Improved sewage treatment fault diagnosis method integrating weighted extreme learning machine

Abstract

The invention discloses an improved sewage treatment fault diagnosis method integrating a weighted extreme learning machine. The improved sewage treatment fault diagnosis method comprises the following steps: S1, aiming at a basic classifier, carrying out assignment on an initial weight of the weighted extreme learning machine by adopting an assignment formula inclined to minority class samples; S2, training the basic classifier; S3, providing a novel integrated algorithm basic classifier weight to update the formula; taking the weighted extreme learning machine as the basic classifier; integrating a plurality of basic classifiers by adopting an Adaboost iteration method; establishing an improved sewage treatment fault diagnosis model; S4, inputting sample data generated in a sewage treatment process and setting the quantity T of the basic classifiers of an integrated algorithm, the optimal kernel bandwidth gamma of the basic classifier and a corresponding optimized regularization coefficient C; establishing a fault diagnosis model of a sewage treatment system to carry out performance test. According to the improved sewage treatment fault diagnosis method disclosed by the invention, the classification of unbalanced data of a plurality of types can be realized and the classification performance of the unbalanced data, especially the classification accuracy of minority classes, is improved; the accuracy of fault diagnosis in a sewage treatment process is effectively improved.

Description

technical field [0001] The invention relates to the technical field of sewage treatment fault diagnosis, in particular to an improved integrated weighted extreme learning machine sewage treatment fault diagnosis method. Background technique [0002] Sewage treatment is a complex biochemical process with many influencing factors. It is difficult for sewage treatment plants to maintain long-term stable operation, and failures may easily cause serious problems such as substandard effluent quality, increased operating costs, and secondary pollution of the environment. Therefore, it is necessary to treat sewage The operating status of the treatment plant is monitored, and operating faults are diagnosed and dealt with in a timely manner. [0003] The fault diagnosis of sewage treatment process is actually a problem of pattern recognition, and the problem of unbalanced distribution of sewage data sets is often encountered in the classification process. The traditional machine lear...

AI Technical Summary

Problems solved by technology

[0002] Sewage treatment is a complex biochemical process with many influencing factors. It is difficult for sewage treatment plants to maintain long-term stable operation, and failures may easily cause serious problems such as substandard effluent quality, increased operating costs, and secondary pollution of the environment. Therefore, it is necessary to treat sewage Monitor the operating status of the treatment plant, diagnose operating faults and deal with them in a timely manner

[0003] The fault diagnosis of sewage treatment process is actually a problem of pattern recognition, and the problem of unbalanced distribution of sewage data sets is often encountered in the classification process

Traditional machine learning methods tend to make the classification accuracy biased toward the majority class, but in actual classification, what is more important is the classification accuracy rate of the minority class, that is, the classification accuracy rate of the fault class.

Images