Abstract:Auto-associative neural network (AANN) is a typical nonlinear principal component analysis method, which is widely used in industry for fault diagnosis purposes, especially in nonlinear systems. However, the basic AANN often suffers from “smearing effects” problems that may lead to misdiagnosis, particularly with regards to the complex faults involving multiple variables. In this work, a new reconstruction-based AANN (RBAANN) method is proposed to enhance the capacity of fault diagnosis. In RBAANN, a generic derivative equation is developed to investigate the effects of AANN model inputs on the prediction error between model inputs and outputs. Based on the derivative equation, the reconstruction-based index for single or multiple variables, which is defined as the minimum prediction error, is obtained by tuning the corresponding model inputs iteratively. However, without the prior knowledge of the real faulty variables, all the possible variable sets need to be evaluated by the reconstruction-based index, and this may result in an exhaustive search and cause a huge computational burden. Thus, a branch and bound algorithm is introduced into RBAANN to solve the variable selection problem. Finally, an efficient fault diagnosis strategy by integrating RBAANN and branch and bound algorithm (BAB-RBAANN) is implemented to further pinpoint the source of the detected faults. This BAB-RBAANN method can handle both single and multiple variable(s) faults for nonlinear systems without prior knowledge efficiently. The effectiveness of the proposed methods is evaluated on a validation example and an industrial example. Comparisons with other methods, including principal component analysis techniques, are also presented.

Fault diagnosis for nonlinear systems via neural networks and parameter estimation

Fault Detection and Diagnosis for Nonlinear System Based on Neural Network On-Line Approximator

Neural Network-Based Model-Free Adaptive Fault-Tolerant Control for Discrete-Time Nonlinear Systems with Sensor Fault

Fault diagnosis algorithm of a class of nonlinear system based on parameter estimation

An approach to nonlinear fault diagnosis based on neural network adaptive observer

Fault estimation approach for a class of nonlinear systems based on Hopfield neural network

Fault Diagnosis Based on Neural Network and Frequency-Domain Kernel Estimation of Nonlinear Circuit

Fault diagnosis of dynamic systems using the ANN-based MMAE method

Actuator Fault Estimation for A Class of Nonlinear Descriptor Systems

An Approach to Fault Diagnosis for Nonlinear System with Different Operating Points

A Fault-tolerant Controller for Unknown Nonlinear System Based on Neural Networks

Model-based fault diagnostics of nonlinear systems using the features of the phase space response

Neural-Network-Based Adaptive Fault Estimation for a Class of Interconnected Nonlinear System with Triangular Forms

Neural network-based fault detection for nonlinear networked systems with uncertain medium access constraint: Application to motor systems

Fault Detection and Identification of Dynamic Systems using Multiple Feedforward Neural Networks

Robust Fault Diagnosis Based On On-Line Adaptive Approximator For Nonlinear Systems

A New Reconstruction-Based Auto-Associative Neural Network for Fault Diagnosis in Nonlinear Systems

Fault diagnosis for a class of nonlinear uncertain systems using deterministic learning approach

Computation-Efficient Fault Detection Framework for Partially Known Nonlinear Distributed Parameter Systems.

Actuator Fault Detection and Estimation for a Class of Nonlinear Systems

Adaptive Fault Estimation and Fault-Tolerant Control for Nonlinear System with Unknown Nonlinear Dynamic