Abstract:In the practical application of fault diagnosis in large-scale chemical systems, due to the danger of faults, the machine learning diagnosis methods used in current research often face the problem of scarcity of fault samples. This makes it difficult for the model to be trained effectively, thus affecting the fault detection rate of the model. Therefore, for the fault diagnosis of chemical system with small sample data, this paper proposes a diagnostic method that uses structural causal model to further combine the diagnostic results of the two machine learning methods, thereby improving global diagnostic performance. With the help of powerful system description ability of structural causal model, we can establish a causal graph for diagnostic process of the chemical system, and accurately construct the corresponding structural equations for the various diagnostic results of the two machine learning methods to combine them, thus improve the fault detection rate. The global results make up for the shortcomings of single machine learning method in small sample chemical system fault diagnosis. The verification on the chemical system simulation platform Tennessee Eastman process shows that, compared with the diagnosis results of the selected Gaussian Naive Bayes method and the K-nearest neighbour method, the results obtained by our method are effectively improved in fault detection rate of each fault in TEP.

Study of Inference Pattern in Fault Diagnosis

The Influence of Fault Diagnostic Precision to Transformer Fault Diagnosability

Intelligent Fault Diagnosis of CNC Machine Tools Based on Rough Set Theory

Sigmoid Model: A Simulation for Inference Process of Engineering Diagnosis

Fuzzy Modeling and Inference in Model-Based Fault Diagnosis

Study on Fuzzy Inference Method for Fault Diagnosis Expert System

Building a Knowledge Base of a Fault Diagnostic System for Chemical Processes

Process fault diagnosis with model- and knowledge-based approaches: Advances and opportunities

A Survey of Knowledge-Based Intelligent Fault Diagnosis Techniques

A general design technique for fault diagnostic systems

A Practical Approach To The Integrated Expression Of Shallow And Deep Knowledge In Fault Diagnosis Systems

Fast Diagnosing Through Diagnostic Graph Analysis

DIAGNOSTIC MODEL OF INSULATION FAULTS IN POWER EQUIPMENT BASED ON ROUGH SET THEORY

Research on the Key Technology of Fault Diagnosis Expert System of Reducer

Hierarchical fault detection and diagnosis method with probability information for complex dynamic system

The research and application of intelligent fault diagnosis methods

A Universal Fault Diagnostic Expert System Based on Bayesian Network

Study on Fault Multi-Symptom-Domain Integrated Diagnosis Based on Fuzzy Integral Theory

Compound Approach for Fault Diagnosis on Nonlinear System

Research on Fault Diagnosis Method Based on Structural Causal Model in Tennessee Eastman Process

Deep Expert Network: A Unified Method Toward Knowledge-Informed Fault Diagnosis Via Fully Interpretable Neuro-Symbolic AI