基于LSTM分位数回归的零碳能源系统不确定性规划方法研究

Uncertainty Plan of a Zero-carbon Energy System Based on LSTM Quantile Regression

零碳能源系统是建设清洁低碳能源体系、促进碳中和目标达成的有效途径。但风、光等可再生能源发电的间歇性和波动性给零碳能源系统的合理规划与科学配置带来巨大挑战。基于此,提出了基于长短期记忆网络(LSTM)分位数回归的零碳能源系统不确定性规划方法。首先,构建了零碳能源系统的物理结构,并对系统内典型设备进行建模;其次,基于区间数理论思想,构建了基于LSTM分位数回归的光伏发电区间预测模型;然后,以系统年总经济成本最小为目标函数,构建了可实现设备配置与运行策略协同优化的混合整数线性规划模型;最后,通过典型算例分析,验证了所提模型应对零碳能源系统不确定性规划的可行性。

Zero-carbon energy system is an effective way to promote the construction of clean and low-carbon energy system and realize carbon neutralization.However,the intermittence and volatility of wind,solar and other renewable energy power generation have brought great challenges to the rational planning and scientific allocation of zero-carbon energy system.In this paper,an uncertainty programming method for zero-carbon energy system based on LSTM quantile regression is proposed.Firstly,the physical structure of the zero-carbon energy system is constructed,and the typical equipment in the system is modeled.Then,based on the interval number theory,the interval prediction model of photovoltaic power generation based on LSTM quantile regression is constructed.Taking the minimum annual total cost as the objective function,the mixed integer linear programming model which can optimize the equipment configuration and operation schedule is constructed.Finally,through the numerical study,the correctness and feasibility of the proposed model is verified.