| As a large lifting machine on modern port terminals,shore bridge is mainly responsible for the loading and unloading of containers on the terminals.The working environment of shore bridge is mostly high humidity,high salt and other harsh conditions,and its working conditions are complex and changeable.Maintaining the good operation of shore bridge is particularly important for efficient operation and safe production of enterprises.The hoisting reduction gearbox is the main transmission mechanism of the shore bridge.The operations of container such as emergency start,emergency stop,acceleration lifting,deceleration landing and crane displacement lead to frequent load impact on the lifting and reduction gearbox of the shore bridge.At the same time,due to the influence of the low-frequency vibration of the shore bridge,the dynamic performance of the lifting and reduction gearbox continues to deteriorate.For the above reasons,the hoisting reduction gearbox has become one of the vulnerable parts of the shore bridge.At present,the fault diagnosis of rotating machinery such as reduction gearbox and bearing has been relatively mature,but the problem of condition monitoring and degradation evaluation of lifting reduction gearbox has not been well solved.How to put forward targeted signal analysis and health status evaluation methods according to the working conditions and signal characteristics of lifting and reduction gearbox of shore bridge is one of the technical difficulties of intelligent maintenance and safety engineering in the construction of smart port.The application of big data and technology of port machinery industry provides a new method and means to solve this problem,which has theoretical guiding significance and engineering application reference value for improving the health management level of port equipment.Based on the massive vibration load spectrum,this paper deeply studies the evaluation technology and method in the process of invisible state degradation.The main research contents and achievements of this paper include:(1)Using the time-domain multi-scale degradation feature extraction method,the degradation characteristics and laws of the health status of lifting and reduction gearbox in different scales in time-domain are studied and revealed,and a degradation feature extraction method based on composite multi-scale fuzzy entropy is proposed.The complex multi-scale analysis is used to decompose the signal in the time domain,and the complexity in each time domain scale is quantitatively evaluated by fuzzy entropy and average fuzzy distance.Simulation and example analysis show that this method can comprehensively characterize the main trend and detail law of degradation in the time domain multi-scale,and provides high-quality characteristic information for the dynamic performance evaluation of hoisting and reduction gearbox.(2)Aiming at the characterization of gearbox degradation law in multi-scale time-frequency domain,the multi-scale degradation feature extraction method in time-frequency domain is further applied.This paper studies and reveals the degradation characteristics and laws of the health status of lifting and reduction gearbox in different scales of time and frequency domain,and puts forward a degradation feature extraction method based on empirical mode decomposition Kolmogorov entropy.In this method,the empirical mode decomposition is used to decompose the signal adaptively in time-frequency.The principal components are selected and the Kolmogorov entropy of each principal mode is calculated respectively to quantitatively evaluate the complexity of the signal in different time-frequency scales.Simulation and example analysis show that this method can characterize the random degree of signal in multi-time-frequency scale and quantitatively describe the law of different performance degradation.For the research on multi-scale degradation characteristics in time domain,it expands the multi-scale feature extraction method in time-frequency domain and enriches the content of degradation feature description.(3)Aiming at the problems of complex operating conditions,strong fluctuation of multi-dimensional degradation feature vector and information redundancy of shore bridge hoisting mechanism,a multi-dimensional degradation feature optimization method based on local weighted smooth point tracing technique principal component analysis(LOWESS PCA)is proposed in this paper.This method can effectively eliminate the random fluctuation in the evolution process of characteristic curve and mine the smooth and accurate degradation main trend,so as to improve the accuracy of degradation state trend analysis of lifting and reduction gearbox and lay a foundation for accurate evaluation of degradation state.(4)Finally,based on the reduced degradation feature vector,a degradation state evaluation method for fuzzy c-means clustering is proposed.Based on the above research on the degradation mechanism and trend change law of the dynamic performance of the lifting and reduction gearbox,the evaluation method applies the anomaly detection method based on sliding window OCSVM to effectively identify the phased dynamic performance of the lifting and reduction gearbox in the degradation process,and is well verified by the whole life data of the rolling bearing.To sum up,through the application research of multi-scale analysis technology,this paper puts forward targeted feature extraction and state evaluation technology,and obtains the effective evaluation of the dynamic performance of the lifting reducer through the analysis of the full fault cycle load spectrum sequence of the on-site shore bridge lifting reducer.The research content and the new method proposed in this paper have reference value in engineering practical application.At the same time,the research carried out in this paper can also provide theoretical and methodological reference for improving the state health management level of similar large machinery. |
PDF Full Text Request