Fault Simulation and Diagnosis for Heater Exchanger Base on BP Neural Network and Performance Parameters Deviation
Download as PDF
DOI: 10.25236/ictmic.2020.012
Author(s)
Shidan Chi, Xudong Gao, Yan Liang, Xundong Hu, Tao Luan, and Yan Gao
Corresponding Author
Shidan Chi
Abstract
There are problem of complex characteristic variables and low diagnostic efficiency in the performance monitoring and fault diagnosis of heaters in thermal systems. This paper prevents a composite diagnostic model based on the BP neural network model. This model can simulate the heat transfer process of heater and performance monitoring and fault diagnosis of heaters. At the same time, the concept of heater target value and benchmarking is proposed. Nine optimal characteristic variables are selected from the 18 parameters that affect real heat transfer process of the heater as input parameter, three variables are output parameter. The results show that: This model can realize the calculation and benchmarking of output parameter’s target value in off-design operating condition. The heater failure diagnosis under off-design operating conditions can be realized by benchmarking result. This model has higher credibility and reliability.
Keywords
Heater, Fault Diagnosis, Neural Network, Target value, Performance Parameter