Fuzzy Self-Turning Of Parameters Online For Human-Simulated Intelligent Controller And Its Application In Cart-Double Pendulum

Hierarchical and multi-mode control structures using the characteristic identification taken from the single mapping structure of the traditional controller, solve the feasibility problems in the control of a complex system. The complex controller, however, has numerous characteristic and control parameters, making it very difficult to design and optimize. A mathematical model of the controlled plant is needed in order to design the controller using an off-line evolutional computation. However, there will always be differences between the model and the real system, due to events that take place in the outside environment. Consequently, the controller parameters need further adjustment before they can be used effectively in the control of a real system. For a nonlinear motion process like the swing-up and stabilization control of a cart-double pendulum, the control parameters need to change with the motion attitude of the two rods of the pendulum. This means that online self-tuning of the control parameters is very important for a fast, high-quality realization of real-time control.Parameters'turning is an important tache in the process of control, the thought of it is in the different phase of systerm response automatically turn the value of control parameters, to adapt the system's reliability, adjustment accuracy and dynamic index request. If the controller with intelligent of adptive, the reserch of complex aystem will be more useful.The performance of self-organizing, self learning and adaptive in fuzzy control,automatically turn the rules and control parameters according to the system. This kind of controller further introduced human's learning and adaptive ability based on the typical controller, makes the control with higher intelligent.In this paper, a fuzzy self-tuning structure of Human Simulated Intelligent Controller's Self-Tuning Level was designed. This structure has fuzzy self-tuning ability, in the swinging up control of a cart-double pendulum system(CDPS); it efficiently accomplishes online control parameters'self-tuning of Human-Simulated Intelligent Controller (HSIC) with multimode control structure and hierarchical structure.What we dealed with is a typical fast, multivariable, nonlinear, and absolutely unstable system, is the ideal platform of teaching and research, many control conception like system's stability, controllability, velocity of convergence and ability of anti-intrrference and so on, can be directly diagramed by pendulum syatem.Based on the HSIC, the main work of this paper are:①Various kinds of PID controller tuning methods are discussed in this paper, and gave a formal description of parameter turning theory in HSIC, then a parameter self-turning theory of HSIC is summed up.②Introduced on-line fuzzy self-tuning into the HSIC controller of pendulum, designed a controller of swinging up and handstand cart-double pendulum.③Finished the simulation and real-time research of swinging up and handstand control of CDPS, results show the controller's effectiveness and robustness of the method.The computer simulate-on and real time control experiments of the swinging up control of a cart-double pendulum system show that the online fuzzy self-tuning of control parameters markedly enhances the robustness and adaptability of HSIC.