UAVs Path Planning Based On Safety Risk Assessment

With the continuous progress of science and technology,the development of the UAV industry has also entered a stage of explosion,and the domestic UAV market has also been opened.During the epidemic period,UAV has also made great contributions in the transportation of goods.UAV is more and more closely related to our life,and it has become the focus of our attention to ensure the safe and fast arrival of UAV in the process of transportation flight.In this paper,the path planning of the UAV is carried out from two aspects,namely the threat to pedestrian safety after the UAV crash and its own safety in the transportation process,including three main tasks.First,a method based on pedestrian safety is proposed.The optimized method of risk assessment is used to weigh the risk of ground personnel and flight time.Above all,predict the flight dynamics model of the crash location of the UAV that loses power at a given altitude and speed,generate the probability distribution of the crash location through Monte Carlo simulation,and conduct the safety risk assessment by combining the crash location distribution and the population density around the flight path;Then through different improvements to the network optimization solution,evaluate and compare a set of quality metrics of multiple sets of dual-objective solutions,and determine the appropriate improvement algorithm;finally plan the path of the case according to the algorithm,and find a time and risk minimized Flight path.Second,optimize the existing UAV trajectory,set a large number of correction points and use the improved NSGA-Ⅱ(Non-dominated Sorting Genetic Algorithm-Ⅱ)to continuously iteratively optimize the UAV’s trajectory,taking into account the UAV’s own maneuvering restrictions,and the current planning The curvature radius of the path is estimated.If the curvature radius of the current node is too small,the constraint condition is not satisfied,and then the iterative optimization flight trajectory is performed,so as to obtain an optimized trajectory with as few correction times and distance as possible.Third,for the risks caused by the influence of the external wind field and the lateral deviation of the UAV’s own attitude in the process of transportation flight,the path data can be calculated by GPS and multiple UAV airborne receivers,providing a data basis for the optimization generation of the transportation path.The flight attitude value is obtained by using the inclination Angle signal and UAV’s top view Angle.Meanwhile,the wind field value of the environment is estimated by airspeed,ground speed and wind speed,and the influence of airframe parameters and environment on the UAV’s flight path is fully considered.Since the exponential function is monotonically decreasing,the least dangerous path can be equivalent to the shortest path generation problem.Finally,Dijkstra algorithm is used to achieve the optimal generation of transportation path.The simulation experiment results show that the proposed method can adjust the generated results in real time according to the GPS signal,so that the optimal output path meets the safest and shortest transportation requirements,and has excellent robustness and applicable scope.