Multi-UAV Task Assignment Benchmark

多无人机任务分配算法测试基准

Introduction

A benchmark for multi-UAV task assignment is presented in order to evaluate different algorithms. An extended Team Orienteering Problem is modeled for a kind of multi-UAV task assignment problem. Three intelligent algorithms, i.e., Genetic Algorithm, Ant Colony Optimization and Particle Swarm Optimization are implemented to solve the problem. A series of experiments with different settings are conducted to evaluate three algorithms. The modeled problem and the evaluation results constitute a benchmark, which can be used to evaluate other algorithms used for multi-UAV task assignment problems.

Notice that three algorithms run at three CPU cores respectively, which means that there is no parallel optimization in this benchmark.

Please refer to the paper to see more detail.

K. Xiao, J. Lu, Y. Nie, L. Ma, X. Wang and G. Wang, "A Benchmark for Multi-UAV Task Assignment of an Extended Team Orienteering Problem," 2022 China Automation Congress (CAC), Xiamen, China, 2022, pp. 6966-6970, doi: 10.1109/CAC57257.2022.10054991.

ArXiv preprint arXiv:2003.09700

Usage

1. Algorithm input and output

Algorithm input includes vehicle number (scalar), speeds of vehicles ($n\times1$ array), target number (scalar $n$), targets ($(n+1)\times4$ array, the first line is depot, the first column is x position, the second column is y position, the third column is reward and the forth column is time consumption to finish the mission), time limit (scalar). The code below is the initialization of the class GA in ga.py.

def __init__(self, vehicle_num, vehicles_speed, target_num, targets, time_lim)

There should be a function called run() in the algorithm class, and the function should return task assignment plan(array, e.g. [[28, 19, 11], [25, 22, 7, 16, 17, 23], [21, 26, 12, 9, 6, 3], [5, 15, 1], [18, 20, 29]], each subset is a vehicle path) and computational time usage (scalar).

2. Evaluate

You can replace one algorithm below with another algorithm in evaluate.py, and then python evaluate.py. If you don't want to evaluate three algorithm together, you should modify the code properly( this is easy).

ga = GA(vehicle_num,env.vehicles_speed,target_num,env.targets,env.time_lim)
aco = ACO(vehicle_num,target_num,env.vehicles_speed,env.targets,env.time_lim)
pso = PSO(vehicle_num,target_num ,env.targets,env.vehicles_speed,env.time_lim)
ga_result=p.apply_async(ga.run)
aco_result=p.apply_async(aco.run)
pso_result=p.apply_async(pso.run)
p.close()
p.join()
ga_task_assignmet = ga_result.get()[0]
env.run(ga_task_assignmet,'GA',i+1,j+1)
re_ga[i].append((env.total_reward,ga_result.get()[1]))
env.reset()
aco_task_assignmet = aco_result.get()[0]
env.run(aco_task_assignmet,'ACO',i+1,j+1)
re_aco[i].append((env.total_reward,aco_result.get()[1]))
env.reset()
pso_task_assignmet = pso_result.get()[0]
env.run(pso_task_assignmet,'PSO',i+1,j+1)
re_pso[i].append((env.total_reward,pso_result.get()[1]))

3. About reinforcement learning

In Env() in evaluate.py, function step is used for reinforcement learning. Because this is still being developed, we cannot supply a demo. If your algorithm is reinforcement learning, you can try to train it with Env(). Your pull request and issue are welcome.

Enhancement

This repository does great enhancement and you can use it for high performance. Thanks to dietmarwo for the nice work.

GA uses numba for a dramatic speedup. Parameters are adapted so that the execution time remains the same: popsize 50 -> 300, iterations 500 -> 6000 For this reason GA now performs much better compared to the original version.
Experiments are configured so that wall time for small size is balanced. This means: increased effort for GA, decreased effort for ACO. For medium / large problem size you see which algorithms scale badly: Increase execution time superlinear in relation to the problem size. Avoid these for large problems.
Adds a standard continuous optimization algorithm: BiteOpt from Aleksey Vaneev - using the same fitness function as GA.py. BiteOpt is the only algorithm included which works well with a large problem size. It is by far the simplest implementation, only the fitness function needs to be coded, since we can apply a continuous optimization library fcmaes. Execute "pip install fcmaes" to use it.
Uses NestablePool to enable BiteOpt multiprocessing: Many BiteOpt optimization runs are performed in parallel and the best result is returned. Set workers=1 if you want to test BiteOpt single threaded.
All results are created using an AMD 5950x 16 core processor utilizing all cores: 29 parallel BiteOpt threads, the other 3 algorithms remain single threaded.
Added test_bite.py where you can monitor the progress of BiteOpt applied to the problem.
Added test_mode.py where you can monitor the progress of fcmaes-MODE applied to the problem and compare it to BiteOpt for the same instance. fcmaes-MODE is a multi-objective optimizer applied to a multi-objective variant of the problem. Objectives are: reward (to be maximized), maximal time (to be minimized), energy (to be minimized). The maximal time constraint from the single objective case is still valid. Energy consumption is approximated by sum(dt*v*v)

robin-shaun / multi-uav-task-assignment-benchmark Goto Github PK