Design of imitating model of mining machine

An importance of solving the scientific problem of identifying deviations of a mining machine in the extraction of potash ore and other ores is determined. The impossibility of direct application of both existing positioning systems inside buildings and underground positioning systems offered on the market for this purposes is shown. That is why difficult conditions during the mining and high vibration. It is proposed to determine the deviation of the mining machine by indications of the distance to the mine wall sensors installed on the sides. The neural network could be used as an identification subsystem in the future. For the learning, a model is needed for simulate data from sensors according with a predetermined deviation. Indications of the sensors is imitated by a simple geometrical way inside the pixel on the monitor screen trailed by a segment of the cutting edge of the mining machine. A simulation model of a two-dimensional underground movement of a mining machine is created. It allows to set deviations of different types and to simulate the indications of distance sensors at the same time. These calculations are based on determining the point of rotation of the mining machine during a small deviation from a straight course of movement. Further, the next position of the machine and the pixels painted by the cutting edge during the movement are determined by an explicit method. The number of pixels between the sensor and the non-shaded area in the direction perpendicular to the axis of the mining machine is evaluated through the scale into the distance to the bottom wall. The error of sensors with a predetermined spread and its statistical distribution is also simulated. The possibility of qualitative identification of evasion by the indications of four sensors and possibility of using the model for learning a neural network are shown.