3维复杂山地环境下旋翼无人飞行器高时效航迹规划策略

发布时间：2018-07-20 10:50

【摘要】：针对3维复杂山地环境中执行无碰撞低空飞行任务的旋翼无人飞行器,提出了一种高时效、低代价的航迹规划策略,设计并采用了改进的稀疏A*算法和生物启发神经动力学模型的融合算法.该算法在稀疏A*全局优化搜索的基础之上融入生物启发神经动力学模型来调整局部航线以加快最优航迹的形成,并运用神经动力学模型来实时获取和处理环境中的局部动态信息,实现了融合算法的在线规划能力,从而解决了传统最优路径搜索算法无法实现的动态规划的难题.通过在3维空间中设置多峰山地,尤其是凹形山体作为障碍进行仿真实验,实验结果表明,该融合算法不仅降低了A*算法的复杂度和耗时,而且改善了生物启发神经动力学模型尚未考虑的代价花费问题,更能够在线应对任务空间中的突发威胁,使旋翼无人飞行器在动、静态障碍物相结合的复杂环境下能够规划出一条安全、快速抵达目标点的低代价且优质的航迹.
[Abstract]:A high time and low cost track planning strategy is proposed for a rotor-wing unmanned aerial vehicle (UAV) which performs a collision-free low-altitude mission in a three-dimensional complex mountainous environment. The improved sparse A * algorithm and the fusion algorithm of biologically inspired neurodynamic model are designed and adopted. Based on sparse A * global optimization search, the algorithm integrates biologically inspired neurodynamic model to adjust local routes to speed up the formation of optimal track. The neural dynamic model is used to acquire and process the local dynamic information in the environment in real time, and the on-line planning ability of the fusion algorithm is realized, thus solving the problem of dynamic programming which can not be realized by the traditional optimal path search algorithm. The simulation results show that the fusion algorithm not only reduces the complexity and time of the A * algorithm, but also reduces the complexity of the algorithm. It also improves the cost and cost problem that has not been considered in the biologically inspired neurodynamic model, and is able to deal with the sudden threat in the mission space online and make the rotoring unmanned aerial vehicle move. In a complex environment with static obstacles, a safe and high quality track can be designed to reach the target point quickly and at low cost.
【作者单位】：湖南大学电气与信息工程学院;机器人视觉感知与控制技术国家工程实验室;
【基金】：国家自然科学基金(61433016)
【分类号】：V279

【相似文献】