Optimal Distribution Network Reconfiguration and DG Allocation Considering Time-Sequence Characteristics of Load and Renewable Generation

The increasing integration of photovoltaic (PV) and wind generation makes distribution system operation more variable throughout the day, so a fixed feeder topology may not keep losses low and voltages within limits at all hours. This paper proposes an improved Human Evolutionary Optimization Algorithm (IHEOA) for simultaneous distributed generation (DG) allocation and distribution network reconfiguration. IHEOA couples the population-based search of the original HEOA with a radiality-preserving decoder and a discrete-aware memetic refinement. The model minimizes active power loss, annual cost, and emissions. In a 24 h framework, PV and wind units are sited and sized once, while the radial topology is re-optimized every hour. For loss minimization with three DG units on the IEEE 33-bus system, IHEOA reaches the global optimum, verified by a mixed-integer second-order cone programming model, in 20 of 30 runs. In contrast, HEOA does not reach it even with an equal computational budget. The coordinated 24 h strategy reduces the mean active power loss by 61.6%, 77.4%, and 44.7% on the 33-, 69-, and 118-bus systems with all voltages within limits. A single daily topology retains most of this reduction, and on the 33-bus system, the fixed DG plan remains robust under forecast errors of up to 20%.