Path Planning Of Multiple Virtual Humans In Complex Maintenance Space

Virtual maintenance is an indispensable research content for the design of ships,and virtual human path planning in virtual maintenance is an important part of virtual maintenance.Ships maintenance with complex virtual environment,repair equipment or tools large space is small,multi-virtual human synergy and so on.In view of the above characteristics,this thesis summarized the existing research content,and studied the path planning of multiple virtual humans in a small space and a complex environment in virtual maintenance.The main research content includes the following aspects.First of all,because the virtual maintenance environment changes frequently according to design adjustments,this thesis adopted the LPA* algorithm based on two-dimensional grid graphs,which can re-plan faster when faced with environmental changes.According to the complex characteristics of the virtual maintenance environment,the threedimensional space was uniformly meshed,and the three-dimensional grid was converted into a two-dimensional grid by means of additional attributes,which laid the foundation for the LPA* algorithm.In addition,considering the large amount of calculations for 3D grids and 2D grids in complex environments,the use of CUDA for parallel calculations of grids and comparison of parallel and serial calculations reduced the calculation time of grid graphs and improved design efficiency.Secondly,in view of the problem that the narrow and conventional path planning in complex environment may not be able to plan a reasonable path,combined with the traffic characteristics of the virtual human,the virtual human is going straight,sideways,crawling or climbing under certain conditions.The virtual human can pass through some special spaces.Combined the virtual human’s way of travel with the LPA* algorithm,and added the influence of the way of travel on the LPA* algorithm to the rhs(s) of the LPA*algorithm,making the algorithm suitable for complex spaces.Next,the influence of the relative position and direction of multiple virtual humans on path planning was studied,and the influence was added to the rhs(s) of the LPA*algorithm.In addition,the path interference of multiple virtual humans and the planning sequence of virtual humans were studied,and the waiting and backtracking strategy was adopted to solve the situation of blocking the way.Then the aforementioned improvements were summarized in the LPA* algorithm.Finally,the feasibility and effectiveness of the algorithm were verified through the secondary development of CATIA and the corresponding software modules were developed.