Fusion And Interaction Technology Of Virtual And Real Models For Aerospace Structural Parts Processing Based On AR

Aerospace structural parts need to meet the extreme service environment and have extremely high requirements for product quality and performance.The processing process involves complex information and has a huge impact on the processing results.The traditional information transmission mode of the processing process has problems such as discrete information sources and backward expression,which makes it difficult to ensure processing efficiency and product performance.The digital twin can realize the virtual and real mapping of the workpiece and achieve the purpose of processing information integration.The augmented reality technology is an efficient humancomputer interaction technology that can fuse virtual information into the physical world.This paper studies an augmented reality-based digital twin process information fusion and interaction technology for aerospace structural parts.The main contents of the paper include:(1)Augmented reality-oriented digital twin machining process information modeling and representation system: In view of the current low integration of machining process information and lack of efficient representation means,first,analyze the workpiece machining process and machine tool machining process based on the idea of ontology modeling;Then the processing information is classified according to static information,dynamic information and reasoning information,and an integrated representation model of processing information is established.Finally,an augmented reality-oriented digital twin processing information representation system is proposed.(2)The augmented reality multi-view generation and interaction of the digital twin processing system: address the problems of information redundancy and difficulty in obtaining information in the digital twin processing process.First,generate different augmented reality view objects according to the current processing time and viewing angle;automatically select the most relevant views to form the augmented reality multiview based on the processing context information;then analyze the human-computer interaction method in the multi-view interaction process Finally,the dynamic generation of multi-view based on the context of the processing process and the multi-view interaction method based on vision,gesture and voice are proposed.(3)Virtual and real fusion registration tracking method: Aiming at the problem of lack of texture on the surface of the workpiece and the complicated processing environment,which makes registration and tracking difficult.First,obtain the different process workpiece models from the workpiece digital twin model and generate matching templates.Then in the observation stage,the LINEMOD template matching method is used to obtain the approximate value of the current camera pose,and the precise pose is calculated by interpolation,and the virtual view object is superimposed to the corresponding position.Finally,real-time tracking without object perspective is achieved by fusing vision and inertial sensor data based on a multi-state Kalman filter algorithm.Experimental results show that this method has high registration efficiency,stable view display and good performance.(4)Digital twin processing information representation prototype system development and case verification based on augmented reality: Designed and developed a digital twin processing information representation prototype system based on augmented reality.The prototype system development includes system architecture design,data acquisition implementation and augmented reality System release.The architecture design includes real-time data reading function,processing equipment control function,real-time view generation function,and virtual and real integration registration tracking function.In the real processing environment,the various data generated during the processing are integrated and the virtual and real registration is visualized,which realizes the efficient visualization of twin data and verifies the effectiveness and feasibility of the method.