Research On The Coordinated Control Of Unit Plant Based On Neural Network Inverse Methods

In present, the control strategies of unit coordinated control system still dependon PID and feed-forward compensation theory. Because of complex control objects,unit plant involves several domains and multivariable in its equipment and principle,which has some features of time-delay and time-varying nonlinear dynamic property.Furthermore, traditional control strategy based on single loop model limits the qualityof the control system. So it is significant to explore advanced control strategy toenhance power process.In this paper, we analyzes deeply the nonlinear multivariable objects, thermaladvanced control strategies problems, then proposes an adaptive inverse control as anew kind of decoupling method, used in the complex, unknown and uncertain of unitcoordination control unit. Through using the two methods, the inverse system controlmethod and neural network identify the multiple input multiple output of nonlinearunit coordination control system. This method is to use adaptive inverse controlwhich is the inverse of the controlled objects as series controller to the dynamiccharacteristics of the system for the open loop control so as to avoid the feedback. Forthat may cause unstable problem. At the same time, this method can work to thedynamic property of the system control and the disturbance of control object handledseparately and each other is not affected. This is a very valuable characteristic. Thecontroller is adaptive, will the control to make objects and the controller can achieveoptimal overall dynamic response. The feedback is only used in the adaptive processby itself, not in order to control system of signal flow, but to control system ofvariable parameter.Neural network adaptive inverse control is a direct adaptive control method ofthe system based neural network model. With this technique we will achieve lowdemand and good robustness. Because of the adaptive capacity, and its traits, the strategy can be applied to the advantages of digital computer control; also can achieved goodquality control in the industrial process control.