融合黄金正弦和曲线自适应的多策略麻雀搜索算法

针对元启发算法中麻雀搜索算法(SSA)的早熟收敛、易陷入局部最优、全局搜索性差等问题进行研究，提出一种融合黄金正弦和曲线自适应的多策略麻雀搜索算法。首先，利用Chebyshev混沌映射初始化种群，使初始解位置分布更为均匀，产生优质初始解，增加种群丰富性；其次，引入黄金正弦和曲线自适应权重改进发现者和加入者位置更新方式，有效协调了全局搜索与局部挖掘能力，加快收敛速度；最后，动态选择随机游走或柯西-t扰动策略对最优麻雀位置进行扰动，提高算法跳出局部最优的能力以及收敛精度。选取14个基准函数进行测试，比较改进算法与其他九个元启发式算法的仿真结果，使用Wilcoxon秩和检验以及MAE(mean absolute error)排序来验证所提改进策略的有效性。结果表明，该算法在全局搜索性、克服局部最优、收敛速度、收敛精度、稳定性都有较大提升。

Multi-strategy sparrow search algorithm integrating golden sine and curve adaptive

To solve the shortcoming of the meta-heuristic sparrow search algorithm, such as early convergence, easy to fall into local optimum, poor global searchability, this paper proposed a multi-strategy sparrow search algorithm integrating golden sine and curve adaptative. Firstly, it used Chebyshev chaotic mapping to initialize the population, so that the initial solution position distribution was more homogeneous, produced high-quality initial solutions, and increased the richness of the population. Secondly, it introduced golden sine and curve adaptive weight to improve the location update method of the discoverer and the joiner, which effectively coordinated global search and local mining capabilities and accelerated the convergence speed. Finally, it selected the random walk or Cauchy-t disturbance strategy disturbed the optimal sparrow position dynamically, which improved the ability of the algorithm to jump out of the local optimum and improved the convergence accuracy. This paper selected 14 benchmark functions for testing, compared the simulation results of the proposed algorithm with other 9 meta-heuristic algorithms, and used Wilcoxon rank-sum test and MAE ranking to verify the effectiveness of the proposed improvement strategy. The results prove that the improved algorithm has a great improvement in global searchability, overcoming local optima, convergence speed, convergence accuracy, and stability.