Performance Evaluation Of Industrial Control System Based On Data Mining

For a long time,the operating system theory research mainly focuses on the study of all kinds of control system method,and the operation of the control system performance evaluation is little research.And there is a widespread realized that if the operating system has no regular maintenance,its performance will gradually degraded over time.This article mainly completes:First,realize the control system performance evaluation by rough set theory based on the theory and features of data mining,improve the accuracy and coverage of the training data by reset attribute reduction algorithm and combined with DCS data,simulate and evaluate overheating control system,verify the correctness of the algorithm,and divide control system performance result into four levels;Furthermore for non-Gaussian systems,A new way of computing benchmark of performance assessment based on entropy index is gave for linear Non-Gaussian system,compared currently used minimum variance evaluation method with different interference noises,reflecting the inadequacy of the minimum variance assessment method,confirm the validity and accuracy of the method.