A Storm Frame Optimization Method for Predicting and Warning the Safety Status of a Shearer

Abstract

Real-time monitoring, prediction, and early warning of operating status during intelligent mining are the key to ensuring stable production. To solve the problem of lag in determining the operating status of a shearer, this study proposes a new method for predicting and warning the real-time operating status of the shearer involving the Storm framework, based on parallel optimization of data processing and the gated recurrent unit (GRU) model based on hyperparameter optimization. First, the GRU model is optimized through hyperparameter optimization to achieve adaptive and accurate prediction and early warning of multidimensional state parameters of the shearer. Second, a virtual machine is constructed to host the Storm framework, parallel optimized real-time processing of data is performed on the Storm framework, and real-time data flow patterns are constructed to speed up data processing and retrieval, ensuring each tuple is fully processed through the topology structure. Finally, the optimized GRU model is embedded into the optimized Storm framework to achieve real-time prediction and early warning of different dimensional data of the shearer. The prediction accuracy, early warning accuracy, and processing efficiency of the Storm platform are used as evaluation indicators to analyze and evaluate the model, verifying the efficiency and applicability of the proposed model. Experimental results show that the model has a prediction accuracy of 93%, an early warning accuracy of 93.05%, and consumes 10 s. It can achieve high performance, low latency, and high precision in predicting and providing early warnings for the shearer's state parameters, greatly improving the efficiency of predicting and early warning the operating status parameters of the shearer. This model realizes real-time prediction and early warning of the shearer's operating status, providing technical support for intelligent mining in coal mines.