A Fast Computation And Optimization Algorithm For Smart Grid Energy System

Increasing penetration of intermittent and variable renewable energy sources (RESs) has significantly complicated smart grid operations. The uncertain nature of RESs may cause increased operating costs for committing costly reserve units or penalty costs for curtailing load demands. In addition, it is often desired to control and coordinate a battery energy storage system (BESS) in an efficient and economical way, especially for islanded microgrid. To address these issues, an approximate dynamic programming (ADP) approach is proposed to investigate the optimal operation of energy systems in islanded microgrid considering stochastic wind energy and load demands. A battery control strategy is also presented to maintain the battery state of charge in a certain range which will help to increase the battery lifetime in the future. The traditional dynamic programming (DP) approach is also implemented to validate the percentage of optimality of the proposed ADP approach for stochastic case studies. The simulation results show that the proposed ADP approach can obtain competitive percentages of optimality with around 50% less computational time compared to the traditional DP approach. Again, the proposed ADP approach is also validated on a large data sample case and achieved 18.77 times faster response than the traditional DP approach