基于分层贝叶斯推理的RV减速器动力学模型修正及动态响应预测方法

Dynamic Model Updating and Dynamic Response Prediction Method of RV Reducer Based on Hierarchical Bayesian Inference

RV减速器作为高端装备核心零部件其性能好坏直接影响装备整机性能。动力学分析是提高RV减速器长期稳定服役能力的重要手段。当前研究在动力学建模时未考虑多源不确定性因素的影响,导致所建立理论模型误差较大,影响后续分析精度。为此,基于分层贝叶斯推理开展RV减速器不确定性度量与传递研究,实现其动力学模型修正及动态响应预测,旨在为准确评估RV减速器动态特性提供有效保障。根据RV减速器结构特点及传动原理,建立整机与核心零部件动力学理论模型,采用Newmark法求解位移、速度和加速度响应。引入待修正参数的多层超参数概率分布,构建整机与核心零部件分层贝叶斯概率模型。充分考虑待修正参数的先验信息,运用过渡马尔可夫链蒙特卡罗法进行识别获取后验分布信息,准确修正RV减速器动力学模型。在此基础上,将参数不确定性进行传递,实现动态响应预测。针对某国产RV-20E型减速器,采用所提方法对其薄弱环节曲柄轴和摆线轮进行分析,得到两者刚度参数的后验分布信息及动态位移响应预测结果。

As a core component of high-end equipment,the performance of RV reducer directly affects the performance of the whole machine.Dynamic analysis is an important way to improve the long-term stability of RV reducer service capacity.The current research work does not consider the influence of multi-source uncertainty factors when modeling the dynamic.This leads to a large error in the established theoretical model and affects the accuracy of the subsequent analysis.Therefore,the uncertainty quantification and propagation of RV reducer is studied based on hierarchical Bayesian inference to realize its dynamic model updating and dynamic response prediction.The aim is to provide an effective guarantee for the accurate assessment of the dynamic characteristics.According to the structure characteristics and transmission principle of RV reducer,the dynamic theoretical models of overall and the core components are established.The Newmark method is applied to solve the displacement,velocity and acceleration dynamic response of each component.By introducing a multi-layer hyperparameter probability distribution of the parameters to be updated,a hierarchical Bayesian probability model is constructed.The prior information of the parameters to be updated is fully considered,and the transition Markov chain Monte Carlo method is applied to obtain its posterior distribution.Thus,the RV reducer dynamics model is accurately updated.On this basis,the parameter uncertainty is transferred to realize the dynamic response prediction.For a domestic RV-20E reducer,the weak points,i.e.,crank shaft and cycloid wheel,are analyzed by the proposed method.The posteriori distribution information of the stiffness parameters and the dynamic displacement response prediction results of both are obtained.