Nonlinear sensor fault diagnosis method based on self-adaptive learning and neural network

Abstract

The invention discloses a nonlinear sensor fault diagnosis method based on self-adaptive learning and a neural network. The method comprises the steps that S1, a single-output consistent observable standard model is established for a system; S2, learning training and standard observation in a normal mode are performed, wherein S21, the diagnosed system is subject to learning training and state observation during normal running; S22, a high-gain observer provided with the neural network is adopted to perform learning training and state observation; and S23, the high-gain observer provided withthe neural network is adopted to perform second learning training and state observation; S3, learning training and state estimation in a fault mode are performed; S4, a mode bank is established; S5, adynamic estimator is established; S6, a state vector in the dynamic estimator is compared with a state vector of the detected system, and a residual error is generated; and S7, the residual error isassessed, so that fault isolation is performed. The method is used for fault diagnosis of the complicated nonlinear system, an unknown system mode can be learnt and quickly simulated, and therefore afault is discovered and approached.

Description

technical field [0001] The invention relates to the field of automatic control, more specifically, to a nonlinear sensor fault diagnosis method based on self-adaptive learning and neural network. Background technique [0002] In the field of automatic control, sensors are the main devices for information acquisition. With the continuous improvement of automation technology, many large-scale automation projects are increasing, and a large number of sensors for parameter measurement and state control have become It cannot be ignored that these sensors composed of precision components are often in harsh working environments, and failures are inevitable. Once the performance of the sensor deteriorates, malfunctions or fails, it will have a serious impact on the subsequent monitoring, control, fault diagnosis and other systems, resulting in misdiagnosis, false alarms, and even immeasurable losses. It is particularly important to study the problem of fault diagnosis of sensors in...

AI Technical Summary

Problems solved by technology

Once the performance of the sensor deteriorates, malfunctions or fails, it will have a serious impact on subsequent monitoring, control, fault diagnosis and other systems, resulting in misdiagnosis, false alarms, and even immeasurable losses

Images