论文部分内容阅读
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.
Aiming at the rotorcraft without collision and low altitude mission in 3-D complex mountainous environment, a high-efficiency and low-cost trajectory planning strategy is proposed. An improved sparse A * algorithm and bio-inspired neuro-dynamics Model fusion algorithm based on the sparse A * global optimization search. The algorithm incorporates bio-inspired neuro-dynamics model to local routes to accelerate the formation of the optimal trajectory, and uses the neural dynamic model to acquire and process the environment in real time Which can solve the problem of dynamic programming that can not be achieved by the traditional optimal path search algorithm.Through the multi-peak mountain, especially the concave mountain in the three-dimensional space as an obstacle The experimental results show that the fusion algorithm not only reduces the complexity and time-consuming of the A * algorithm, but also improves the cost of the biological-inspired neuro-dynamics model that has not been considered, and can deal with the emergencies in the task space online Threatening to enable the rotor UAV to plan an emergency under the complex environment of moving and static obstacles , Quick access to the target point of the low costs and high quality track.