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

The running of cooling dehumidifier is characterized by strong coupling, large delay and nonlinearity, so it is not easy to establish a precise quantitative model for fault diagnosis. Aiming at this problem, a fuzzy classifier optimized by adaptive genetic algorithm (AGA) is proposed for the dehumidifier fault diagnosis. Firstly, the data acquisition and experiment system is built and the dehumidifier work statuses are simulated. Secondly, the fuzzy classifier for fault diagnosis is built. The classifier fuzzy rules and membership functions are step-wisely optimized by AGA to improve the model output precision, and a novel nearby mutation operator is proposed in order to extract the rules more accurately. Finally, the fuzzy classifier is validated and also compared with the conventional fuzzy classifier. The results demonstrate that this proposed model optimized by AGA is not only effective for the dehumidifier fault diagnosis, but also has advantages over the conventional model.