Operating Optimization of Multi-energy System From Perspective of accuracy and real-time Performance

Aiming at the research of optimization of multiple energy system operation, an linear matrix modeling method was proposed. In view of system engineering, to solve nonlinear programming problem due to the involvement of scheduling coefficients, the energy flows were taken as a state variable between energy conversion devices. And a linear programming model for multiple energy system was established, and two kinds of objective functions, which are point-by-point optimization strategy and full-point optimization strategy respectively, are presented in the paper. The application and analysis for a typical regional triple supply system of electricity, cold and heat illustrated that, this linear programming matrix model is simple and efficient, and has better real-time performance compared with the existing nonlinear programming model while ensuring the calculating accuracy. The point-by-point optimization strategy has the same accuracy as the full-point operation optimization strategy, but the computational real-time performance is much better than the full-point optimization strategy. Moreover, the sensitivity analysis of energy prices and component efficiency at partial load on system operation costs can be carried out based on the proposed method. After the optimal operation strategy produced by the linear programming matrix model is trained by the general regression neural network (GRNN), the average computational time for one day can be further reduced to 0.0057 seconds. The proposed method has great application potential for operating optimization of multienergy system from perspective of accuracy and real-time performance.