Solution of the problem of odometric positioning of a mining machine under the ground by using a Kalman filter

In the modern mining industry, an urgent technical challenge is the introduction of automatic systems that provide orientation and positioning of mining machines during the development of industrial seams. There are several basic technologies used for positioning combines underground, but their scope is limited by various mining, geological and technological factors. In the conditions of industrial development of VKMKS seams, the vast majority of them are not suitable. Aim. To develop a new approach to the problem of odometric positioning of a mining machine under the ground, as well as to create a simulation model that allows with the required degree of accuracy to determine the current and predicted distance of the miner from the start of production in conditions of noisy measurements. Materials and methods. As a technical solution to the task, the use of BLE (Bluetooth Low Energy) technology is proposed: iBeacon beacons will be dropped in the direction of the combine's movement, and a sensor attached to the rear of the loading bunker will read the distance to the beacon. For simulation modeling of uncertainty during the movement of the combine, the hypothesis of the normal distribution of the speed of movement on sections of random length was considered. When simulating the dropping of the beacon, the hypothesis was used that the scattering value of the beacon upon falling is a two-dimensional normally distributed random variable. Noisy measurements were generated by a stochastic process with increasing scatter boundaries as the sensor moved away from the beacon. The Kalman filter was used as a tool for processing measurement noise. Results. A model has been created that simulates random speeds of the combine's movement on sections of random length, and also a random spread when throwing off Bluetooth beacons has been simulated. To generate sensor measurements, an algorithm has been developed that takes into account the increase in the noise level of the readings when moving away from the nearest dropped beacon. To process the simulated measurements and correctly determine the distance of the beacon-sensor, the Kalman filtering algorithm was used. Conclusion. The proposed approach and the created simulation model make it possible, with a given degree of accuracy, to determine and predict the distance to the withdrawing shearer when mining industrial seams.