Positioning Constraints And Intelligent Path Planning Reserched In Virtual Assembly

Part assembly is an important part of industrial production.With the increasing demand for high-quality assembly training,higher requirements have been put forward for the operating skills of mechanical assemblers.Therefore,the study of complex positioning and intelligent virtual assembly systems has important theoretical and practical value.This paper takes the assembly process of compressor components in aviation engines as the research object,studies the positioning constraints of parts in virtual assembly,and proposes to apply the proximal policy optimization algorithm in reinforcement learning to the part assembly path planning process.Based on this,an immersive virtual assembly system for compressors has been developed and implemented,and the effectiveness of the proposed method has been verified.The main research and achievements of this paper are as follows:(1)Model construction and optimization: A 1:1 compressor model for part positioning constraints and immersive assembly systems was constructed in 3ds MAX.The compressor model was processed to remove repeated parts,and the assembly structure of compressor parts was described using an assembly tree.The assembly sequence of compressor parts was calculated.In Unity 3D,various part models were optimized to solve the problem of device lag in immersive virtual reality systems.(2)Study on positioning constraints of compressor parts assembly: Starting from the spatial pose and spatial degree of freedom constraints of parts,the assembly pose constraints and spatial degree of freedom changes in the part assembly process were studied.Positioning algorithms were designed and simulated for three types of parts,and correct guidance was provided for assembly operations.Finally,the virtual assembly environment was built,and the above positioning constraint process was implemented in C# scripts.The assembly system can dynamically recognize the assembly constraint relationship between parts,solving the problem of poor realism due to the lack of positioning constraints in immersive environments.(3)Using the compressor as the object,the HTC VIVE virtual reality device and the VRTK virtual reality engine are used to develop and validate the assembly system.The system achieves diverse human-computer interaction mechanisms and a highly immersive virtual environment.Modules for mechanical structure learning and assembly operation training are developed.(4)Development of immersive virtual compressor assembly system: Using the compressor as the object,the HTC VIVE virtual reality device and VRTK virtual reality engine were used to develop and validate the assembly system,achieving a diverse human-machine interaction mechanism and a highly immersive virtual environment.A mechanical structure learning module and an assembly operation training module were developed.Based on the above methods,an immersive virtual compressor assembly system has been developed,which effectively solves the problem of poor realism in the assembly process in immersive environments due to the lack of positioning constraints.The proximal policy optimization algorithm is used to automatically generate assembly motion guidance paths for parts,reducing the cognitive and assembly burden on people during virtual training,and realizing the transformation of virtual assembly systems from focusing on functionality to focusing on user interaction experience.