Performance Assessment Of Model Predictive Controller And Fault Diagnosis Of Batch Process

Model predictive control is widely used in actual industrial processes, but the performance will go bad gradually because of various factors, so the research on performance assessment of model predictive controller has been attracting widely attention. The benchmark designing and the constrained processing is an essential issue.Fault detection and diagnosis is important means to ensure the equipment operation safe and stable. At present there are many chemical processes control is batch process. The product of batch process is closed related to people's life. Therefore, to ensure safe and reliable operation of batch process for real-time fault detection and diagnosis is necessary.The work includes following aspects:1,A constrained generalized minimum variance controller and the related performance assessment approach is derived based on a moving horizon minimum variance performance assessment approach and generalized minimum variance control principle. The generalized minimum variance controller can account for hard constraint on process variables, and it can account for soft constraint compared with minimum variance controller. The simulations based on the Wood-Berry distillation column model show that the generalized minimum variance performance assessment approach is more practical significance than the minimum variance approach for model predictive controller and the validity of the approach is demonstrated.2,Develop a fault diagnosis software for batch process using VC++ on the visual studio development platform which is integrated dynamic multiway principal component analysis, and T2 statistic, SPE statistics, contribution chart, and Exp-PCA phased approach. Diagnostic results show that the approach can identify the correct segmentation points, detect the fault generation, and diagnos the fault sources.3,Pass balance is used in the reboiler furnace of fractionation tower. The practical application results show that the control system works well.