Research On PID Controller Performance Monitoring,Assessment And Optimization Technology

The fact that DCS and FCS technology have been more and more mature and widely used, provides engineers with opportunities to maintain and return the Control systems. The intense competition and extended Profit Point force many companies into placing more and more attention on maintaining and improving system performance in everyday or even every operation. Maintaining PID system with a superior performance is an important tool to ensure the efficient normal daily operation, and cope with the typical fluctuations and anomalies in the system. Next, great PID performance enables advanced control technology to run by the optimized border, closely tracking optimized system orbit, as a great surpport for the system to maintain high-quality operation and achieve planned operational goals. PID controller supports the basic operation of systems and brings to the upper advanced control system the most direct and significant impact.Therefore, more attention are putted on the monitoring, assessment and optimization of PID control loop. As the controllers are initially designed and turned according to certain performance requirements. As time goes by, working conditions of most industrial scene changes, making many PID controller performance degradation. In order to solve time-varying problems of PID loop, in this article, we propose a model mismatch monitoring and assessment optimization algorithm based on statistics. This method uses model detection technology to solve squared deviation problem, employs PID accessibility assessment of minimum-squared variance techniques to assess and optimize the system. The major research work and innovation points of this article are follows:1,On the basis of reading domestic and international literatures, this charter discusses the technical background,the development trend and current situation of performance assessment techniques China and abroad, makes detailed summary of its status and existing problems. Pave the way and lay the theoretical basis for follow-up Chapters. 2,Use moving-window BJ model-based prediction error algorithm to complete closed-loop identification based on Gauss-Newton method. In the end, simulation examples demonstrate the accuracy of BJ based model prediction error method parameters of the closed-loop identification.3,In order to resolve the problem of declining performance of PID controllers caused by system time-varying in operating system conditions, this paper presents a specific method, namely: PID cycle assessment optimization algorithm. This algorithm uses the system closed-loop input and output data, employs the PID minimum variance criterion to assess the performance of PID controller, calculates PID controller parameters; compares its theoretical minimum variance with the reality one to judge whether the on-line optimized PID parameters to be used. Simulation verified the effectiveness of this method.4,This article proposes a statistical model mismatch monitoring,assessment and optimization algorithm. This algorithm makes use of online Model detection method, responds immediately when the system model changes, combines with PID loop evaluation and optimization algorithm to debugging the system. By this way , it makes PID cycle assessment optimization algorithm specifically applied to control problems arising from the time-varying conditions. As can be seen from simulation figure, this algorithm has a very strong identification of mutations on the system, makes the optimization algorithm respond quickly, therefore, improving the capability of optimizing and debugging the time-varying systems to a great extent. 5,At last, summarizes the full-text, and lookes to the future research work.