数字孪生驱动的高保真水工闸门剩余寿命预测

Digital Twin-Drivn High-Fidelity Remaining Useful Life Prediction for Hydraulic Steel Gates

针对水工钢闸门这种长寿命金属结构寿命预测难以预测的问题,提出了一种基于数字孪生技术的方法,获取闸门的全寿命周期数据,以达到剩余寿命预测的目的。首先利用有限元模型建立水工闸门的数字孪生体,并通过监测系统实时获取弧形钢闸门的静动力响应特性,在此基础上,以振动频率和结构应力作为响应,基于贝叶斯理论对闸门进行实时不确定性模型修正,最后利用修正的高保真模型获取闸门全寿命周期数据集,对闸门进行剩余寿命预测。结果表明:修正后的参数对应响应与实际响应偏差相比于仿真初始值更小,且剩余寿命预测结果与真实剩余寿命预测结果相似,在真实寿命预测的95%置信区间内,表明了方法的有效性和可行性。

Aiming to address the challenge of difficult lifespan prediction for durable metallic structures like hydraulic steel gates,a method⁃ology based on digital twin technology is proposed.This approach seeks to acquire life cycle data for the gates to enable accurate prediction of remaining useful life.Initially,a digital twin of the hydraulic steel gate is established using finite element modeling.Real-time static and dy⁃namic response characteristics of the radial steel gate are obtained through a monitoring system.Building upon this foundation,vibration fre⁃quency and structural stress are utilized as responses.Bayesian theory is then employed to perform real-time uncertainty model adjustments to the gate,based on the responses.Ultimately,by utilizing the refined high-fidelity model,a dataset encompassing the gate′s entire lifecy⁃cle is generated,enabling accurate prediction of the remaining useful life of the gate.The outcomes of this study indicate that the responses corresponding to the adjusted parameters exhibit smaller deviations from actual responses compared to initial simulation values.Moreover,the predicted remaining useful life results closely align with the actual predictions within a 95%confidence interval for real lifespan forecasts.This underscores the effectiveness and feasibility of the proposed methodology.