Research On Path Planning For Data Collection By Using Uavs

Path planning is one of the important issues in robotics.Different kind of robots and tasks lead to lots of different path planning problems.With the development of Wireless Sensor Networks(WSNs)and UAV,more and more applications combine with the two ones is presented in all walks of life such as precision agriculture,rescue and search,wildfire management.Lots of research problem is followed after it.In the traditional WSN,the way of data collection easily caused the problem of energy hole,as the sensor nodes near the sink nodes cost too much energy on forwarding sensed data to the sink node.With the development of correlated technologies,UAV becomes smaller,faster,more intelligent and flexible,the energy hole problem can be solved efficiently by applying it to collect data.Due to the constraints of battery technology at present,the energy available to execute tasks for UAV is greatly limited.In order to make UAV complete the task more quickly and efficiently,its path must be routed to reduce its energy consumption on flight and data collection.This topic mainly solves the path planning problem for data collection using UAVs,namely,the path planning of sensor nodes coverage.In terms of the problem,we build the correlated model contains system model,environment model,energy consumption model of UAV.By using the grid method,the monitor area where the sensor nodes are distributed is discretized into several grids.Each grid on each height represents a hover point.The UAV will select several ones from these hover points as data collection points to complete the task of data collection.How to determine the number,location and traversal order of these data collection points,and makes the flight path composed with these points cover all the sensor nodes and minimize the energy consumption of UAV,is the focus of this problem.In terms of the problem,we divided the problem into two sub-problems to solve: 1)Calculate the optimal subset of data collection areas;2)Calculate the optimal flight path based on the set calculated in 1).For the first subproblem,we solved it by the greedy algorithm;for the second subproblem,it was solved by the improved Shuffled Frog-Leaping Algorithm.Based on the data of UAV tested in real environment,the experiments are conducted from two views on the recent algorithms such as VD-PSO,TSP-GA and so on.In terms of the data acquisition about UAV,we employed the SDK provided by DJI company to make the UAV fly automatically and the test data is collected based on it.In terms of the simulation experiment results,it shows that the proposed algorithms have good performance on the energy consumption in the case of less data load and more sensor nodes.