A Study On A Self-adaptive Fuzzy PID Controller And The Application

The performances of conventional PID controller depend on accurate system model, so it is difficult to meet the control demand of industrial process, which is complex, time-varying and nonlinear. The Fuzzy PID controller(FPID) is one of the key research technologies of intelligent control and advanced PID control, which has the combine the good dynamic response of the fuzzy control with the high steady-state accuracy of PID control. In large degree, the design of Fuzzy PID controller depends on the designers'experience, and because it lacks of systematic design method, once the rule base is constructed, it can not be updated, so the FPID is still can not adjust with the changing operating conditions and the disturbance.Because of the problems of the FPID design, a kind of adaptive fuzzy PID controller is proposed and the design process is introduced. Then the proposed controller is applied into the LED Source-meter System and the simulation of solar cell Maximum Power Point Tracking. The main researches are as follows:⑴The dynamic performance and steady-state accuracy of PD type Fuzzy controller(FPD) is mutual constraint, a linear combination of PD and PI is used to construct a PID type fuzzy controller(FPID). The simulation results show that FPID can effectively eliminate steady state error, and preserve the good dynamic performance of Fuzzy control.⑵As in the system response, the proportional, integral and derivative control parameters of the FPID influence each other, which makes it is difficulty to extract the control rules and tune the control parameters. The variable universe method is adopted, and a self-organizing mechanism is introduced on the basis of the FPID, and a Self-organizing Fuzzy PID Controller (SFPID) is designed. The self-organizing mechanism with the normalized acceleration as the indicator of system response is used to adjust parameters of the fuzzy PID controller (FPID) in an online manner. A common second-order system in industrial process is selected as the simulation object, and the SFPID characteristics are researched, including servo tracking, dynamic adjustment performance, ability to adapt to the changes of system parameters and the random interference. The simulation results show that SFPID has a large dynamic adjustment range, and its static and dynamic performance, robustness, anti-disturbance capability are better than PID and FPID.⑶SFPID is a multi-parameter, strongly coupled system, the optimization of related parameters needs large searching space. So a adaptive GA is proposed to optimize the fuzzy rules and membership functions. Firstly, a measure is given to reduce the optimized parameters. Then an adaptive GA is introduced, with elitist strategy adopted in earlier genetic evolution stage and a nonlinear selection method based on Roulette Selection used in the later period, the optimum individuals are preserved and population diversity is increased. An adaptive crossover mutation operator is applied to make the probability of crossover and mutation adjusted with evolution stage changing. Finally, simulation experiments is carried out, the results show that in the same control accuracy the optimized SFPID has more fast system response and shorter steady-state response time, compared with the normal SFPID.⑷LED has variety kinds, a highly nonlinear current-type PN junction, is a power device. The optical and electrical characteristics testing of LED need to meet different kinds of LED testing, with a large dynamic range of measurement and control. Test LED of LED products need to adapt to different, with a large dynamic measurement and control range. The SFPID is applied to set and control the testing conditions. First of all, the recursive least squares method is used to identify measuring circuit of LED testing system. And the control parameters of SFPID based on the comprehensive survey of different types of LED products. Furthermore, experiences are carried out to test the functions of sourcing constant voltage, sourcing constant current and the sweeping. The experiences results show that the SFPID has better performance and robustness.⑸The output characteristics of solar cell is typical nonlinear, it is difficult to get the maximum power using the analytical method, in order to SFPID is adopted to adjust the operating point of the electronic load. The simulation results show that the control algorithm can achieve the change of the MPP, with the light intensity and temperature changing, and the tracking curve is smooth, has good precision.