将多目标优化问题与粒子群算法相结合，提出了一种用于求解2个目标函数的优化算法，即双目标函数粒子群算法(two objective particle swarm optimization，TOPSO)。直流输电单极运行时产生的地电流会导致接地极附近交流变压器铁心磁饱和，影响变压器的安全稳定运行。变压器中性点经小电阻接地抑制直流偏磁时，会使邻近站的变压器中性点直流增大。基于该算法，对目标电网的变压器接入电阻阻值进行网络优化配置；避免某些变电站接地极电流过大或者接入电阻阻值过大。以泸州电网为例，并与NSGA-II算法的计算结果相对比，验证了该方法的准确性和可行性。

Monopole operation of HVDC transmission system can bring on transformer DC bias. This problem would be more serious, after commission and operation of UHVDC transmission system in China. Transformer grounding via resistance is an efficient way to limit DC bias current. But that could cause DC bias of other neighbouring transformers. This paper presented an improved Particle Swarm Optimization algorithm for two objective optimization problems (TOPSO). Based on this method, a group of optimal resistance value could be found to inhibit severe DC bias of all object power grid transformers. And those resistance values could be as small as possible. To validate the approach, with TOPSO and NSGA-II the network configuration of small resistances of all Luzhou power system transformers of China were calculated. The result shows that TOPSO is efficient and feasible for two objective optimization problems.

国家自然科学基金联合基金(10476022)。