Research On Intelligent Control Technology Of Hydraulic Turbine Governor

The hydraulic turbine regulating system is an extremely complex closed loop control system with the characteristics of nonlinear,time-varying,non minimum phase and so on,and it is difficult to establish a precise mathematical model.At present,the hydraulic turbine regulating system of most hydropower stations in our country is still using conventional PID control or conventional PID control loop based control strategy.Although the control strategy has the advantages of simple and easy operation and strong robustness,it does not have the power of on-line adaptive tuning parameters,that is,it can not adapt to the adjustment of PID parameters with the dynamic changes of the operating parameters of the control system.It will cause the speed fluctuation of the controlled unit,so that the frequency of the power grid will be swaying.It can also be seen that the hydraulic turbine governing system is an important part of the water turbine generator set and even the power grid.The quality of the regulating quality of the unit will be directly related to the quality of the electric energy.In order to solve the above problems,a variety of intelligent control algorithms have emerged in the control field and have been widely used in other industries and achieved good results,but the intelligent control in the hydraulic turbine regulation system has not achieved more ideal results.The reason is that there are many problems in intelligent control during the actual production process,which can not meet the special requirements of hydraulic turbine governing system.Therefore,in view of the defects of the BP neural network PID control,this paper uses the principal component analysis method to optimize the BP neural network,in order to improve the convergence speed and generalization ability of the network.The design of a new controller(main component-BP neural network PID control)is applied to the simulation analysis of the water turbine regulation system,and is compared and analyzed with the conventional PID control,fuzzy PID and BP neural network PID control,which proves the superiority and feasibility of the new control strategy.First,through the analysis of the structure and principle of the hydraulic turbine regulating system,the mathematical model of the approximate linearized hydraulic turbine governing system in the small wave transition process is established.Then the two typical working conditions are simulated and analyzed with the 4 control strategies respectively using MATLAB and Simulink simulation platform.The results show that the principal component-BP neural network PID is controlled by the main component.The transition process of the step response curve is obviously better than the other 3 kinds of control.The performance indexes,including the adjustment time,overshoot and the frequency of oscillation,have been greatly improved.Therefore,this research has certain theoretical significance and application value.At the end of the paper,we summarize the paper systematically,and make a brief explanation and prospect for future research directions and problems to be solved.