情绪认知驱动的社会情感优化算法研究

Research on Social Emotional Optimization Algorithm Driven by Emotional Cognition

摘要: 针对社会情感优化算法仅依靠社会评价值优化策略而导致的动态适应性差、搜索策略过渡生硬等问题，提出一种基于情绪认知驱动的社会情感优化算法(EC-SEOA)。算法引入满足、欢喜、愉悦、平静、焦虑、烦躁、疲劳七级类脑连续情绪机制，通过情绪激励值实现情绪状态自适应判定，并给出标准化阈值与推荐参数取值；构建情感吸引力–排斥力协同交互模型，增强种群多样性；设计包含工作记忆、趋势预判与维度注意力的高级认知决策模块，形成情绪–认知闭环调控机制。在Sphere、Rosenbrock、Rastrigin、Griewank标准测试函数上与多种智能算法对比，结果表明：所提算法收敛更快、寻优精度更高、稳定性更强，在多峰与高维复杂优化问题中优势显著，为复杂工程优化提供一种新型类脑智能优化方法。

Abstract: Aiming at the problems of poor dynamic adaptability and rigid transition of search strategies existing in the social emotional optimization algorithm (SEOA), which only relies on the social evaluation value for optimization, an emotional cognition-driven social emotional optimization algorithm (EC-SEOA) is proposed. The algorithm introduces a seven-level brain-like continuous emotional mechanism, including satisfaction, joy, delight, calmness, anxiety, irritated, and fatigue. It realizes the adaptive judgment of emotional states through emotional incentive values, and provides standardized thresholds and recommended parameter values. A collaborative interaction model of emotional attraction-repulsion is constructed to enhance population diversity. An advanced cognitive decision-making module including working memory, trend prediction, and dimensional attention is designed to form a closed-loop regulation mechanism of emotion-cognition. Comparative experiments with various intelligent algorithms are carried out on the Sphere, Rosenbrock, Rastrigin, and Griewank standard test functions. The results show that the proposed EC-SEOA has faster convergence speed, higher optimization accuracy, and stronger stability, and exhibits significant advantages in multimodal and high-dimensional complex optimization problems. This study provides a new type of brain-like intelligent optimization method for complex engineering optimization.