FPGA Implementation Of Model Prediction Controller For High-speed Target Plasma

In the process of returning to the ground,the hypersonic vehicle in near space will be affected by plasma sheath,which will cause communication interruption and abnormal monitoring.To solve the above problems,it is necessary to build an experimental device on the ground to simulate the reentry environment of the target aircraft and to study the exact impact of plasma sheath on communication and detection.The core of ground test facility is a highspeed target plasma generator,which can generate a high-speed target plasma that meets the requirements under control.Therefore,it is of great significance to study accurate and efficient high-speed target plasma control systems.In this paper,the application of model predictive control in high-speed target plasma control system is studied.For the engineering requirements of high-speed target plasma control system,the hardware implementation of model predictive controller on FPGA is also studied.In this paper,the model predictive control is applied to the high-speed target plasma control system,and the model predictive controller is designed based on FPGA.For the quadratic programming problem of model predictive control,this paper uses dual active set algorithm to complete the optimal solution.After implementing the FPGA-based model predictive controller,this paper uses d SPACE to build a real-time semi-physical simulation platform to complete the simulation verification of the controller.The specific work includes the following three aspects:1.Complete the design of model predictive controller for high-speed target plasma.According to the control requirements of high-speed target plasma,the corresponding mathematical model is established.Then,based on this design model,the controller is predicted,and the dual active set algorithm,which is simpler and easier to implement,is used to solve the quadratic programming problem.The off-line simulation test of high-speed target plasma model predictive controller is completed by using MATALB to verify the control effect of the controller.2.Aiming at the problem that model prediction control is slow in online calculation,this paper implements model prediction controller based on FPGA to improve the calculation speed of model prediction control.On the basis of floating-point IP core as the bottom module,the matrix operation module,problem solving module,objective function module,multiplexing module and data transmission module are compiled by Verilog HDL.After instantiating each sub-module under the top module,ISE is used to analyze,synthesize and layout the cabling,and the overall design of model predictive controller is completed.3.Based on d SPACE,a hardware-in-the-loop simulation platform is built to perform realtime hardware-in-the-loop simulation experiments of plasma density and rate tracking for high-speed Le targets.Firstly,the hardware-in-the-loop simulation model of high-speed target plasma is built by using MATALB/Simulink.After compiling the model,the hardwarein-the-loop simulation platform is built by combining d SPACE and FPGA.Then,real-time simulation experiments of plasma density and velocity tracking for high-speed targets are carried out based on the platform.The experimental results verify the real-time performance and effectiveness of the model predictive controller based on FPGA.