In the AC-DC interconnection system, each DC transmission channel can produce large power loss while transmitting a large amount of electric energy. By coordinating and optimizing the voltage and reactive power control measures of multi-circuit extra/ ultra HVDC transmission system, the power loss of the whole AC-DC system can be effectively reduced. This paper presents a multi-objective reactive power optimal control method for AC-DC system considering the detailed power loss characteristics of the converter stations. A multi-objective reactive power optimal control model for AC-DC interconnected system is established to minimize the total loss of the whole AC-DC transmission system including the loss of the main equipment in the converter stations and the sum of the squared voltage deviation of all key nodes. An adaptive weighted sum algorithm is presented, which is combined with the GAMS/CONOPT solver to obtain the evenly distributed Pareto optimal solutions set for the multi-objective optimization problem. According to the fuzzy membership and entropy weight of each Pareto optimal solution, the compromised optimal solution is determined from the Pareto optimal frontier curve, which is used as the multi-objective reactive power optimal control scheme. Test results of a 6056 bus AC-DC interconnection system show that Pareto optimal solutions set obtained from the proposed method is evenly distributed, and the obtained multi-objective optimal control scheme can effectively reduce the operational power loss of the converter stations and improve the voltage quality of key nodes. This work is supported by National Natural Science Foundation of China (No. 51207056) and Natural Science Foundation of Guangdong Province (No. 2015A030313233).