Summary
A new evolutionary technique named modified cultural algorithm (MCA) has been implemented for the reduction of real power loss by optimal capacitor allocation. This evolutionary algorithm has provided promising results for 9‐bus, 34‐bus, and 69‐bus standard distribution systems dealing with the real power loss. It has been observed that situation and history knowledge influence the population space with the updating of belief space. Comparative study has been done to establish the effectiveness of MCA with other methods such as cultural algorithm, heuristic, fuzzy‐logic, and hybrid methods. The real power loss has been observed to be reduced from 783.8 to 677.54 kW and from 221.5 to 160.9 kW for 9‐bus and 34‐bus distribution systems, respectively. It is also shown that MCA has outperformed cultural algorithm as well as real‐coded genetic algorithm and particle swarm optimization for loss as well as cost reduction in 69‐bus distribution system. Copyright © 2013 John Wiley & Sons, Ltd.