摘要
由于状态空间模型进化算法(SEA)易受初始种群的影响,精度不高,容易早熟等问题。因此,提出了一种基于反向学习的状态空间模型进化算法(OLSEA)。通过对状态进化矩阵G重新构造实现全局搜索,增强了全局探索和局部搜索能力;算法结合了反向学习策略,提高了算法搜索效率,增强了跳出局部最优的能力;利用8种基准测试函数对算法有效性分析。仿真实验表明,OLSEA在搜索能力,收敛精度和计算结果的稳定性等方面均有大幅提升。
To overcome the shortcomings of evolution algorithm based on sate-space(SEA) such as easy to be affected by the initial population,premature convergence, low accuracy and poor ability to escape from local optimum.Therefore,this paper proposed a state-space model evolution algorithm based on opposite-learning(OLSEA) Sate evolution G was reconstructed to achieve global search,which enhances the global exploration and local search capabilities;The algorithm combines opposite-learning strategy to improve the search efficiency of the algorithm and enhance the ability to jump out of the local optimum;The effectiveness of the algorithm is analyzed through 8 benchmark test functions. Simulation experiments show that OLSEA has greatly improved its search capability,convergence accuracy and stability of calculation results.
出处
《工业控制计算机》
2021年第2期91-93,共3页
Industrial Control Computer
关键词
状态空间模型
进化算法
反向学习
state-space
evolutionary algorithm
opposition-based learning
