Research on attitude monitoring and decision-making of hydraulic support based on FBG sensor and BP neural network

Abstract

Hydraulic supports are essential equipment in coal mining operations, directly affecting the safety of miners and the efficiency of the mining process. However, due to the complexity and dynamic nature of the mining environment, timely and accurate monitoring of hydraulic support posture has become a significant challenge. Recent advancements in fiber optic sensing technology, specifically Fiber Bragg Grating (FBG) sensors, have provided a promising approach to monitor the posture of hydraulic supports. Despite progress in monitoring techniques, decision-making regarding the support’s posture remains underexplored. This paper proposes a comprehensive monitoring and decision-making system for hydraulic supports, integrating FBG sensors and Back Propagation (BP) neural networks. FBG sensors are used to collect real-time data on critical parameters such as angles, displacements, and pressures of the top beam and base of the supports. The BP neural network then processes this data to establish a predictive model, allowing for real-time decision-making regarding the support posture. The system’s efficacy is demonstrated through experiments conducted at the 101 working face of the Longde Coal Mine, where real-time predictions of support posture have shown high accuracy, offering valuable support for safe and efficient mining operations.