An innovative multi-objective scheduling for virtual power plants employing a robust decision-making tactic

Addressing the issue of output uncertainty in renewable energy generation units within the context of environmental and economic dispatch problems for virtual power plants, this paper proposes a robust optimization strategy based on uncertainty boundary decision-making and develops a multi-objective self-adaptive evolutionary algorithm for solving the problem. Initially, the concept of robust cost is introduced to establish an environmental and economic multi-objective optimization dispatch model for a virtual power plant that includes photovoltaic stations and wind turbine units. Subsequently, mathematical theories such as the Lagrangian transformation are employed to eliminate uncertain variables in the dispatch model, resulting in a deterministic multi-objective optimization model. Building upon this, a robust optimization method based on uncertainty boundary decision-making is derived, taking into account the practical needs of dispatch scenarios. Meanwhile, to meet the comprehensive requirements for solution convergence and distribution in multi-objective optimization problems, a multi-objective self-adaptive evolutionary algorithm is obtained by introducing a full-genome-based recombination mechanism and a mutation operator that incorporates adaptive learning. Finally, the rationality of the proposed dispatch strategy and the applicability and superiority of the proposed algorithm are verified through specific case studies.