基于改进烟花算法的资源分配

Resource allocation based on improved fireworks algorithm

资源分配问题作为一个NP-Hard问题,在云计算、无线电、卫星调度、多无人机协同作业等领域皆有研究需求,是一个共性的数学问题。烟花算法作为一种智能优化算法,具有求解大规模资源分配问题的能力,但也存在求解精度低等问题。为了提高传统烟花算法的计算效率和全局寻优能力,提出一种改进烟花算法,用遗传算法中的变异算子替代高斯变异操作,并增加模拟退火流程。最后在多无人机协同作业任务分配数学模型上进行仿真验证,实验结果表明在收敛速度以及计算精度方面,该算法均优于其余3种烟花算法。

As a NP-hard problem, the resource allocation problem is a common mathematical problem in cloud computing, radio, satellite scheduling, multi-UAV collaborative task allocation, and many other fields. As an intelligent optimization algorithm, the fireworks algorithm has the ability to solve large-scale resource allocation problems, but also has some problems, such as low solving precision and probability to fall into locally optimal solution. To improve the computational efficiency and global optimization ability of the traditional fireworks algorithm, this paper proposes an improved fireworks algorithm, which uses the mutation operator in genetic algorithm to replace Gaussian mutation operation in the traditional fireworks algorithm and adds the simulated annealing process in each iteration. Finally, the performance of the algorithm is verified by simulation on a mathematical model for multi-UAV cooperative task assignment, and the results show that the algorithm is superior to other three fireworks algorithms in terms of convergence speed and solving precision.