Fault Diagnosis For Power Plant Condenser Based On Optimized ELM Algorithm

The power industry is the basic supporting industry of the national economy.The condenser is an important part of the steam turbine generator set of the power station.The condenser device composed of it is used as the cooling device in the steam cycle of the steam turbine,so the fault diagnosis of the condenser is crucial,with realistic security and economic significance.This paper describes the current situation and development trend of fault diagnosis for power plant condenser,and analyzes the condenser system and its common fault categories and the signs of faults.Using MATLAB as the development tool,the Extreme Learning Machine(ELM)algorithm is the core.Firstly,the singular value decomposition method and principal component analysis method are used to determine the number of hidden layer nodes of the ELM algorithm.Then,the improved genetic algorithm and improved gray wolf algorithm are used to optimize the input layer weight and threshold of ELM algorithm.The improved artificial bee colony algorithm is used to optimize the kernel function parameter matrix and penalty coefficient of nuclear limit learning machine,combined with DS evidence theory.Finally,based on the above theoretical methods,the fault diagnosis model is established,and the fault symptom data set training model is adopted.Then,the trained model is used to diagnose the test instance,and the fault category of the condenser is obtained.In addition,this paper develops a condenser fault diagnosis system based on B/S structure,which can realize fault category diagnosis based on fault symptom data.The condenser fault diagnosis method proposed in this paper can effectively identify the fault category,which can provide reliable support for the condenser fault repair personnel and minimize the maintenance cost to ensure the safe operation a nd economic benefits of the power station.