Planning of Multi Energy-Type Micro Energy Grid Based on Improved Kriging Model

The increasing complexities of multi energy-type micro energy grid (MEG) integrated with distributed renewable energy resources require more effective planning method. This paper presents an improved Kriging model for the planning of MEG to satisfy user's demands in cooling, heating, and electrical energy. First, a generic MEG model containing energy supply devices (combined cooling, heating, and power system, and energy storage systems) and energy supply networks is established. Second, the improved Kriging model combined with the Latin hypercube sampling method is proposed for searching the MEG optimal configuration to minimize the total annual cost. Third, for the sake of completeness and practicality, the sample points are updated by a novel mixed infill-sampling criterion comprised of minimum surrogate-model point criterion, trust region criterion, and mean square error criterion. The optimal configuration and operation schemes are obtained simultaneously in the case study. Eventually, the numerical results indicate that the proposed method could efficiently solve the optimal planning problem in contradistinction to three other scenarios regarding the Kriging model.