Research And Simulation Implementation Of Multi-UAV Cooperative Path Planning

Unmanned aerial vehicle(UAV) path planning is the core of multi-UAV cooperative control, with the increase of popularity rate of UAVs, the people whose living environment is more complex contact UAV more frequently, which also gives a challenge to overall control of the UAV, and makes single UAV, even multiple UAVs cooperative path planning research has become to be a hot issue in recent years the same time, the communication of UAVs in three-dimensional space and the complexity of multiple-degree-of-freedom navigation make more challenge to Multiple UAVs cooperative path planning. Multiple UAVs cooperative path planning refers to the coordination to complete the task on the basis of a single drone multi-objective programming method, it is particularly prominent when set of tasks is huge and there is an order in finishing the tasks. Common multi-UAV path planning focus on building environment and the shortest path optimization, or focus on solving the performance problems of the shortest path algorithm, or focus on the design of collaborative rules. In this thesis, optimization and multi-aircraft cooperative aspects of the existing methods improved, the model makes improvements in environment establishment, shortest path algorithm optimization and multi-aircraft coordination compared with the existing methods, and analyze the results of simulation experiments to draw relevant conclusions. In this thesis, the main contents and innovative points are as follows:1.This thesis designs some The threat distribution model of several threat sources and defines the concept of Comprehensive threat field based on two dimentional Voronoi graph according to these models. This thesis expands the traditional single threat source model, improves some of the threats and distribution model, and adjusts constantly threat expression according to common sense to obtain a Voronoi environmental threats field which is close to reality and has high calculating speed.2.This thesis explores the Single UAV path planning with ant colony algorithm in the Voronoi graph with comprehensive threat field, and designs an improved ant colony algorithm, which made two main improvements based on classical ant colony algorithm. In order to accelerates the algorithm convergence, this algorithm puts forward to a way to update pheromone according to the best path in fixed number of iterations. To make the performance of the algorithm faster, this algorithm puts a heuristic factor information in the target point, this information is helpful for ants to shorten the time in finding the target point. In the unified environment, the experiment compares the performance of the algorithms and the solutions of improved ant colony and classical shortest path algorithm.3. This thesis uses K-trajectory smoothing algorithm to smooth the initial path, and presents a collaborative rule based on a range of distance according to the characteristics of a smooth trajectory. This rule defines when and how the UAVs will cooperate with each other. As the core content of this thesis, and based on the above points, we establishs a set of "Environmental Modeling-initial path planning-smoothing the path-collaborative planning" process, and achieves a simulation experiment.