Fault diagnosis for cooling dehumidifier based on fuzzy classifier optimized by adaptive genetic algorithm

Area of Science:

• Engineering
• Artificial Intelligence
• Control Systems

Background:

• Cooling dehumidifiers present complex operational challenges due to strong coupling, significant delays, and nonlinearity.
• Establishing precise quantitative models for fault diagnosis in these systems is difficult.

Purpose of the Study:

• To propose and validate a fuzzy classifier optimized by an adaptive genetic algorithm (AGA) for effective dehumidifier fault diagnosis.
• To improve the precision and accuracy of fault diagnosis models for cooling dehumidifiers.

Main Methods:

• Development of a data acquisition and experimental system to simulate dehumidifier working statuses.
• Construction of a fuzzy classifier for fault diagnosis.
• Step-wise optimization of fuzzy rules and membership functions using AGA, incorporating a novel nearby mutation operator.

Main Results:

• The adaptive genetic algorithm optimized fuzzy classifier demonstrated effectiveness in dehumidifier fault diagnosis.
• The proposed model showed superior performance compared to conventional fuzzy classifiers.
• The novel nearby mutation operator improved the accuracy of rule extraction.

Conclusions:

• The AGA-optimized fuzzy classifier is a viable and advantageous tool for cooling dehumidifier fault diagnosis.
• This intelligent approach overcomes the limitations of traditional quantitative models for complex systems.
• The method offers enhanced precision and reliability in identifying dehumidifier faults.