Normalized Normal Constraint Algorithm Applied to Multi-objective Security-constrained Optimal Generation Dispatch of Large-scale Power Systems with Wind Farms and Pumped-storage Hydroelectric Stations

What is it about?

A multi-objective security-constrained optimal generation dispatch (MOSCOGD) model for large-scale power systems with wind farms and pumped-storage hydroelectric (PSH) stations is proposed. In this model, fuel consumption, emission content of atmospheric pollutants, and power purchase costs are taken into account as objective functions, and network security under basic status and N − 1 criterion conditions, and the operational limits of PSH resources are included in the constraints. The difficulties associated with discrete variables for describing the operational characteristics of PSH resources are avoided by introducing two continuous variables, generation power and pumping power, which satisfy complementary constraints. The normalised normal constraint algorithm is used to transform the three-objective optimal dispatch model into a series of single-objective optimisation models, which are solved by the interior point method, resulting in a series of evenly distributed Pareto optimal solutions (POSs) and the complete Pareto frontier surface. The computational efficiency is greatly enhanced by employing a checking–adding–checking again-adding again scheme to address network security constraints. Moreover, parallel computing is employed to improve the computation speed for solving the POSs of the MOSCOGD model. Test results on an actual large-scale power system and the modified IEEE 39-bus system demonstrate the effectiveness of the proposed method.

Featured Image

Why is it important?

A multi-objective security-constrained optimal generation dispatch (MOSCOGD) model for large-scale power systems with wind farms and pumped-storage hydroelectric (PSH) stations is proposed. The normalised normal constraint algorithm is used to transform the three-objective optimal dispatch model into a series of single-objective optimisation models, which are solved by the interior point method, resulting in a series of evenly distributed Pareto optimal solutions (POSs) and the complete Pareto frontier surface.

Perspectives