Research On A Ground Mobility System Based On Urban Combat Simulation

In urban combat simulation,the manoeuvring system,as one of the basic modules for controlling the movement of units in the scenario,is both a prerequisite for the reliable movement of manoeuvring units in the scenario and a core indicator for assessing the reliability of the simulation results.In a large scale urban combat scenario with a complex environment,it is a challenging task to plan the paths of many movable units in real time and to avoid obstacles dynamically during the movement.The difficulties are mainly in three areas: Firstly,the scenes in the map are complex and a single data format cannot organise all the scenes in the map,requiring modelling with many different forms of map data and effectively bridging the modelled map types together;secondly,the huge search space derived from the huge map and the time consumed to solve it grows exponentially with the number of nodes,making path search algorithms such as A*,D* and Dijkstra’s unable to meet the real-time requirements of the system;thirdly,the mutual interference of manoeuvring units moving through the scenes and the change in the static environment during combat,resulting in dynamic obstacle avoidance problems.In response to these issues,the paper focuses on the following:(1)In view of the large scale and complex environment of urban combat simulation scenes,a single data format map cannot organize the scenes used in the map,this paper constructs a classified and hierarchical map system to provide effective map support for pathfinding and obstacle avoidance tasks.(2)In order to meet the real-time requirements of pathfinding,this paper proposes a variety of improved pathfinding algorithms.In local pathfinding,a pathfinding algorithm based on pre-training and segmentation pathfinding is proposed,which significantly improves the efficiency of local pathfinding;in road network pathfinding,a route construction method using clustering algorithm to build sub-networks and construct subroutes is proposed,which significantly improves the efficiency of road network pathfinding;in indoor pathfinding,a median algorithm for indoor pathfinding is proposed by adding auxiliary lines,which significantly improves the efficiency of indoor pathfinding;in indoor pathfinding,a median algorithm is proposed by adding auxiliary lines,which significantly improves the efficiency of indoor pathfinding.(3)In order to improve the efficiency of obstacle avoidance,various improved obstacle avoidance algorithms are proposed in this paper.An eight-phase diagram obstacle avoidance judgement algorithm is proposed,together with a rule-based behavioural tree model for obstacle avoidance decision,which significantly improves the local obstacle avoidance efficiency;a bidirectional marked link and rule-based behavioural tree road network obstacle avoidance model are proposed;for indoor obstacle avoidance,a rule-based median obstacle avoidance algorithm is proposed,and an improved ORCA algorithm based on the social force model(SFM)is proposed.