Design And Research On Diesel Speed Governor Based On Q-learning And Variable Universe Fuzzy-PID

The electronic governor plays an important role in the diesel engine. It is not the best strategy to control the diesel,a complex system with the characteristics of nonlinear,multi-input and multi-input, and variability,through traditional PID control,which is also known as linearity control. In addition, with the change of external circumstance, the PID parameters applied in orthodox cannot be change of external circumstance, as a consequence, the rotate speed of diesel cannot back to the set speed.An integrated control algorithm which combines the Q-learning and variable universe fuzzy is proposed in this thesis. The performance of the speed governor is improved via adjusting the PID parameters, and this adjustment is completed by variable universe fuzzy-PID controller. But as the time goes on, the control function of variable universe fuzzy-PID controller will be ditortion, which results in control inaccuracy error. The PID becomes more precise as the Q-learning a reinforce learning algorithm, is introduced to the optimization design. The mean value model for D6135 diesel engine is built by Matlab/Simulink. The simulation results showed that the speed control performances attained by Q-learning and variable universe fuzzy-PID controller is better than the traditional PID.