Research On Early Warning Method Of Tailing Pond Based On BN-SVM

Tailings pond is composed of intercepted valley mouth or exclosure while damming,which is an important production facility and a major hazard source in mine engineering;its safe and stable operation is very important for the operation of mining enterprises and local ecological security.In the event of dam collapse of tailings pond,it will cause great harm to the lives and properties of the people in the lower reaches,while leading to soil erosion and heavy metal pollution and other ecological problems.It is of great practical significance to construct the tailings pond monitoring and early warning system to monitor the running status of the tailings pond in real time and to early warn and forecast the potential hidden danger.At present,the early warning of tailings pond is mainly for the saturation line,pond water level and other single factors,ignoring the internal relations between these factors,lacking of evaluation and prediction to comprehensive risk rating.Therefore,it is very necessary to study the multi-factor early warning method of tailings pond;this paper focuses on studying the multi-factor early warning method of tailings pond,and the main contents are as follows:(1)Based on the comprehensive analysis to the current research status of the monitoring and early warning of tailings pond the tailings pond at home and abroad and the characteristics of tailings pond engineering,the risk index system of dam collapse of tailings pond is established by the correlation analysis method.This paper presents a method to evaluate the risk of dam collapse of tailings pond based on structural equation modeling and fuzzy comprehensive evaluation;where the structural equation modeling is used to determine the correction of the five factors,such as overtopping break,unstability break,seepage damage and foundation management for safety,and the weight of each impact index,and then conduct risk assessment to the tailings pond through the fuzzy comprehensive evaluation method.The result shows that the evaluation result is in accordance with the actual situation of the tailings pond,and provide technical support for the monitoring and early warning of tailings pond.(2)Established the early warning model of safety of tailings pond based on Bayesian Network(BN)and Support Vector Machine(SVM),determining the relationship between the key influence factors of dam collapse of tailings pond by BN;on this basis,combining the Support Vector Machine model to learn and train the sample data set,and predict the influence factors of dam collapse.Finally,comprehensively evaluating the risk level of dam collapse of tailings pond,issuing a corresponding level of early warning.The result shows that the precipitation,the pond water level,the burial depth of saturation line,the dam body displacement,the length of the dry beach,the seepage flow and the discharge capacity can affect each other to jointly act on the stability of the dam body of tailings pond.The prediction accuracy of SVM is different for different influence factors,where the prediction accuracy of pond water level is the largest,and the prediction accuracy of discharge capacity is the smallest,and the average prediction accuracy is more than 95%;the overall prediction accuracy is high and the prediction stability is strong.Therefore,the early warning method of risk of dam collapse of tailings pond,which has integrated the Bayesian Network and the Support Vector Model,has provided an effective way for the early warning of risk of dam collapse of tailings pond.(3)Based on the ASPNET3.5 architecture,the tailings pond monitoring and early warning system is developed and built by the C # language in the environment of Visual Studio 2010 Web Developer,realizing the automatic collection,storage and management of on-line monitoring data.Based on the monitoring data,the evaluation and early warning model of the dam risk level of dam collapse of tailings pond is used to forecast,evaluate and early warm the change trend of comprehensive indexand the risk level of dam collapse of tailings pond,and conducting digitized and intelligent management and monitoring to tailings pond.