Online Identification and Monitoring Method for Damage of Hydraulic Gates Driven by Digital Twins
A digital twin-based online identification and monitoring method for hydraulic gate damage is proposed to address issues related to complex identification processes,poor real-time performance,and limited visualization in the hy-draulic gate damage identification process.Firstly,a method for acquiring twin data of the gate is introduced,and a conv-olutional neural network fused with attention modules is used to extract features from the twin data and construct a real-time data-driven gate damage identification model.Subsequently,digital twin technology is employed to build a virtual re-ality scene for hydraulic gates,with the damage identification results mapped onto the virtual geometric model of the gate,facilitating real-time visual monitoring of gate damage.Finally,this method is applied to a hydraulic gate experi-mental platform,and a gate damage identification system is developed using the Unity3D platform.The results demon-strate that the constructed hydraulic gate damage identification model achieves an accuracy rate of over 95% in identifying damage states,with an average system identification time for damage of 136.68 milliseconds.This method exhibits high accuracy and fast analysis speed,confirming the feasibility of its application in hydraulic gate damage identification.
万方数据
维普
