From performance to prediction: extracting aging data from the effects of base load aging on washing machines for a machine learning model

The standard testing of washing machine performance faces reliability challenges, which stem from the uncontrollable degradation of base load. This study provided a quantified standard for base load age to enhance the reliability and stability of the washing machine performance testing process. First, the impact of base load aging on the cleaning performance and water extraction performance was investigated. Simultaneously, the changes in reflectance of the base load were recorded. Then, MLR and ANN models were developed using the remaining moisture content and reflectance of the base load to predict base load age. It was found that the water extraction performance is more easily affected by the aging of the base load than the cleaning performance. In addition, ANN has better performance in predicting base load age, resulting in an R_c² of 0.978, RMSEC of 9.295 h, and R_cv² of 0.972 h, RMSECV of 10.639 h for the front-loading washing machine, and R_c² of 0.958, RMSEC of 12.948 h, and R_cv² of 0.940, RMSECV of 15.572 h for the top-loading washing machine. This study establishes a theoretical foundation for optimizing base load age regulation. Further study could expand the sample size to enhance the robustness and generalizability of the proposed method.