Research On Intelligent Control Technology Of Electric Load Simulator

Load simulator is a hardware-in-loop simulation device which is used to simulate theaerodynamic loads suffered by new missiles and other aircrafts when they flight. How to inhibit thesurplus torque caused by the active exercise of actuator is the technical difficulty which need to besolved in electric load simulator. This paper studied the control of electric load simulator deeply.Firstly, the system composition and basic working principle of electric load simulator is analyzed,and the mathematical model is built based on its mechanism. Through the simulation analyzing of thesurplus torque, we can see that the existence of surplus torque greatly reduces the loadingperformance of system. The speed feed forward compensation control system based on PID isdesigned, which improves the dynamic performance of the system. However, as the sytem has suchcharacteristics such as strong disturbance, parameter time varying and nonlinear, the control algorithmhas some limitations.Secondly, the basic property of chaos is analyzed. Through studying the control parameter ofLogistic mapping and Lorenz equation, control chaos is quantitatively studied. By simulating andanalyzing the Lyapunov exponentand and the measure entropy individually, the principle judgingchaos is obtained and the Hopfield neural network mode is built. Then, a negative feed back termwhich has chaotic characteristics is introduced and chaotic neuron model is obtained.Thirdly, the chaotic neuron model based on annealing strategy is built. Through simulation, thewhole process about chaotic state which degenerates with a form of inverted bifurcation by theattenuation of self-feed back is presented. The control strategy combining the chaotic neuron based onannealing strategy and traditional PI control strategy is designed. The best P and I control parametercan be obtained by global optimization of the chaotic neuron. The simulation results show that thiscontrol strategy is better than traditional control strategy on inhibiting the surplus torque andimproving the loading performance of the system.Finally, the human-computer interface of the electric load simulator based on VC++and thesoftware of the PI control algorithm based on the chaotic neural network is designed on the industrialcontrol computer in this paper. Some relative experimental researches are also done.