基于神经网络的减振器性能仿真

现有减振器模型多为简化模型，无法有效应用于工业生产中减振器阻尼力特性模拟.基于某系列汽车筒式液力减振器的实验数据，利用BP神经网络技术，建立了减振器内部结构参数与阻尼力-速度特性之间的非线性映射模型.该方法只需将减振器零部件特征参数作为模型输入，就可以得到阻尼力-速度特性，可以直接用于工业生产中减振器性能测试.自主开发的减振器性能仿真软件测试结果表明，该模型具备很好的阻尼力-速度特性模拟和预测能力，可以减少设计过程中的实验重复次数，提高了设计效率，模型精度可以通过增加实验数据得到不断提高.

Performance simulation of shock absorbers based on neural network

Traditional simplified models of shock absorbers cannot effectively simulate the force-velocity performance of shock absorbers in industry fields.By using back-propagation(BP) neural network's nonlinear mapping ability,a nonlinear model to describe the relationship between the structure parameters and the force-velocity performance of shock absorbers was established based on the experimental data of vehicle tube hydraulic shock absorbers.The model can be directly applied in product tests with absorbers' characteristic parameters as inputs and force-velocity performance as output.Test results obtained by the self-developed simulation software illustrate that the model can successfully simulate and predict the force-velocity performance of absorbers and improve the design efficiency.With increased test data,the model precision can be further improved.