Research Of Control System Monitoring And Diagnostics Based On Cloud Theory

Industrial process controllers can’t achieve the design performance and even seriously deteriorate after a long time run. Without timely monitoring of performance and take effective measures to repair the controller, it will reduce the economic efficiency of the control system. In recent year, the research of control performance assessment has attracted increasing attention. This thesis from the uncertainty in the control system starting, the introduction of Li Deyi academicians cloud theory to express the uncertainty in the control system, take full advantage of the fuzziness and randomness of the information, completion of the mapping between the qualitative and quantitative. The method based on cloud transformation (and AHP) achieves assessment function, to make the results more intuitive. In the diagnosis, it gives a new method of fault diagnosis combining cloud model with the traditional association rules technology to achieve control system monitoring and diagnostics based on cloud theory. This thesis includes:(1)The review of the current research status and development trends in control performance assessment and fault diagnosis technology is presented.(2)A review is presented of cloud theory of history, research status and development trends, and introduces the basic concepts of the cloud theory and its generation method and basic computing.(3)Cloud theory application in complex control system. The concept "multi-dimensional cloud" is introduced. Express each control loop with a one-dimensional cloud model, multi-dimensional cloud model is the synthesis of one-dimensional cloud model.By comparition of the multi-dimensional cloud model between test cloud model and standard cloud model, we get the similarity coefficient, then to arrive at the results of the assessment of system performance. Prove the effectiveness of the method through simulation.(4)AHP is introduced on the basis of cloud theory, to build a more reasonable functional hierarchy system.The method that expresses evaluate set and relative value of index through the Cloud model is easier to effectively deal with uncertainty information of system, to achieve a wider range of practical applications.(5)It gives a definition of cloud association rules, mining methods and algorithms combining cloud model with the traditional association rules technology. Fault detection mechanism is proposed based on cloud association rules, and then conduct a pneumatic actuator simulation examples. The results indicate that pneumatic actuators fault detection and diagnosis method based on cloud theory has good reliability and robustness and is batter able to deal with fuzziness and randomness issues on the detection and diagnosis and the method in fault detection has a good timeliness.