An event-driven Spike-DBN model for fault diagnosis using reward-STDP

• Published in: ISA transactions 2023-09, Vol.140, p.55-70
• Main Authors: Liu, Ying, Wang, Xiuqing, Zeng, Zihang, Zhang, Wei, Qu, Hong
• Format: Article
• Language: English
• Subjects: Event-driven; Manipulator failure diagnosis classification; Spike-DBNs; Spiking encoding; STDP
• ISSN: 0019-0578; 1879-2022
• DOI: 10.1016/j.isatra.2023.06.008

Deep neural networks (DNNs) have shown high accuracy in fault diagnosis, but they struggle to effectively capture changes over time in multivariate time-series data and suffer from resource consumption issues. Spike deep belief networks (spike-DBNs) address these limitations by capturing the change in time-varying signals and reducing resource consumption, but they sacrifice accuracy. To overcome these limitations, we propose integrating an event-driven approach into spike-DBNs through the Latency-Rate coding method and the reward-STDP learning rule. The encoding method enhances the event representation capability, while the learning rule focuses on the global behavior of spiking neurons triggered by events. Our proposed method not only maintains low resource consumption but also improves the fault diagnosis ability of spike-DBNs. We conducted a series of experiments to verify our model’s performance, and the results demonstrate that our proposed method improves the accuracy of fault classification of manipulators and reduces learning time by nearly 76% compared to spike-CNN under the same conditions. •An event-driven spike-DBN is proposed for fault diagnosis.•The event-driven spike-DBN is based on evt-Siegert neuron model.•The core of evt-Siegert is Latency-Rate encoding.•An event-driven reward-STDP learning algorithm is developed for network training.•The proposed method is demonstrated via UCI execution failures dataset.