Multi-objective Operation Optimization For Off-grid Regional Integrated Energy System

Nowadays,with the rapid economic development,energy shortage,environmental pollution,greenhouse effect and other problems are becoming increasingly prominent.Many countries are intensively developing renewable energy technologies order to achieve the goal of energy conservation and emission reduction.In the electric power industry,integrated energy system has a promising development prospect,with high energy efficiency,low emissions and other advantages,and presents an important solution for the sustainable development of electrie power industry.When large power grid fails,the regional integrated energy system will be switched from grid-connected operation to off-grid operation.However,the randomness and volatility of power generation and power within the regional integrated energy system(RIES)are large,which puts forward higher requirements for its operation strategy.This thesis is focused on temporary off-grid operation of RIES.The goal is to ensure adequate supply of load,with the operation economy and CO2 emissions as the optimization objective.The ofF-grid operation strategy is optimized with dynamic programming algorithm.In order to address the real-time control problem,this thesis proposes the RIES based on stochastic model predictive control and develops a multi-objective optimal operationstrategy.The main results are as follows:(1)Established operating model of off-grid distributed power supply and energy storage model;described in details the constituents of regional integrated energy system;compared the similarities and differences of the control methods of each subsystem under grid-connected operation mode and ofF-grid operation mode;simplified the modeling of each subsystem according to its operation characteristics;and thus established the operation models for later multi-objective optimization of the operation strategy.(2)Proposed an optimal strategy for the off-grid operation of RIES based on dynamic programming algorithm;solved multi-stage decision optimization problem,in which the optimization objectives are economic operation,power supply reliability and CO2 emissions.With the battery State of Charge(SOC)as the state variable,the operation strategy of RIES is obtained for its 4-hour off-grid operation.(3)Proposed an operating condition prediction model for RIES based on multi-time scale synchronous prediction method.The forecasting time horizon is analyzed for the RIES operating conditions.It is determined that the prediction values at adjacent moments are moderately or highly correlated with a forecasting time horizon of 25 minutes.The sample state space was constructed based on the sampled operating conditions.The forecasting results under different forecasting time horizon are compared for wind speed,solar radiation intensity and class I electrical load working conditions using single-time-scale cyclic prediction method.The errors in terms of Root Mean Squared Error(RMSE)are analyzed.(4)Proposed optimal operation strategy for off-grid RIES based on stochastic model predictive control.The online rolling optimization framework is constructed with Markov Chain Monte Carlo(MCMC)method as the prediction model and Dynamic Programming algorithm as the optimization model.The Stochastic Model Predictive Control(SMPC)strategy and Dynamic Programming strategy were simulated and compared under the same operating conditions,and their performance in terms of operation economy and CO2 emission was comprehensively analyzed.