Mining Machine Simulation Model Refinement According to the Engines Energy Consumption Data

Abstract

The importance of mining factories planning for ore extraction is indicated. The created simulation model with open and normative data is presented. Significant deviations of the simulation results from the real data of the mining machines monthly operating time of are shown. The goal is to refine the model with the available mining machine operation real data. As a source of data, it is proposed to use miming machines engines energy consumption. The source this information is described. Information about the mining machine engines is given. That information can be used for mining machines analyzing to determine the times of the states. There are 8 such engines in total, their sequential switching on and off allows us to understand the time of 4 technological operations that make the greatest contribution to the mining machines operating time. A fragment of the algorithm which allows to determine the operations time is given. Two combines were selected with the production differs most from the model in the direction of increase and decrease. For both combines the duration of ore breaking and ore loading from the storage hopper to the self-propelled car is determined. First, an analysis was made for the mining machine that breaks more ore than was obtained in the calculations. The analysis showed the existence of a previously undescribed state of the mining machine-simultaneous ore breaking into the storage hopper and ore loading from the storage hopper to a self-propelled car After including that state in the model, a new calculation was made. Calculation shows increasing in the model total output. In addition, analysis was made of the existing stochastic delay during ore breaking. Analysis shows that statistical distribution stochastic delay can corresponds to a negative binomial distribution. Further, the study of data for a mining machine with the producing less than the model amount was carried out. It allows to study the stochastic delay during ore loading. It is shown that such a delay can also correspond to a negative binomial distribution. The simultaneous inclusion of described delays significantly reduces the total estimated production amount. In conclusion, it is discussed that the simultaneous addition of a new state and stochastic delays with a negative binomial distribution to the model significantly reduces the modeling error.