Study of the possibility of controlling ore averaging by methods of classical control theory

The importance of the problem of ore averaging at potash enterprises and the search for the optimal set of measures to eliminate the problem are shown. The problem of a large spread of insoluble residues in the potash enterprises of the Verkhnekamsky District is identified. At the moment, it is solved by bunker averaging, but this does not always work effectively. It was suggested to use the previously described method of meaningful distribution in the warehouse and targeted sampling depending on the composition. A mathematical model of loading and unloading of the warehouse was constructed; algorithms and calculation of the coordinates of the point of discharge and extraction of ore were proposed, depending on the content of insoluble residue and potassium chloride in the ore. This method excludes the possibility of manufacturing defects and carries out the averaging of raw materials in an optimal way. According to the indicators in the simulation model, targeted sampling in the warehouse reduces the percentage spread of insoluble residues in the ore. It was decided to investigate the sampling process in the warehouse for identification. Purpose of work is to test the possibility of controlling the sample as a conventional technological object using a proportional-integral-differentiating controller. To do this, the control object was identified, namely: a single impact jump was applied to the system input. Materials and methods. The standard impact was modeled on a previously developed warehouse simulation model, where the geometric parameters of the warehouse, the physical parameters of the ore elements, as well as the parameters of the nozzle and scraper movement are set. With its help, potassium chloride from ore is conducted. The results of the ore sampling are recorded for the initial installations, and then after a five percent jump. The simulation results are presented as a normalized graph for comparing the results and determining the behavior of the system. Result. The resulting array of values was moved to the previously developed transfer function calculator. Based on the values found, a smoothed normalized graph was constructed, which had to be identified. As a result of this work, the transfer function of the first-order aperiodic link with a delay was obtained. Conclusion. When analyzing the graphs, a conclusion about the validity of the obtained function was made. Based on the obtained arrays of values, an error of 6,5% was calculated. The transfer function has been identified, so the sample in the warehouse can be controlled using a proportional-integral-differentiating controller.