Research On Intelligent Evaluation System For Metal Structure Of The Quayside Container Crane Based On Data Fusion

With the globalization of trade and the diversification of the market,the port has increasingly shown its pivotal position.As the main force in the port terminal,the port crane has undertaken most of the handling work.The quayside container crane is a typical representative of the port crane.Its structure is complex and the working speed is fast.Once an accident occurs,the consequences will range from the crane damaged slightly to the one destroyed and deaths of workers.At the same time,with the increasing service life of the crane and the corrosion of moisture and salt in air,the failure rate will increase year by year due to the decrease of metal structure performance.Therefore,it is of great significance to carry out the safety evaluation of the quayside container crane.In this paper,the skeleton of the crane—metal structure is taken as the research object.The intelligent evaluation method based on data fusion is studied aiming at the problems that the existing safety evaluation methods are too subjective,the evaluation results being not real-time,and the evaluation system is not intelligent enough.The purpose of the research is to improve the objectivity and accuracy of the evaluation results and the intelligence of the evaluation process.The main work of this paper is as follows:(1)Data fusion and artificial intelligence algorithms are used to construct an intelligent evaluation method of the crane based on data fusion.The algorithms of information layer,feature layer and decision layer of data fusion structure are explained clearly,and the functions of each layer are planned.(2)The BP neural network is trained by using the index data of metal structure of the crane which is recorded during the overhaul of Zhanjiang Port.Aiming at the deficiency of the index data of the evaluation result of the whole crane,a method based on the Miner rule for the evaluation result is proposed.Because of the defect of gradient descent method,some improved BP neural network algorithms are compared by data training,and the conclusion that LM algorithm converges fastest is drawn.Then according to the superiority of LM algorithm,the LM-BP neural network of the metal structure index of the crane is established and the evaluation result of the whole crane is obtained.(3)The entropy weight method is introduced into the fuzzy analytic hierarchy process for objective weighting.For the problem of missing information caused by calculating membership degree,the entropy weight method is improved and the fuzzy analytic hierarchy process based on improved one is obtained.After the objective weight and the subjective one are combined,the evaluation value of the crane is obtained.The D-S evidence theory is used to make fusion judgment due to the evaluation value obtained by the improved fuzzy analytic hierarchy process being quite different from the one obtained by the LM-BP neural network.Finally,a relatively reasonable conclusion is obtained.(4)According to the LM-BP algorithm neural network,fuzzy analytic hierarchy process based on improved entropy weight method and D-S evidence theory,the intelligent evaluation program for the crane is compiled by using C++language,which realizes the real-time and intelligent evaluation for the crane.