Application Of Artificial Neural Network In Rubber Damper Design

Rubber dumpers have excellent damping and damping properties and are widely used in vibration reduction systems.Therefore,with the rapid development of China's rail transportation,bridge building and other fields in recent years,its application is becoming more and more extensive.At the same time,the design and manufacture of rubber damper are also higher.At present,the design method of rubber shock absorber is mainly based on experience design and finite element method.Among them,experience design is mainly based on trial and error design,with long design cycle,high cost and low efficiency.The finite element method has the advantages of simplifying the design process and improving the design efficiency.However,because the development of the rubber viscoelastic theory is not perfect and the constitutive model of the material is inaccurate,the result of the finite element analysis has a certain error.In view of the above problems,this paper studies the application of artificial neural network technology in the design of rubber damper.By using the strong nonlinear fitting ability of artificial neural network,a prediction model is established to predict the dynamic and static performance of rubber damper in a certain range,so as to achieve the auxiliary rubber damping.The main work done in this article is as follows:1.Through consulting a large number of relevant documents,the present situation of the design theory and methods of rubber damper at home and abroad,the development status of artificial neural network technology and the application of artificial neural network technology in the field of rubber materials are summarized and summarized.2.By studying the vibration damping principle of rubber damper and aiming at the design process of the existing rubber damper,this paper puts forward two ideas of using artificial neural network to assist the new product design of rubber damper and the optimum design of rubber damper.3.the theory and method of rubber formulation design were studied.The influence of the dosage of reinforcing agent,filler and vulcanizing agent on the damping performance of rubber shock absorber was studied by orthogonal experimental design.The results show that carbon black has the greatest influence on the damping performance of rubber shock absorbers.4.The structure and principle of BP,RBF and GRNN artificial neural network model are studied.The performance index of artificial neural network and the objective function of damping performance are determined,and the process of using artificial neural network to assist the new product design of rubber damper is designed in detail.According to the test data,different prediction models of formula shock absorber performance are established by using artificial neural network technology.Through the comparative analysis of the prediction results,the optimal prediction model is identified as the BP network model.The experimental results show that the average prediction error is around 5%,which proves that the artificial neural network technology can achieve the purpose of designing new products for assisting rubber shock absorbers.5.Detailed design of artificial neural network aided rubber damper product optimization design process.According to the test data,a prediction model of the performance of the rubber material with fixed structure based on different neural networks is established.Through the comparison and analysis of the prediction results,the optimal prediction model is identified as the GRNN network model.The experimental results show that the average prediction error is about 7%,which proves that the artificial neural network technology can achieve the purpose of optimizing the rubber damper design.