基于故障树和贝叶斯网络的液压支架故障风险分析

Fault risk analysis of hydraulic support based on fault tree and Bayesian network

煤矿作为复杂社会技术系统,其智慧化发展使得正常安全生产对于机械安全可靠性的要求越来越高,有必要对煤矿关键机械设备的故障风险开展准确快速分析。为了解决液压支架故障风险评估中存在的主观性大、结果不准确问题,提出了基于故障树和贝叶斯网络的综合风险分析方法。首先建立液压支架故障树,并将其映射为贝叶斯网络,获得所有事件的概率后与贝叶斯网络的条件概率相结合进行故障风险推断,利用Netica软件实现后验概率推理与敏感性分析。实例验证表明,该方法能较准确地对煤矿机械故障风险进行定量分析,得到影响液压支架故障发生因素的重要程度。

As a complex social technology system,the intelligent development of coal mine makes the requirements of normal safety production for mechanical safety and reliability higher and higher.Therefore,it is necessary to carry out accurate and rapid analysis on the failure risk of key mechanical equipment in coal mines.In order to solve the problems of subjectivity and inaccurate results in risk assessment of hydraulic support failures,a comprehensive risk analysis method based on fault tree and Bayesian network was proposed.Firstly,the fault tree of hydraulic support was established and mapped to a Bayesian network.The probability of all events was obtained and combined with the conditional probability of the Bayesian network to infer the fault risk.The posterior probability inference and sensitivity analysis were realized using Netica software.The example shows that this method can accurately analyze the fault risk of coal mine machinery,and obtain the importance of the factors affecting the fault occurrence of hydraulic support.