A concept of automatic tuning of longwall scraper conveyor model

The modeling of machines and their operation modes is a key approach for optimization of their performance as well as for avoiding unwanted operational states which may lead to the occurrence of faults, and finally, to the breakdown. The developed model of a machine should be always parametrized, i.e. the certain number of parameters should be selected in the certain ranges. The most of the parameters can be selected based on engineering documentation and experts' knowledge, however, for some of them this knowledge cannot be directly acquired which leads to the parameter uncertainty. One of the approaches allowing selection of these uncertain parameters is a tuning procedure of a model. The paper deals with a concept of heuristic optimization method for automatic tuning of key parameters of longwall scraper conveyor model. In the first part of the paper, the evolutionary algorithm for tuning this model is proposed. In the case study, the merits and limitations of the evolutionary approach are analysed. The obtained results prove that the proposed approach of tuning of the considered model has high practical potential and it may be applied in real mining conditions.