| 摘 要: 针对传统灰狼优化算法(GWO)求解精度低、收敛速度慢和易陷入局部最优的缺点,提出一种基于混合扰动的非线性灰狼优化算法。首先,该算法采用一种新型非线性控制参数策略和基于余弦变换的惯性权重策略在探索和开发阶段取得平衡;其次,为了增加种群在算法后期的多样性,采用t-分布概率扰动策略;最后,对灰狼最优个体进行扰动,增强算法跳出局部最优的能力。为了验证改进GWO算法的性能,在10个基准函数上对其进行测试,通过与4种基本智能算法、4种单一改进策略的GWO算法及4种其他改进算法进行比较,结果表明:改进GWO算法在90%的测试函数上优于4种基本智能算法,在50%的测试函数上优于4种单一改进策略的GWO算法,在50%的测试函数上优于4种其他改进算法。 |
| 关键词: 灰狼优化算法;控制参数;t-分布 |
|
中图分类号: TP301.6
文献标识码: A
|
| 基金项目: 国家自然科学基金资助项目(72174121);国家自然科学基金资助项目(71774111). |
|
| Nonlinear Grey Wolf Optimization Algorithm Based on Hybrid Perturbation |
|
LI Zhengnan, QIN Jiangtao
|
(Business School, University of Shanghai for Science & Technology, Shanghai 200093, China)
486453073@qq.com; qinjiangtao_usst@126.com
|
| Abstract: Aiming at the shortcomings of traditional Grey Wolf Optimization algorithm (GWO), such as low solution precision, slow convergence speed and being easy to fall into local optimum, this paper proposes a nonlinear Grey Wolf Optimization algorithm based on hybrid disturbance. Firstly, a new nonlinear control parameter strategy and an inertial weight strategy based on cosine transform are used to achieve a balance between exploration and development. Secondly, in order to increase the diversity of the population in the later stage of the algorithm, a t-distribution probability perturbation strategy is used. Finally, the Grey Wolf optimal individual is perturbed to enhance the ability of the algorithm to jump out of the local optimum. In order to verify the performance of the improved GWO algorithm, it is tested on 10 benchmark functions and compared with four basic intelligent algorithms, four single improved strategy GWO algorithms and four other improved algorithms, the results show that the improved GWO algorithm is superior to the four basic intelligent algorithms on 90% of the test functions. It is better than four single improvement strategies on 50% test function and better than four other improved algorithms on 50% test functions. |
| Keywords: Grey Wolf Optimization algorithm; control parameters; t-distribution |
