【摘要】 针对现有智能优化算法在求解电网故障诊断解析模型时存在的易于陷入局部最优和种群质量低等问题，提出一种改进二进制增益共享知识算法(improved binary gaining-sharing knowledge-based algorithm, IBGSK)。首先，根据故障诊断规则，构建一种包含完备故障信息的完全解析模型。其次，将离散工作机制融入改进算法的种群更迭中，以避免发生空间脱节。然后，结合进化种群动力学思想(evolutionary populationdynamics, EPD)，引入一种自适应交叉算子，以提高种群质量和增强算法的全局寻优能力。最后，通过特征选择和故障诊断仿真实验对算法性能进行评估。结果表明：IBGSK算法相较于其他优化算法，在特征选择问题上具有更高的计算效率、更强的全局寻优能力和泛化能力；在求解电网故障诊断解析模型上具有更优的诊断可靠性、时效性和收敛性。更多还原

【Abstract】 There is a problem that existing intelligent optimization algorithms tend to fall into local optima and low population quality when solving power grid fault diagnosis analytical models. Thus an improved binary gain-sharing knowledge-based algorithm(IBGSK) is proposed. First, a complete analytical model considering complete fault information is constructed from the fault diagnosis rules. Second, a discrete working mechanism is integrated into the population replacement of the improved algorithm to avoid spatial disconnection. Then, combined with the idea of evolutionary population dynamics(EPD), an adaptive crossover operator is proposed to improve the population quality, thereby enhancing the global optimization ability of the improved algorithm. Finally, the performance of algorithms is evaluated by feature selection and fault diagnosis simulation experiments. The results show that the IBGSK algorithm has higher computational efficiency, stronger global optimization ability, and generalizability in feature selection problems than other optimization algorithms. It has better diagnostic reliability, timeliness, and convergence in solving the analytic model of power grid fault diagnosis.更多还原