Application of effective gravitational search algorithm with constraint priority and expert experience in optimal allocation problems of distribution network

Abstract

Optimal allocation problem of distribution network (OAPDN), which attracts much attention of electric enterprise, contributes to the flexible and environmental-friendly power supply by rationally introducing distributed generations (DGs) and shunt capacitors (SCs). To smoothly solve OAPDN problems, several effective measures such as advantageous schemes guidance (ASG) mechanism are presented and integrated into the proposed modified gravitational search algorithm with expert experience (MGSA-EE). Multiple OAPDN experiments essentially indicate that the MGSA-EE with higher efficiency and stronger exploration capability reduces the power loss on 33, 69 and 119 node networks by 94.15%, 98.10% and 84.59%, which is superior to most published technologies.Furthermore, multi-objective OAPDN problem which simultaneously considers two or more goals is also studied. Compared with the single-objective one, it is more in line with the diverse demands of actual electricity market, but the difficulty is greatly increased as well. On this basis, this paper extends MGSA-EE to an innovative multi-objective MGSA-EE (MMGSA-EE) algorithm by the suggested non-inferior sorting strategy with constraints-prior.Most typically in multi-objective OAPDN experiments on large scale networks, MMGSA-EE achieves high-quality scheme that concurrently reduces power loss and voltage deviation of 69-node network by 94.07% and 98.69%. Meanwhile, several quantitative indicators also prove that the proposed MMGSA-EE has competitive advantages over the original algorithm in terms of Pareto fronts, node voltage profiles and execution time.In general, MGSA-EE and MMGSA-EE algorithms provide efficient tools for exploring superior DG/SC configuration schemes, and are of great significance to fill research gaps in multi-objective optimizations of distribution network.