.The nonlinearity of the control system increases due to multiple vibration sources in the space.And it is not easy to maintain the controller manually in the space.In this paper,the traditional PI control and fuzzy PI control are combined to realize and improve the self-tuning fuzzy PI controller.The hardware experiment results of FSM control show that the designed fuzzy PI controller has higher control precision,faster system convergence and better control performance than the traditional PI controller,whether it is step response or sinusoidal signal tracking.In this paper,the area average method is proposed to solve the center point value of edge fuzzy sets online.Through the fuzzy algorithm is improved to improve its reliability.In terms of FPGA design method,the floating-point control algorithm is realized by using System Generator model design,which improves the development efficiency and practicability of FPGA and expands the application of FPGA in the control field.For the requirements of satellite laser communication,the parallel floating-point fuzzy PI control algorithm is realized by FPGA,which improves the processor performance,operation accuracy and control accuracy.It provides important engineering application value for the establishment and maintenance of satellite laser communication link. |
