论文部分内容阅读
采用数据挖掘中BP(back propagation)神经网络模型来研究逆向卸荷膜片式减压器的结构参数与稳定性能之间的依赖关系,得到结构参数变化,尤其是多结构参数耦合变化下减压器的稳定性结果.其中稳定性对阻尼孔直径、膜片刚度非常敏感,对弹性元件材料的阻尼系数、低压腔有效长度较为灵敏.由此提出减弱振荡的各种措施:增大阻尼孔直径、增大膜片刚度、在一定范围(标准值的6.5倍)内增大弹性元件材料的阻尼系数、增大低压腔有效长度、减小阀芯质量.数值实验误差分析表明:该模型不存在过拟合、局部最优的情况,其预测结果是可靠的,可为减压器的设计和系统分析提供决策支持.而且,该模型对不同类型的数据集具有通用性,可以用来研究其他部件的结构参数与性能指标的依赖关系.
BP (back propagation) neural network model in data mining is used to study the dependence of the structural parameters and the stability of the reverse unloading diaphragm pressure reducer, and the structural parameters are obtained. Especially, The stability of which is very sensitive to the diameter of the orifice and the rigidity of the diaphragm and is very sensitive to the damping coefficient of the elastic element material and the effective length of the low pressure chamber.Therefore, various measures for reducing the oscillation are proposed: increasing the diameter of the orifice , Increasing the rigidity of the diaphragm, increasing the damping coefficient of the elastic element material within a certain range (6.5 times of the standard value), increasing the effective length of the low pressure chamber and reducing the valve plug quality.Numerical experimental error analysis shows that the model does not exist Overfitting, local optimum, and predictive results are reliable and can provide decision support for the design and system analysis of the pressure reducer.Moreover, the model is versatile for different types of data sets and can be used to study other The structural parameters of components and the dependence of performance indicators.