针对鹈鹕优化算法收敛速度较慢、初始化过程随机产生初始种群导致种群多样性差，在后期易陷入局部最优等问题，提出了一种融合三维螺旋飞行和混合反向学习策略的鹈鹕优化算法。首先使用Gauss映射初始化种群，提高种群多样性；其次利用三维螺旋飞行和混合最优最差反向学习策略，加强算法跳出局部最优的能力；最后，引入自适应平衡因子与自适应步长，提出鹈鹕坠落策略，以模拟捕食过程中群体的微小变化。最后，通过12个基准函数和实际案例对IPOA（improved pelican optimization algorithm）进行测试，并与8个仿生算法进行对比，测试结果与Wilcoxon符号秩和检验结果均表明IPOA收敛精度与稳定性等各项性能都有所提升，具有明显优势。

The pelican optimization algorithm (POA) has some defects, such as slow convergence speed, poor population diversity, and premature. To overcome these shortcomings, an improved algorithm was propped. The improved algorithm was proposed incorporating 3D spiral flight and a hybrid backward learning strategy. Firstly, the Gauss mapping was used to initialize the population to improve the population diversity. Secondly, the three-dimensional spiral flight and the hybrid optimal-worst backward learning strategy were used to strengthen the ability of the algorithm to jump out of the local optimum. Finally, the adaptive balance factor and adaptive step size were introduced to propose a pelican falling strategy to simulate the small changes. In the Finally, IPOA was tested by 12 benchmark functions and actual cases and compared with 8 bionic algorithms. Test results and Wilcoxon signed rank sum test results showing that IPOA has improved convergence accuracy and stability.