Research And Development On Eddy Current Dynamometer Loading Control System

As torque loading device, eddy current dynamometer is often used in power driveline system which consists of vehicle engine and gearbox. The loading precision and response speed of dynamometer loading system decide the result of engine bench test. The research and develop for eddy current dynamometer loading system of test equipment has important practical value.The overall function of eddy current dynamometer loading control system was firstly ensured. We designed the controller and driver of loading control system and gave out the controlling program. Then the hardware and software platform of loading system were finished.On the basis of analysis of operating principle, we gave out the output expression of torque about current and speed, and proved it with a large amount of experimental data which measured from field. This expression was feasible as mechanism model and applied to research and analysis of system.With single neural self-adaptive PID as system controller, double layers dynamic recurrent neural network and state delay input dynamic recurrent neural network as identifier, we designed single neural self-adaptive PID control algorithm based on dynamic recurrent neural network. In use of experimental data measured from field, we trained neural network and verified the result of identification. We designed constant torque experiments and constant speed experiments to simulate this algorithm. The simulation results showed that this algorithm can achieve constant torque control and constant speed control for loading system. Compared with the two dynamic recurrent neural networks, state delay input dynamic recurrent neural network has quicker response speed and stronger approximation capability.Eddy current dynamometer loading control system designed in this paper has been applied in real system already. It can achieve torque loading for power driveline device and has good loading precision and response speed.