Research On Fault Early Warning Of Blasting Equipment Based On Nonlinear State Estimate Technique

The blower is an industrial production auxiliary equipment and plays an important role in many fields such as metallurgy,power generation and chemical industry.Once a fault occurs,it will directly affect the economy and safety of industrial production.The parameters of the blast equipment are various during operation.How to extract valuable information from a large amount of data to monitor the status,fault diagnosis and fault warning of the blast equipment has become a hot research topic.Based on the above background,this paper selects the fault warning of the blast equipment as the main research,and uses nonlinear state estimate technique(NSET)to model the operating state of the blower,and realizes the fault warning through the double sliding window similarity method,which can be divided into the following aspects:First of all,the article introduces the NSET modeling principle in detail,and analyzes the structural characteristics and common faults of the blower system.This paper mainly studies the data preprocessing method for failure warning of blast equipment,the modeling data is preprocessed,and the main component analysis is used to simplify the modeling parameters,and the important monitoring parameters with high contribution rate to the principal component are obtained.A preliminary screening of the original vibration signal is performed by wavelet decomposition technique to select the time period during which the equipment is in normal operation.Secondly,This paper studies the operating state model of blast equipment based on NSET.According to the actual situation,the similarity operator and process memory matrix of NSET model are optimized respectively.The normalized Euclidean distance is used as the similarity operator,a method of constructing a process memory matrix based on Mahalanobis distance and equidistant sampling is proposed.Through simulation experiments,The validity of the model is verified,and the improved NSET model has higher prediction accuracy.Finally,a fault warning scheme based on the similarity of double sliding windows is proposed.In order to capture the dynamic development process of faults more intuitively,this paper defines the similarity function between observation vector and prediction vector to replace the traditional residual threshold method,and uses the analytic hierarchy method qualitatively and quantitatively to determine the weight of each monitoring variable.In order to improve the sensitivity and accuracy of fault warning,a double sliding window statistical method is used to statistically analyze the similarity sequence.The fault warning threshold is determined according to the minimum average similarity of the sliding window in the normal state,and the alarm information is sent if the average similarity corresponding to the new observation vector exceeds the fault warning threshold of the double sliding window.This article takes an accident of a blower factory in Hunan as an example for experimental research.The result proves that the scheme can effectively reduce the false alarm and discover the early abnormality of the blast equipment in time,and win valuable time for fault repair.