Probabilistic reliability optimization using hybrid genetic algorithms

In this paper transmission power system structure optimization is performed via a minimal spanning tree based encoded fuzzy logic self-controlled hybrid genetic algorithm (GA). During the redundancy optimization of the power system network a binary encoded GA is used for a modified transmission network expansion problem, finding the optimal power line type with respect to the net present value (NPV) of minimal investment cost, operating costs and load flow constraints. Each individual is evaluated by a minimal state probability reliability estimation algorithm verifying a certain minimal reliability constraint. A developed improvement algorithm is used for individuals not satisfying a reliability constraint. A recently developed fast reliability calculation algorithm, computing energy not supplied, and the obtained NPV of the transmission network expansion problem are utilized as minimization function. The algorithm is applied to a real world sub transmission system in order to discuss strategies for future system expansions.