Research On Fault Detection Technology Of Industrial Control Equipment Based On Data-driven

With the increasing complexity and cost of industrial control systems and the rapid development of information technology,the tolerance of industrial control equipment to performance degradation,productivity decline,and hidden safety hazards is getting lower and lower.Detecting the industrial control equipment failures immediately is important to the safety industrial control system and reduction maintenance costs are significant.With the development of the Industrial Internet,more and more industrial control equipment is connected to the network,and the data collection of the operating status of industrial control equipment is also easier,which provides a research basis for data-driven fault diagnosis.Faced with a large amount of equipment status data,traditional industrial control equipment fault diagnosis methods have been unable to meet the actual needs in terms of safety,accuracy and real-time performance.Aiming at the above problems,this paper studies the fault detection methods of industrial control equipment from the perspectives of safety,reliability,and accuracy.First,This paper introduces and summarizes the existing fault diagnosis theories,and analyzes the advantages and disadvantages of fault data preprocessing methods.In view of the industrial Internet environment,many industrial control equipment are exposed to the Internet,industrial control equipment is easy to be attacked,equipment operating status data is easily tampered,and fault data sources are not reliable.This paper proposes a fault data watermark design scheme based on invisible characters,which can ensure the security of the device status data without affecting the appearance of the faulty data.Aiming at the problems that the current industrial control equipment diagnostic algorithm is fragmented,the diagnostic architecture is not easy to expand,and the diagnostic algorithm is difficult to support massive data,a set of Dev Ops and convolutional neural network-based fault diagnosis systems is proposed,which has high scalability and can make accurate predictions of faulty inputs.Aiming at the problems that the industrial control environment is unstable and the working parameters of industrial control equipment are susceptible to change,in order to achieve intelligent prediction,a fault diagnosis method for industrial control equipment based on integrated learning is proposed.This method has strong self-adaptability,and can well face the problem of industrial control environment changes,and achieve intelligent fault detection.