基于BP神经网络的套筒式串联推力轴承组均载性能及承载能力细化研究

Detailed Research on Load Uniformity and Load-bearing Capacity of Sleeve Type Tandem Thrust Bearings Based on BP Neural Network

为完善套筒式串联推力轴承组优化设计理论,为大型、高端推力轴承组国产化提供理论支持,利用有限元数值模拟方法,对轴承组进行静态力学分析,研究轴承组几何设计参数对其均载性能和承载能力的影响。通过BP神经网络建立轴承组几何设计参数与轴承组均载率的数学模型,并用于轴承组的优化设计。对部分理论研究内容进行了实验验证,并进行了误差分析。

To perfect the optimization design theory, realize the localization of large-type and high-quality thrust bearings, enormous static mechanical analysis was done on the thrust bearings via finite element modeling analysis to study the influence on the load uniformity and load-bearing capacity from geometric parameters of the thrust bearings. The mathematic model of geometric parameters and load uniformity was established by BP neural network, and it was used to optimize the design of bearings. The theory was proven by experiment and the error was analyzed.