Research On Energy Optimization Scheduling Strategy Based On Swarm Intelligence Algorithm

The emergence of artificial intelligence has promoted the development of energy information networks and energy flow planning.The traditional manual regulation of heating equipment operation mode no longer meets the operational requirements of today’s fine adjustment.By dividing the operational strategy more precisely,it can meet the needs of power grid peaking,and at the same time improve the economics of operation.This paper proposes a method for formulating the optimal operation strategy of integrated energy system based on load mode.Taking a heating plant in a university as an example,the performance gap of the new and more classical multi-objective intelligent optimization algorithm in the optimization of energy system operation is analyzed in the paper,and selects the NSGAII algorithm with better performance to calculate the system optimization strategy.Then the clustering obtains the typical load mode of the building in the plant.The K-means clustering algorithm is combined with the index analysis to establish the typical class.Then the load is classified into the typical class by the SVM classifier to obtain the typical user heating load mode.Finally,the load based on the cluster is obtained.The model uses the NSGAII algorithm to optimize the equipment power-heat model based on the gray box method with the lowest operating cost and the lowest operating energy consumption.Finally,the Tossian gray correlation degree entropy weight method is used to make multiobjective decision making,and the optimal operating strategy point on the Pareto front is obtained to realize the planning of the unit operation strategy.The optimal operation strategy under different load modes is obtained,and the heating load under different load change modes is analyzed.The working mode of the unit based on different level of load mode is also obtained.As a verification of the optimized operation strategy,the research compares the operational strategy obtained by the test with the optimized strategy.It is found that the pattern recognition method is used to classify the building load data,and the accuracy of the classifier can reach 98.7%.In addition,the final operation strategy obtained by multi-objective optimization can save 15.58% of the running cost and save the operating energy consumption by 14.75% compared with the actual operation.Through the optimization results,it can be found that the energy system of the building can be finely adjusted in a planned manner,which can achieve a balance between operating costs and system power consumption.The energy system with the hot water storage tank will have greater adjustment flexibility.