The Research Of Coal Mine Ventilator Failure Prediction Strategy

Coal mine ventilators belong to the rotating machinery in the industrial and mining establishments and occupy an inconvenient and decisive position in the process of safety production and economic construction. But there is some fault in using these mechanical equipment, and there muse be imponderable casualties and economic loss once the coal mine ventilators out of order. So it is necessary to monitor, diagnose and predict the running statue of coal mine ventilators during the actual operation. Failure prediction is a method of estimation that use to judge the ventilators are normal or not at the next moment according to the rule summarized by the precious running status. This theoretical research has great practical significance on decreasing coal mine accident and increasing safety production and economic construction.Because the treatment effect of non-linear and non-stationary signal is not very good when using the single prediction algorithm, and there are some other disadvantages such as incompletely filtering, narrow applicability and low prediction accuracy. At this point, the article presents a combinational algorithm to predict coal mine ventilators’ fault based on wavelet and SVM. Firstly, the vibration signal measured by vibration sensor uses wavelet analysis or wavelet packet analysis to de-noise, filter and decompose and then extract the characteristic value of breakdown signal reconstructed by approximation coefficients, which is used as input vectors of SVM; Secondly, SVM can train input sample and establish forecast model and then estimate the true value of the next moment; At last, this model can apply to LabVIEW epistasis software well which can be connected seamlessly with Matlab. The experiment of the article is based on the ventilators of Huaibei Yang Zhuang mine, it mainly includes gathering vibration data, creating experiment model and applying in the process of actual use.The article presents a set of complete theoretical analysis of failure prediction on coal mine ventilators, and it also points out the research direction of monitoring, diagnosing and predicting the rotating machinery’s running status. The prediction method presented by the article can apply to normal rotating machinery, so the method can widely use on steam turbine, compressor and motor. It has practical application value on increasing the working time and safety production of rotating machinery.