Condition monitoring method of power plant equipment based on improved bp neural network

Abstract

A condition monitoring method of power plant equipment based on improved BP neural network. Aiming at the complexity of the status of power plant equipment, the neural network is used for status monitoring, but the neural network has the problem of being easily trapped in a local optimum. First, the neural network is improved to obtain the global optimal solution. According to the real-time requirement of power equipment data in the power plant and the advantage of less calculation time when genetic algorithm is optimized, the neural network is optimized by genetic algorithm. The invention proposes a genetic algorithm based on the individual migration-expansion mechanism, which accelerates the convergence speed while ensuring the convergence to the global optimal solution. After that, the present invention utilizes the BP neural network optimized by IM-EMGA to monitor the state of the power plant equipment, so that the omen of failure can be found in time. The simulation experiment of the model of the present invention shows that the method has better convergence speed and global optimization ability than the BP neural network optimized by simple genetic algorithm.

Description

technical field [0001] The invention relates to a state monitoring method for power plant equipment, in particular to a state monitoring method for power plant equipment based on an improved BP neural network. Background technique [0002] Because the working state of power plant equipment changes without regularity, it is difficult to use the existing mathematical model to estimate it. Neural network has excellent nonlinear fitting ability, suitable for fitting complex working conditions, but it is easy to fall into local optimal solution, each fitting result may be different, genetic algorithm can handle any form of objective function and Constraints, it has good global optimization ability and fast convergence speed, but genetic algorithm also has premature phenomenon. [0003] The working state of the power plant equipment directly affects the operation of the power plant, and the detection of the power plant equipment has become an important issue in the research of th...

AI Technical Summary

Problems solved by technology

[0002] Since the working state of power plant equipment changes irregularly, it is difficult to use the existing mathematical model to estimate it

Neural network has excellent nonlinear fitting ability, suitable for fitting complex working conditions, but it is easy to fall into local optimal solution, each fitting result may be different, genetic algorithm can handle any form of objective function and Constraints, good global optimization ability and fast convergence speed, but genetic algorithm also has premature

Images