Research On Collision Detection In The Visual Simulation Of Air Cushion Vehicle

As a significant part of Air Cushion Vehicle (ACV) Simulator, the visual simulation system should directly display the onshore and offshore movement process of ACV in virtual environment. In the ACV's motion simulation process, real-time and accurate collision detection is important for enhancing the authenticity and immersion of virtual environment. This paper initially analyzes the advantages, disadvantages and applicable conditions of collision detection techniques used in the process of visual simulation developing, and expands the techniques to meet the detection demand of ACV motion simulation system.For collision detection of ACV and static objects, ACV is simplified as a collision rectangular frame by building eight LOS types of isector on the outmost edge of ACV. The method can detect static objects roundly. The motion of ACV is more realistic and closer to the actual situation both in the visual and functional effect, meeting detection requirements of real-time and accuracy.For collision detection of ACV and dynamic objects, it is carried on in two stages. First of all, by using bounding sphere algorithm, initial collision detection is achieved to prune away nonintersecting object pairs clearly and reduces the computational complexity of collision detection. Then, based on AABB hierarchical box, more detailed collision detection is achieved for the possibly intersecting object pairs. The dendriform hierarchical structure of the model is established in the modeling process to achieve the pretreatment of the data structure which collision detection depends on, and bounding box is gotten by using software developing platform OpenGL Performer API, which can optimize the structure algorithm of hierarchical box and accelerate collision detection. Two simulation experiments established verify that the method can provide efficient and real-time detecting results for the system in certain precision.For the real-time amphibious movement of ACV, three points matching algorithm of ACV and offing in the marine environment and six points matching algorithm of ACV and terrain in the terrestrial environment are designed, the position of ACV is modified real-time. The application of the algorithms improves reality of the entire virtual scene.Finally, simulation experiment is designed among ACV driving-console, visual simulation system and movement platform. Simulation shows that the designed collision detection methods can satisfy real time and accuracy requirements.