Control System Performance Evaluation And Optimization And Its Application

For a long time, the research on the control theory has mainly focused on various control strategies instead of the performance assessment and optimization of running controllers. Meanwhile, it is realized that the performance of control system will degenerate as time goes by if there is no periodical maintenance. Inferior control capability will influence system validity, reduce product quality and increase manufacture cost. Therefore, only those process control systems which have been well designed, tuned and maintained could bring stable benefits.Many control loops exist in electric power system. And the dynamic characteristics of dynamo groups, various technology equipments and auto-control devices are very complex. In the actual system, controllers will be considered eligible as long as the variety of control parameters stays within certain limits no matter the changing trend in the interior system. There are no measurable guidelines on quantity of control performance so that the modification will be implemented only when the system suffers distinct deterioration. While in practice any system capabilities changes gradually so that it is necessary to make a measurable guideline to assess auto-control systems and monitor the real-time changes. But the auto-control system performance guideline made from the views of safety and economy is still blank in electric power system. So the control performance assessment and optimization not only is important theoretically but also accords with the requirements of complex industry process control.In this paper, the control performance assessment and optimization isintroduced into the electric power system and mostly studied.1. The design of synthesis assessment cost function. The minimum variancecontrol algorithm is used on the stochastic performance assessment, andtraditional dynamic guideline is used on the deterministic performanceassessment. The synthesis assessment performance function is gained using the two assessment methods, which is satisfied with the stochasticperformance and deterministic performance.2. Optimization strategy based on self-learning neural network controlalgorithm and self-tuning immune fuzzy control method, that makes thecontrol systems have a certain adaptive capacity. Thereby, the controlperformance assessment and optimization can be implemented.Furthermore this way is introduced into the control systems of the powerplant.In the Chapter 3, the prevalently used control performance assessment method at present is introduced. And the weighting parameters are introduced. Via the weighting parameters, the synthesis assessment performance function is acquired using the two assessment guidelines including minimum variance and tradition performance index.Chapter 4 discusses that self-learning neural network algorithm and immune fuzzy control method are used to adjust the parameters of the PID controller, which can optimize the control system. In the chapter 5, we use the weighting parameters of the synthesis assessment function to acquire the adjusted value of PID controller according our request. Through the simulation experiment, the validity of synthesis assessment performance function is proved, and the controller parameter optimization using neural network and immune fuzzy methods control is also proved.