Research On Intelligent Algorithm Improvement In UAV Path Planning

As a type of unmanned aerial vehicle(UAV)that can be remotely controlled and navigated,drones have been widely used in the military field due to their good operability and economy.However,as the complexity of UAV missions increases,the demand for their autonomous flying capabilities has also grown.Therefore,this paper focuses on rotary-wing drones and explores the problems of global flight path planning based on an improved sparrow search algorithm in threedimensional environments and local path replanning based on an improved artificial potential field method.The research contents of this article are outlined below:(1)A model for UAV trajectory planning in a three-dimensional environment was established.The issues and planning requirements involved in trajectory planning were analyzed,and spatial and threat models for UAV flight were developed.Corresponding trajectory evaluation functions were designed,taking into account the flight constraints of UAVs.To ensure that the planned trajectory meets the UAV’s flight conditions,3rd-order B-spline curve optimization was utilized to optimize trajectory waypoints,laying the groundwork for trajectory planning research.(2)Improvement of the Sparrow Search Algorithm(SSA)for Global UAV Path Planning.To address the shortcomings of the SSA in terms of weak global convergence ability and low search accuracy,a hybrid optimization strategy for the SSA is proposed.Firstly,the Circle chaotic mapping is used to initialize the sparrow population,increasing the diversity and search capability of the initial population.Secondly,a dynamic weight adjustment is introduced during the finder’s position updating process to balance the algorithm’s global and local optimization capabilities.Thirdly,the Levy flight strategy is incorporated into the scout’s position updating formula to improve the algorithm’s convergence speed in the later iterations.Finally,a simulated annealing strategy is used to update the individual’s optimal position,which helps the algorithm escape from local optima.The enhanced algorithm was evaluated for its efficacy using eight benchmark functions under the fusion mechanism described above.Simulation experiments were conducted by integrating the improved algorithm with UAV trajectory planning,which demonstrated that the enhanced algorithm significantly increased the success rate of UAV missions and satisfied the flight requirements of the UAVs.(3)Improvement of Artificial Potential Field(APF)Method for Local UAV Trajectory Replanning.To address the drawbacks of local minimum values and unreachable objectives in traditional algorithms,an improved strategy is proposed.Firstly,the attractive potential field function is reconstructed.Secondly,a distance adjustment factor between the drone and the target point is added to the repulsive potential field to solve the problem of reaching the target point,and a relative velocity potential field between the drone and obstacles is added to improve the drone’s dynamic obstacle avoidance capabilities.Finally,a direction perturbation strategy is introduced to improve the UAV’s ability to escape from local minima.Under the above integration conditions,simulation experiments are conducted in a three-dimensional environment to validate the improved algorithm,which is then applied to UAV path planning to improve the UAV’s ability to handle both static and dynamic threats.The results show that the improved algorithm enables the UAV to quickly avoid unknown threats and return to the predetermined trajectory,safely and stably reaching the target point.(4)A UAV Path Planning System was designed and developed.An UAV path planning system was built using the APP design module in Matlab,and the system’s design modules and functional requirements were introduced.By setting the required functional parameters for trajectory planning in the system,global UAV trajectory planning and local replanning were achieved,and the planning results were displayed on the interface.Through simulation experiments,the applicability of the interface built in this paper was verified,and the decisionmaking ability of users was improved,thus achieving human-machine interaction.Simulation experiments have proven the feasibility of the entire research scheme,which can provide optimal trajectories for unmanned aerial vehicles(UAVs)that meet global planning and local replanning requirements.This is advantageous for applying the system in actual threedimensional environments,allowing UAVs to avoid static threats and dynamic emergent threats,and safely and stably reach the target point.This expands the application fields of UAVs and has certain research significance and value.