Introduction Coal mine roadheaders operate under complex conditions characterized by prolonged exposure to high vibration and dust levels, resulting in a high failure rate. Traditional fault diagnosis methods suffer from issues such as low diagnostic accuracy and poor real-time performance. This study suggests an intelligent diagnosis model for coal mine roadheader faults based on the artificial fish swarm algorithm, particle swarm optimization, and a backpropagation neural network in an attempt to improve the precision and effectiveness of fault diagnosis for coal mine roadheaders and guarantee safe equipment operation. Methods Initial data is processed through statistical methods, correlation analysis, and normalization. A backpropagation neural network is selected as the fundamental diagnostic model, with the artificial fish swarm algorithm and particle swarm optimization algorithm introduced to perform global optimization of its initial weights and thresholds. The network is trained using forward propagation and error backpropagation mechanisms, and its outputs are converted into fault probability distributions. Results Experimental outcomes indicated that the training loss and test loss values of the research method differed by 0.01. Its classification accuracy remained consistently above 95% across varying TS proportions. In practical application testing, the best fitness and AF of the research method both exceeded 0.8 overall. Its omission rate reached a stable value of 9.8% at an 80% load rate. Discussion The above results demonstrate that the research methodology exhibits high diagnostic accuracy and efficiency, effectively addressing issues such as insufficient precision and low detection efficiency in traditional approaches. This enhances the safety and reliability of intelligent fault diagnosis for coal mine roadheaders.
Feng et al. (Mon,) studied this question.