Abstract: To address the issue of optimal allocation of distributed generation and shunt capacitors in distribution networks, a multi-objective siting and sizing model is established to minimize network loss and optimize cost. To effectively solve this multi-objective model, a multi-objective dual-guided lens opposition-based learning differential evolution algorithm (MODGLDE) is proposed. MODGLDE employs lens imaging to conduct reverse learning on the initialized population. Meanwhile, it integrates the bootstrapping strategy based on convergence and diversity indicators into the mutation strategy, and balances the global search and local exploitation capabilities of the algorithm by using the preferences for pre-evolution and post-evolution. In this manner, it can effectively address the issues of uneven distribution in randomly initialized populations, high randomness in mutation, and slow convergence speed in the differential evolution algorithm. The simulation results of the IEEE 33-bus system indicate that the MODGLDE algorithm not only reduces active power loss and improves network voltage distribution, but also exhibits good stability and effectiveness.