Research On UAV Maritime Search And Rescue Path Planning

The shipping is the carrier and foundation for global economy.With the development of science and technology,the navigation status of ships has been greatly improved.However,the navigation safety of ships is still subject to many factors,and ship accidents are difficult to avoid.Maritime search and rescue(MSAR)is an important international obligation undertaken by China according to "International Convention on Maritime Search and Rescue".Maritime search is the important part of providing maritime rescue,as well as the most expensive,risky aspect of the MSAR process.As an emerging equipment,the UAV has many advantages,the cooperation mode of multi-UAV can effectively improve search efficiency.This paper takes the UAV as the maritime search unit,and researches on search and rescue path planning of multi-UAV.The main work is as follows:(1)With the research objective of determine the best search area at sea.Based on the analysis of the forces on the drift targets and leeway model,according to the general form of the Markov process,a stochastic particle ensemble drift model is proposed based on Monte Carlo approach with the perturbations are added to leeway coefficients,wind field and flow field.The crosswind leeway jibing mechanism is added to the simulation process.The numerical computation method of the ensemble drift model is presented,and a search area prediction simulation example is given to validate the effectiveness of the ensemble drift model.(2)With the research objective of the environmental entropy value and the cumulative probability of success,the UAV detection probability model is analyzed.Based on the diversity properties of drift targets,a search area is established for UAV.A multi-UAV distributed model predictive control(DMPC)method is presented based on model predictive control,combined with the UAV status model.Constructing a calculation model under the consideration of UAV's communication and search profits.The simulation results show the rationality and validity of the model.