Identification Of Moving Loads Based On The PSO-SVM

Due to the influence of the moving load of the vehicle and the overload of the transportation,the bridge is prone to fatigue damage and even collapse,threating the safety of bridge seriously.So,identification of moving vehicle loads on bridges to determine the magnitude and type of moving loads on the bridge has important theoretical significance and application value for the health monitoring and daily maintenance,safety evaluation and traffic planning.Moving load identification technology can get the real-time load information of moving vehicles on the bridge.Compared with the traditional weighing system,the vehicle data can be measured without destroying the bridge structure and without interrupting the traffic,and can effectively reduce the cost,to help improve the non-traffic law enforcement capacity of the bridge,and effectively reduce the occurrence of overload and reduce the damage to the bridge caused by overloaded vehicles.It can provide more accurate data for the health monitoring of the bridge,and provide the basis for the design of the bridge and the research of the load spectrum,so the moving load identification technology is playing more and more important role.Compared with the traditional theoretical equation to solve the identification method,artificial intelligence machine algorithm-Support Vector Machines has the advantages of simple model structure and strong non-linear processing,so this paper proposes a SVM forecasting model and its application in identification of moving loads,and optimizes the traditional SVM prediction model,optimizes the relevant parameter of SVM with particle swarm optimization algorithm,builds a improved SVM prediction model,forecasts the moving loads on the two simply supported girder bridges of the highway bridge in Haiyan area with the PSO optimized SVM prediction model,traditional SVM prediction model and neural network model,and compared the results,it shows that optimized SVM model prediction has higher precision,and makes better effect.