【作者】 刘喆；

【导师】 马思乐；

【作者基本信息】 山东大学 ， 电子信息（专业学位）， 2023， 硕士

【摘要】 无人机路径规划是保证无人机飞行安全性、高效性和稳定性的关键技术。然而在复杂的高维环境中,路径规划过程变得更具有挑战性。传统算法在应对这些挑战时可能面临局限性。此外,现有无人机路径规划评估模型大多只考虑路径约束或简单的威胁约束,这会使得评估模型的准确性无法得到保证,并最终影响路径规划的质量。针对以上问题,本文的主要工作如下:(1)本文提出了一种基于改进的麻雀搜索算法(CSSA),针对原算法存在的收敛性能差的问题,我们引入了 Singer混沌映射和Levy飞行策略优化麻雀搜索算法,以提高算法的收敛性并避免局部最优,进一步将该算法应用于灰色预测模型(GM(1,1))的改进。由于灰色预测模型背景值计算公式的系数为固定值,因此模型的灵活性较差,通过利用该算法对传统的GM(1,1)模型动态生成背景值,从而降低了误差,有利于将模型进一步应用于后续的路径规划问题。总的来说,本文提出的改进算法解决了原算法本身易陷入局部最优的问题,并且提高了灰色预测模型的预测精度。(2)本文构建了一种基于改进灰色预测模型的无人机能耗预测模型。由于现有的无人机路径规划评估模型大多只考虑简单的约束,而忽视了能量消耗在无人机飞行中的重要作用,因此带来路径规划结果的不准确性和不稳定性问题。为了解决这些问题,本文根据实际飞行数据建立了能耗预测模型,该模型考虑了无人机的飞行距离和角度与能耗的关系。除此之外,同时建立多种约束和威胁的综合评价函数,为无人机提供了更加优化的路径。该能耗模型为旋翼无人机后续路径规划的基础。(3)本文设计了一种将改进的麻雀搜索算法与能耗预测模型相结合的二维地形无人机路径规划方法。目前大多数算法无法满足无人机路径规划对精度和稳定性较高的要求。针对存在的问题,本文将提出的改进算法与能耗预测模型应用于二维路径规划。仿真结果表明,该算法能综合考虑能耗、路径、精度等需求,找到复杂环境下的有效路径。该方法为无人机路径规划提供了一种更加准确、高效的解决方案。(4)本文提出了一种将改进的麻雀搜索算法与能耗预测模型相结合的三维环境下无人机路径规划方法。三维环境约束条件相比二维更加复杂,对规划效率要求更高。本文提出了一种改进的麻雀搜索算法(S-SSAE),该算法利用Sinusoidal混沌映射实现种群均匀分布,利用t分布改进了麻雀搜索算法的更新公式,以及精英反向策略对种群进一步更新位置,这三种方式共同解决了算法收敛速度慢,易陷入局部最优的问题。该算法提高了无人机三维路径规划效率以及效果。更多还原

【Abstract】 Unmanned aerial vehicle(UAV)path planning is a key technology for ensuring UAV flight safety,efficiency,and stability.However,in complex high-dimensional environments,the path planning process becomes more challenging.Traditional algorithms may face limitations in dealing with these challenges.In addition,existing UAV path planning evaluation models only consider path constraints or simple threat constraints,which may compromise the accuracy of the evaluation model and ultimately affect the quality of path planning.In view of the above problems,the main work of this paper is as follows:(1)This paper presents an improved Sparrow search algorithm(CSSA)to address the problem of poor convergence performance in the original algorithm.We introduce Singer chaotic mapping and Levy flight strategy to optimize the Sparrow search algorithm to improve its convergence performance and avoid falling into local optima.Furthermore,we apply this algorithm to the improvement of the grey prediction model.Due to the fixed coefficient of the background value calculation formula in the grey prediction model,its flexibility is limited.By using the proposed algorithm to dynamically generate the background value for the traditional GM(1,1)model,the prediction error is reduced,which is beneficial for further application of the model in subsequent path planning problems.In summary,the proposed improved algorithm solves the problem of local optima in the original algorithm and improves the prediction accuracy of the GM(1,1).(2)This paper proposes an UAV energy consumption prediction model based on an improved grey prediction model.Existing UAV path planning evaluation models mostly consider simple constraints and overlook the important role of energy consumption in UAV flight,leading to inaccuracies and instability in path planning results.To address these issues,we establish an energy consumption prediction model based on actual flight data,which considers the relationship between the flying distance,angle,and energy consumption of the UAV.Additionally,a comprehensive evaluation function is established that considers multiple constraints and threats,providing optimized paths for the UAV.This energy consumption model serves as the basis for subsequent path planning for rotary-wing UAV.(3)This paper presents a two-dimensional terrain unmanned aerial vehicle(UAV)path planning method that combines the improved Sparrow search algorithm and the energy prediction model.Most existing algorithms cannot meet the high requirements for accuracy and stability of UAV path planning.To address this issue,this study applies our proposed improved algorithm and energy prediction model to two-dimensional path planning.Simulation results show that the algorithm can comprehensively consider energy consumption,path,accuracy,and other requirements to find an effective path in a complex environment.This method provides a more accurate and efficient solution for UAV path planning.(4)This paper proposes a three-dimensional UAV path planning method that combines an improved Sparrow search algorithm and an energy consumption prediction model.Compared to two-dimensional environments,three-dimensional environments have more complex constraints and higher requirements for planning efficiency.An improved Sparrow search algorithm(S-SSAE)is proposed in this article,which uses a Sinusoidal chaotic mapping to achieve uniform distribution of the population,an updated formula based on the t-distribution to improve the Sparrow search algorithm,and an elite reverse strategy to further update the population’s positions.These three methods together solve the problems of slow convergence speed and easy local optimization that the original algorithm faces.The proposed algorithm improves the efficiency and effectiveness of UAV three-dimensional path planning.更多还原