Bacterial Foraging Algorithm for Optimal Power Flow in Dynamic Environments

Optimal power flow (OPF) problem has already been attempted as a static optimization problem, by adopting conventional gradient-based methods and more recently, nonconventional ones, such as evolutionary algorithms. However, as the loads, generation capacities and network connections in a power system are always in a changing status, these static-oriented methods are of limited use for this issue. This paper presents a new algorithm, dynamic bacterial foraging algorithm (DBFA), for solving an OPF problem in a dynamic environment in which system loads are changing. DBFA is based on the recently proposed BFA which mimics the basic foraging behavior of E. coli bacteria. A selection scheme for bacteria's reproduction is employed in DBFA, which explores the self-adaptability of each bacterium in the group searching activities. DBFA has been evaluated, for optimizing the power system fuel cost with the OPF embedded, on the standard IEEE 30-bus and 118-bus test systems, respectively, with a range of load changes which occurred in different probabilities. The simulation results show that DBFA can more rapidly adapt to load changes, and more closely trace the global optimum of the system fuel cost, in comparison with BFA and particle swarm optimizer.