Shearer memory cutting strategy research basing on GRNN

Shearer height adjusting is a key technology for coalmine unmanned working face. On the basis of establishing Shearer working face mathematical models, this paper determined related parameters influencing the Shearer height adjusting, then analyzed traditional Shearer memory cutting strategy and pointed out its shortcomings. Aiming at changing technical limitations of Shearer height adjusting currently, this paper proposed a new Shearer memory cutting strategy based on GRNN(General Regression Neural Network). According to height adjusting data acquired from Shearer working face, we use MATLAB to analyze the new Shearer memory cutting algorithm, results shows GRNN network approximation error is ±0.02m, and GRNN network prediction error is ±0.025m. The experimental result shows that the new Shearer memory cutting strategy has higher prediction accuracy and better generalization ability.