Research On Optimal Working Parameters Control Method Of Drawing Machine Based On Improved BP Neural Network

With the rapid development of science technology and economic society,the development of electrified railway has become the ultimate way to realize the modernization of railway network.High speed trains obtained a steady stream of power from the traction power supply system through pantograph networks to ensure high speed and normal operation of train.There would be major trafic accidents if the power supply was not available for the high speed trains.The contact wire was an indispensable transmission medium for pantograph catenary of electrified railway.it was necessary to bear strong vibration and fast temperature and other difference tests.If the contact wires were broken in the train operation,it would cause the train not to get the current.Aiming at the instability problem of contact wires in the production process,for the purpose of producing high quality contact wire products and considering the influencing factors in the cold drawing process,the intelligent algorithm is applied to carry out the research and analysis of the contact wire cold drawing equipment.The main contents of this paper are as follows:(1)Research on factors affecting stable drawing of contact wires products and modeling of contact line in cold drawing process.A mathematical model between tension and speed was established.Meanwhile,according to the principle of equal flow rate of material flowing through the cold drawing die,a mathematical model between speed and the diameter of the cold drawing die was established,and the key factors of stable drawing are determined..(2)Based on improved BP algorithm for drawing speed stability control,the real-time control parameters are adjusted online.In the process of operation,the algorithm of standard BP neural network has slow convergence rate,and it was easy to fall into the local minimum areas.Considering the influence of error increment and relative error,a new adaptive learning rate function formula was build.In order to increase the possibility of the local minimum area of the error surface of the network escaping,an adaptive momentum factor function formula was build.(3)The optimum working parameters matching of the contact wires in cold drawing process studying.The BP neural network was applied to the optimum matching of the working parameters in the cold drawing process of the contact wires,and the intelligent matching of the main working parameters was realized.DDE communication and MPI protocol were used to achieve mutual communication between data,and the neural network achieved the matching of cold drawing parameters.(4)The control system of the contact wire drawing machine and the real-time monitoring system studying.According to the working flow of drawing machine,S7-300 series PLC and configuration software STEP 7 were used to design the control program,and the correctness of the control program was verified.Through the upper computer displaying,the drawing process control on the MCGS on-line monitoring was realized.