At this stage, when studying the economic and environmental load distribution of the power grid, the conversion of the multi-objective optimization problem into a single-objective optimization problem generally requires traversing the reorganization of the weights for comparison. Moreover, the obtained weight combination has a low precision, and there is still a certain gap between the optimal weight combination. In view of this situation, an adaptive multi-objective weighted particle swarm optimization (AMWPSO) algorithm is proposed that uses the number of generator sets and the number of optimization targets to determine the particle dimension, and uses the data of the IEEE 30 bus system including 6 units to account for and ignore the network. Solve the load distribution model under two conditions. The solution results show that the proposed optimization algorithm can automatically generate the weight ratio of each target on the one hand, avoiding the tedious weight comparison calculation. On the other hand, compared with the combination of manually set neighboring weights, the calculated target weights have high accuracy and perform better under the two comprehensive targets of power generation cost and pollutant emission.