Precision-Controlled Conservative Repair for Loss-Constrained Economic Emission Dispatch

Economic emission dispatch (EED) and combined economic emission dispatch (CEED) are multi-objective power system optimization problems in which Pareto-front quality must be interpreted together with physically feasible power balance. Conventional feasibility repair can keep final populations feasible under a loose tolerance while leaving avoidable loss-coupled balance residuals. This paper presents precision-controlled conservative repair (PCR), a bounded repair layer that repeatedly redistributes the residual in proportion to available generator capacity, recomputes transmission losses after every correction, and stops at a strict tolerance or iteration cap. PCR is evaluated with the non-dominated sorting genetic algorithm II (NSGA-II), the multi-objective evolutionary algorithm based on decomposition (MOEA/D), and multi-objective particle swarm optimization with crowding distance (MOPSO-CD) on the IEEE 30-bus 6-unit EED and SAGE/Basu 10-unit CEED benchmarks. Thirty independent seeds are used per condition. Relative to the matched current-repair baseline, PCR improves median inverted generational distance in all six benchmark–algorithm pairs and median hypervolume in five, while reducing mean absolute balance mismatch from 0.0078–0.075 MW to 0.000019–0.000059 MW. All final populations remain feasible under the 1 MW reporting tolerance. The gains are accompanied by mean runtime increases of 3.1–9.0 s. A cap study shows that 50 strict iterations is nonbinding in the two tested systems, whereas five iterations is frequently insufficient. These controlled algorithmic comparisons support PCR as a reproducible repair refinement with an explicitly measured runtime cost.