Coal and rock identification method based on Kalman optimal estimation of load data of rocker arm pin axle of shearer

The shearer's coal and rock identification technology is the basis of intelligent control. The existing coal and rock identification method for the site environment and testing equipment requirements are higher. The actual fully mechanized working face is difficult to meet the necessary conditions. In order to solve the above problems, a coal and rock identification method based on the Kalman optimal estimation of the shearer rocker pin axle load data is proposed. On the basis of not increasing external auxiliary instruments and equipment, the rocker pin axle sensor is used to replace the existing pin axle to sense the coal and rock load, which can better adapt to the environment. By measuring the strain data of the rocker pin axle sensor located at the connection between the rocker arm and the connecting frame, the Kalman optimal estimation method is used to reduce the noise of the load data. The load intervals of the shearer under different working conditions such as cutting coal and rock are separated from each other. By judging the interval of the real-time load value, the coal and rock interface can be identified. The identification of coal and rock is verified on a fully mechanized experimental platform. The load of the rocker pin axle at the upper end of the coal wall side along the shearer traction direction is analyzed in three stages: no-load, cutting the coal wall and cutting the rock. The results show that before the load data is processed, there is overlap between the load intervals of cutting the coal wall and cutting the rock, and the coal and rock interface identification cannot be completed accurately. After the load data is processed by the Kalman optimal estimation algorithm, the load intervals under no-load, cutting the coal wall and cutting the rock conditions are separated from each other. In addition, the load interval length of each working condition is shortened by 65.6%-83.3%, and the mean square deviation is reduced by 66.5%-72.9%. The data fluctuation is smaller, which effectively improves data identification. In practical engineering applications, the expected load stress range in the coal seam cutting state can be set according to the method. Once this range is exceeded, it is judged that this is not a cutting the coal wall state and thus coal and rock identification is achieved.