Early Warning Method for Power Station Auxiliary Failure Considering Large-Scale Operating Conditions

Abstract

Large-scale deep variable load of thermal power units will increase the hidden trouble of auxiliary equipment failure. Timely detection of minor faults and early warning are conducive to the safe operation of units. With the large-scale deep variable load of units, the operating conditions of auxiliary equipment also change in a large range, and the operating parameters fluctuate in a large range, which directly affects the accuracy of fault early warning algorithm based on operation data. In order to solve this problem, this paper uses a multivariate state estimation technique (MSET) algorithm for fault early warning. Firstly, the cluster method is used to divide the operating conditions of units, and process memory matrix is constructed under different operating conditions for modeling and calculation. The simulation results show that compared with the traditional method using one process memory matrix this method significantly improves the accuracy of fault early warning.