Research On Fault Intelligent Warning Method And Its Application

There exist some faults in most aspects of industrial production.Once a failure occurs,it will inevitably affect the normal course of industrial production,and it will result in a certain amount of property damage,even environmental pollution.Therefore,fault early warning becomes a better choice to reduce drilling risk.Because there are certain limitations for manual predicting faults,it is a current direction to adopt intelligent technology for fault early warning.So it is very important to study the intelligence warning algorithm.And Drilling process is the high-risk and high-cost system engineering.Especially in the complex and variable geological conditions,the faults of drilling may occur at any time.So the study of intelligent early-warning algorithms and the development of an intelligent early-warning system have significant practical significance for the drilling process.The main content of this article,1.Because the time-series data collected in the drilling have the unstable inner points and irregular outer points,and the least-squares method cannot get the perfect result of linear fitting according to these time-series data,this paper proposes a dynamic piecewise linear fitting algorithm based on random sample consensus.The experimental results verify the validity and practicability of the proposed method.2.As the real-time data of drilling process has the characteristics of high dimensionality and large disturbance of parameters,this paper proposes a moving window sparse Principal Component Analysis(MWSPCA)to monitor faults.And it extracts the degrees of contribution of the parameters with high correlation to the fault.3.A case-based reasoning(CBR)algorithm based on decision tree is proposed to solve the problem of low efficiency of case retrieval in case-based reasoning.Firstly,the decision tree is used to judge the type of the fault according to its characteristics to narrow the scope of case retrieval.Secondly,an integrated similarity matching retrieval algorithm is proposed.Finally,the similarity matching results are used to sort the cases,and the optimal solutions are provided for reference by the experts.4.The paper designs and implements an intelligent warning system for the fault of drilling process based on C/S architecture,including three modules,data processing,fault monitoring and fault diagnosis.