Research on Coal Mine Gas Safety Evaluation Based on D-S Evidence Theory Data Fusion

Abstract

In order to improve the accuracy of coal mine gas safety evaluation results, a gas safety evaluation model based on D-S evidence theory data fusion is proposed, and multi-sensor fusion of gas safety evaluation is realized. First, the prediction results of the weighted least squares support vector machine are used as the input of D-S evidence theory, and the basic probability assignment function of each sensor is calculated by using the posterior probability modeling method, and the similarity measure is introduced for optimization. Secondly, aiming at the problem of fusion failure in D-S evidence theory when fusing high-conflict evidence, the idea of assigning weights is used to allocate the importance of each evidence to weaken the impact of conflicting evidence on the evaluation results. In order to prevent the loss of the effective information of the original evidence after modifying the evidence source, a conflict allocation coefficient is introduced on the basis of fusion rules. Finally, a gas safety evaluation example analysis is carried out on the evaluation model established in this paper. The results show that the introduction of similarity measures can effectively eliminate high-conflict evidence sources; the accuracy of D-S evidence theory based on improved fusion rules is improved by 2.8% and 15.7% respectively compared to D-S evidence theory based on modified evidence sources and D-S evidence theory; as more sensors are fused, the accuracy of the evaluation results is higher; the multi-sensor data evaluation results are improved by 63.5% compared with the single sensor evaluation results.