钻机系统故障智能诊断方法

Study on the Method of Intelligence Fault Diagnosis System for Drill

由于地层的复杂性和隐蔽性,很难用精确的数学模型模拟钻机的工作过程。目前钻机故障只能凭人的经验进行识别和处理,不能充分利用检测数据对设备运行优劣的趋势、故障发生的原因、部位及程度进行预估,也不能给出相应专家治理意见。在对钻机故障进行研究的基础上,将人工智能和专家系统应用于石油钻井中,研究钻机系统故障智能诊断的理论和方法,建立了钻机常见故障图形化模糊神经网络专家知识库,提出了由专家规则、模糊逻辑、神经网络有机结合构成的智能推理机,使钻机故障诊断系统适用于多变量、多参数、多过程的复杂系统。

It is very difficult to simulate the motion process of drill by accurate mathematics model for the complexity and invisibility of stratum. The fault of drill is usually identified and disposed by personal experience so far. This means it can not estimate the trend of the equipment running good or not and the reason conduced the fault or location and degree of fault by data measured. Farther, it can not give the expert suggestions. Based on the studying of drill fault, artificial intelligence and expert system have been used in the petroleum drilling engineering, the theory and method of fault diagnosis intelligence system for drill have been studied. It also constitutes the expert knowledge database of graphic Fuzzy Neural Network for the familiar fault of drill. The intelligence reasoning machine which consists of expert rule, Fuzzy logic and artificial Neural Network have been bring forward. And the fault diagnosis system it makes for drill can be applied in complex system which includes multi-variable, multi-parameter and multi-process.