During path planning, it is necessary to satisfy the requirements of multiple objectives. Multi-objective synthesis is based on the need of flight mission and subjectivity inclination of decision-maker. The decision-m...During path planning, it is necessary to satisfy the requirements of multiple objectives. Multi-objective synthesis is based on the need of flight mission and subjectivity inclination of decision-maker. The decision-maker, however, has illegibility for under- standing the requirements of multiple objectives and the subjectivity inclination. It is important to develop a reasonable cost performance index for describing the illegibility of the decision-maker in multi-objective path planning. Based on Voronoi dia- gram method for the path planning, this paper studies the synthesis method of the multi-objective cost performance index. Ac- cording to the application of the cost performance index to the path planning based on Voronoi diagram method, this paper ana- lyzes the cost performance index which has been referred to at present. The analysis shows the insufficiency of the cost per- formance index at present, i.e., it is difficult to synthesize sub-objective flmctions because of the great disparity of the sub-objective fimctions. Thus, a new approach is developed to optimize the cost performance index with the multi-objective fuzzy optimization strategy, and an improved performance index is established, which could coordinate the weight conflict of the sub-objective functions. Finally, the experimental result shows the effectiveness of the proposed approach.展开更多
无人机的自主飞行是无人机相关研究的重点方向,如何在复杂环境中快速分析环境,并规划一条安全可行的路径,是该方向的研究目标。针对传统路径搜索算法存在的路径不平滑问题,采用三阶B样条曲线进行预规划航迹。在欧式有符号距离函数(Eucli...无人机的自主飞行是无人机相关研究的重点方向,如何在复杂环境中快速分析环境,并规划一条安全可行的路径,是该方向的研究目标。针对传统路径搜索算法存在的路径不平滑问题,采用三阶B样条曲线进行预规划航迹。在欧式有符号距离函数(Euclidean Signed Distance Functions, ESDF)地图提供的梯度信息的基础上,分别在平滑、碰撞和可行性上设计约束方程,实现轨迹动态重规划。针对路径动态更新中,时间间隔变化产生的控制点不再符合约束的问题,采用各向异性曲线拟合方法,实现时间再分配,保障在动态更新路径的过程中,新产生的路径与原路径相似且具有同样的可行性。实验证明,该算法实现了无人机的自主路径规划与优化,能够进行动态避障,面对复杂环境具有鲁棒性。展开更多
文摘During path planning, it is necessary to satisfy the requirements of multiple objectives. Multi-objective synthesis is based on the need of flight mission and subjectivity inclination of decision-maker. The decision-maker, however, has illegibility for under- standing the requirements of multiple objectives and the subjectivity inclination. It is important to develop a reasonable cost performance index for describing the illegibility of the decision-maker in multi-objective path planning. Based on Voronoi dia- gram method for the path planning, this paper studies the synthesis method of the multi-objective cost performance index. Ac- cording to the application of the cost performance index to the path planning based on Voronoi diagram method, this paper ana- lyzes the cost performance index which has been referred to at present. The analysis shows the insufficiency of the cost per- formance index at present, i.e., it is difficult to synthesize sub-objective flmctions because of the great disparity of the sub-objective fimctions. Thus, a new approach is developed to optimize the cost performance index with the multi-objective fuzzy optimization strategy, and an improved performance index is established, which could coordinate the weight conflict of the sub-objective functions. Finally, the experimental result shows the effectiveness of the proposed approach.
文摘无人机的自主飞行是无人机相关研究的重点方向,如何在复杂环境中快速分析环境,并规划一条安全可行的路径,是该方向的研究目标。针对传统路径搜索算法存在的路径不平滑问题,采用三阶B样条曲线进行预规划航迹。在欧式有符号距离函数(Euclidean Signed Distance Functions, ESDF)地图提供的梯度信息的基础上,分别在平滑、碰撞和可行性上设计约束方程,实现轨迹动态重规划。针对路径动态更新中,时间间隔变化产生的控制点不再符合约束的问题,采用各向异性曲线拟合方法,实现时间再分配,保障在动态更新路径的过程中,新产生的路径与原路径相似且具有同样的可行性。实验证明,该算法实现了无人机的自主路径规划与优化,能够进行动态避障,面对复杂环境具有鲁棒性。
