中文摘要：

为解决配网系统中因三相不平衡造成的三相电压不对称、配电变压器出力降低、网络损耗增加等一系列问题，通过综合考虑配网系统中各节点的负荷曲线特征，建立长效三相平衡优化换相模型，弥补了常规基于单个时间点负荷换相模型需要频繁换相的不足；然后，提出一种基于分子动力学原理的新型模拟结晶算法求解模型，以在进化后期获得更高的个体多样性，从而可有效避免陷入早熟，快速求得全局最优换相方案：最后，通过对IEEE34节点配电系统进行三相平衡优化换相计算，验证了所提方法的有效性及优越性。

英文摘要：

In order to solve a series of problems which are caused by the three-phase unbalancing in the distribution networks, such as asymmetry in three-phase voltage, reduction in distribution transformer output and the increase in network losses, a new optimal phase swapping model for long-term three-phase balancing of distribution networks was built in this paper by considering load curve characteristics of each bus. The model can cover the shortage of conventional models based on the node loads of distribution networks at a particular time that frequent operations for phase swapping are required. Meanwhile, based on the principle of molecular dynamics, a new simulated crystallizing algorithm （SCA） was proposed. SCA can keep higher diversity of individuals in the later evolution process to prevent premature convergence effectively and obtain global optimal phase swapping scheme rapidly. Finally, optimal phase swapping calculations for three-phase balancing with IEEE 34-bus test feeders were carried out to verify the validity and superiority of the proposed method.