Emulation And Prediction Of The Cold Roll Forming Force Base On BP Neural Network

High strength steel (HSS) has excellent material properties, and because it caters to the lightweight concepts of automobile, such as safety, energy saving, and economy at the present stage, HSS has been widely used in automobile industry, In the manufacturing process of the automotive structural parts, the roll forming process has high productivity and irreplaceable advantages compared with other process. With the extensive application and high-speed development of roll forming process, there need the corresponding theoretical basis. To carry out the research of metal deformation rules and characteristics of HSS roll forming has the important theory significance and the practical value.In order to optimize the craft parameter of cold roll forming, this paper puts forward a new method reflect the law between cold roll forming force and craft parameter. The method combines artificial neural network and finite element emulation for cold roll forming. On the basis of the existing roll forming force calculation theory and technology characteristics, this paper analyzes the basic principle of roll forming, and obtained the main process parameters that have an impact on roll forming force. According to the single factor experiment method, the paper designed the roller flowers and the corresponding rollers, and establishes the roll forming emulation model to simulate, and a BP neural network is constructed with the sample data from the simulation.Based on the high strength steel complex and multi-pass bending characteristics, the paper built the equivalent plastic plate to improve the computing efficiency of finite element. According to the uniaxial tensile test data, the paper obtained the high strength steel material model data through mechanical derivation. After finite element simulation, extracted each pass forming force, and there is a one-to-one relationship between with the corresponding process parameters. The data is carried out the necessary processing, which is used to train the construction of BP network, after repeated debugging network parameters, obtained a more ideal prediction accuracy.This paper combined artificial neural network with finite element emulation for cold roll forming, and emphatically discussed influence rules of the workpiece thickness and forming angle. The results show that the neural network can better reflect the nonlinear characteristics between the forming force and the process parameters. The neural network has very strong mapping ability, and reach very high forecast accuracy, thus it can be used for roll forming process optimization.