Electrical Fault Diagnosis Method for Tunneling Equipment Based on Fault Simulation and Optimized Particle Swarm Algorithm

In response to data scarcity constraints in electrical fault diagnosis of tunneling equipment, this study proposes an intelligent diagnosis model based on fault simulation and improved particle swarm optimization. By constructing dynamic equations to accurately simulate fault effects and designing the Lévy flight random walk strategy with dynamic neighborhood topology optimization, the model achieves a 34.7% convergence acceleration while maintaining population diversity. The results show that the proposed model significantly outperforms comparison models in dynamic load tracking error (0.13%) and vibration suppression (28.7 dB), with parameter identification accuracy (MSE=0.19) and response latency (27.9 ms) surpassing benchmarks. It achieves the lowest false alarm/missed detection rates in the industry (1%) while reducing hardware power consumption (0.9 W) and memory (57 MB) by over 34.5%, supporting edge deployment. The physics-data fusion framework addresses generalization issues under load abrupt mutation and high-frequency interference. In summary, this model achieves high-precision and low-latency diagnosis with minimal resource consumption, promoting intelligent maintenance of tunnel engineering equipment.