Abnormal Detection Of Boiler Main Steam Temperature Based On KPCA And LSSVM

With the development of industry,the application of large boilers is becoming more and more extensive,and the safe operation of boilers is becoming more and more important.The main steam temperature system is a very important part of the boiler system.Once the main steam temperature system is abnormal,the boiler will have problems.At present,the abnormality of the main steam temperature system,especially the abnormality of the superheater,is mostly completed through macro and micro inspection,chemical analysis and other methods.Although the accuracy rate is guaranteed,the response is delayed,and it is often impossible to analyze the abnormality in the first time..In response to this situation,this paper proposes to design an abnormality detection method for the main steam temperature system based on the kernel principal component analysis(KPCA)and the least square support vector machine(LSSVM)classification model based on the monitoring variables of the thermal process.The main content of the paper is as follows:(1)For the abnormality of the main steam temperature system,it is proposed to use the combination of KPCA and LSSVM algorithms to build a detection model,that is,first use KPCA for data feature extraction,and input the processed data into the LSSVM classification model for classification.Use the public data set to test the classification effect of the KPCA-LSSVM model under different parameters.It is found that parameter changes have a great influence on the classification accuracy of the model.Appropriate parameters can greatly improve the performance of the model;(2)Aiming at the problem of blindness of LSSVM model parameter selection,the improved sparrow algorithm(CDE-SSA)is proposed for parameter optimization.First,the CDE-SSA algorithm is designed on the basis of the sparrow algorithm: in view of the poor global search ability of the sparrow algorithm and the shortcomings of the unstable distribution of the initialized sparrow,the chaos theory is merged into the sparrow algorithm to establish the chaotic sparrow algorithm model;for the discoverers in the algorithm The defect of the position update method is improved,and the new mutation operator is used to replace the original operator to improve the practicality of the algorithm;in order to balance the global search ability and local search ability in the early and late stages of the algorithm,the cosine factor is introduced to improve the proportional operator in the Sparrow algorithm,Introduce the number of iteration information into the iteration process.The performance of the CDE-SSA algorithm is compared with the PSO algorithm and the SSA algorithm through the test function.The results show that the CDE-SSA algorithm has better performance in the optimization accuracy and optimization stability;then the CDESSA algorithm is used to determine the LSSVM parameters Find the best,build the KPCA-CDE-SSA-LSSVM anomaly detection model,and apply it to the main steam temperature system anomaly detection scene,the detection accuracy rate reaches98%,which proves the rationality of the algorithm in this paper;(3)On the basis of the above theory,using JAVA and MATLAB mixed programming method,the main steam temperature anomaly detection platform based on the Spring Boot framework is realized,and the technical personnel are assisted in the abnormal analysis,which makes this article more realistic.