Research on a road roughness classifier built by the wavelet neural network using an immune algorithm to generate initial values

Accurate classification of road roughness is critical for road maintenance and vehicle driving dynamics improvement. However, existing neural network-based classifiers suffer from problems such as large fluctuations in recognition accuracy and difficulty in determining structural parameters. To this end, this study proposes a classifier that incorporates an immunization algorithm to generate initial values for wavelet neural networks, aiming to overcome the limitations of the existing techniques and to improve the accuracy and stability of the classification of road roughness. The results show that the classifier achieves 99.94 % recognition accuracy while utilizing a smaller number of training samples, which validates the effectiveness of the classifier.