基于预训练的大模型赋能场景规划和双层嵌套的多能互补系统优化调度

doi:10.19595/j.cnki.1000-6753.tces.230354

Abstract
Figure/Table
References (0)
Related Citation (15)

Download: PDF (2243 KB) HTML (1 KB)
Export: BibTeX | EndNote (RIS)

Abstract

Renewable energy source (RES) power generation has strong stochastic volatility, and large-scale grid-connection will have an impact on the consumption of RES and scheduling operation planning of the power system, which is not conducive to stable economic operation. Constructing multi-energy complementary systems represents a viable solution to enhance energy utilization and mitigate dispatch operational risks. Scholars at home and abroad have conducted extensive research on the problem of multi-energy complementarity, and the following problems still exist at this stage: (1) In the regional power grid with large RES integration, thermal power and conventional hydropower units will frequently start-stop, climb over the limit, adjust the overdraft. Meanwhile, technical difficulties such as the high cost and large capacity integration of chemical energy storage have not been broken through. (2) The performance of existing power forecasting models needs to be further improved, and accurate prediction of wind, solar and load power is the premise of efficient utilization of RES. To address these issues, this paper proposed a pre-trained large language model (PLLM) based on empowering scenario planning and bi-level nested optimization of wind-solar-hydro-thermal-pumped storage complementary scheduling model.
Firstly, after cleaning and normalizing the data, fixed step time series input tokens are constructed to enhance the model's capability to capture local data features and reduce the redundancy of input information. The prompt word pattern framework is innovatively designed to integrate the context feature information with the input tokens to guide the PLLM reasoning process. In order to improve the wind-solar-load integrated power prediction performance of PLLM, an efficient fine-tuning strategy is introduced to adjust the relevant weight parameters. Subsequently, The K-means clustering algorithm integrated with PLLM is constructed to generate supply-demand coupling operation scenarios conforming to the power system scheduling production simulation according to the integrated power prediction results, which provides scenarios support for the next optimal scheduling model. Finally, a bi-level nested optimization method is proposed for the coordinated operation of each power source in wind-solar-hydro-thermal-pumped storage complementary system. The upper level optimizes the stability of the output power of the system to guarantee the consumption of RES, while the lower level optimizes the low-carbon economic scheduling of the system and the thermal power unit commitments to realize the economy and environmental protection.
According to the test results of the actual data sets, compared with the traditional neural network prediction models gate recurrent unit (GRU), long-short term memory (LSTM), temporal convolutional network (TCN) and bidirectional long-short term memory (BiLSTM) integrated network TCN-BiLSTM. The evaluation indexes of the model's prediction performance, mean absolute error (MAE), root mean absolute error (RMSE) and mean absolute percentage error (MAPE) are decreased by an average of 46.45%, 42.08% and 23.25%, respectively, and R² increased by an average of 7.26%. The superiority and adaptability of the proposed PLLM in power prediction are demonstrated. The simulation results of a base project in a northwestern province demonstrates that the proposed bi-level nested solution method effectively coordinates the rapid regulation capabilities of conventional hydropower and pumped storage to respond to real-time fluctuations in wind-solar-load integrated power output, and realize the economy and low-carbon of the system operation on the basis of ensuring the stability of the output power. After pumped storage is involved in the complementary operation, the average values of the source load fluctuation smoothing factor, system operating cost and CO₂ emissions are reduced by 13.92%, 1.06% and 1.24%, respectively, the average values of total number of thermal power units, start-stop times and start-stop costs are reduced by 2 times, 4 times and 20,800 yuan, respectively. When pumped storage participates in the peaking operation of the coupled system, it can significantly improve the flexible adjustment margin, alleviate the peaking pressure of thermal power units, and greatly reduce the operating cost and CO₂ emissions of the system.

Key words： Integration of wind-solar-load power forecasting pre-trained large language model bi-level nested optimization multi-energy complementary

Received: 06 March 2025

PACS:	TM61
	TM732

	Service

	E-mail this article
	Add to my bookshelf
	Add to citation manager
	E-mail Alert
	RSS
	Articles by authors
	Wang Kaiyan
	Zhu Hengtao
	Jia Rong
	Ming Bo
	Dang Jian

Cite this article:

Wang Kaiyan,Zhu Hengtao,Jia Rong等. Pre-trained Large Model Empowering Scenario Planning and Bi-Level Nested Optimization for Optimal Scheduling of Multi-Energy Complementary System[J]. Transactions of China Electrotechnical Society, 0, (): 20250354-20250354.

URL:

https://dgjsxb.ces-transaction.com/EN/10.19595/j.cnki.1000-6753.tces.230354 OR https://dgjsxb.ces-transaction.com/EN/Y0/V/I/20250354

[1] 李晖, 刘栋, 姚丹阳. 面向碳达峰碳中和目标的我国电力系统发展研判[J]. 中国电机工程学报, 2021, 41(18): 6245-6259.
Li Hui, Liu Dong, Yao Danyang.Analysis and reflection on the development of power system towards the goal of carbon emission peak and carbon neutrality[J]. Proceedings of the CSEE, 2021, 41(18): 6245-6259.
[2] 任大伟, 侯金鸣, 肖晋宇, 等. 能源电力清洁化转型中的储能关键技术探讨[J]. 高电压技术, 2021, 47(8): 2751-2759.
Ren Dawei, Hou Jinming, Xiao Jinyu, et al. Exploration of key technologies for energy storage in the cleansing transformation of energy and power[J]. High Voltage Engineering, 2021, 47(8): 2751-2759.
[3] 姜云鹏, 任洲洋, 李秋燕, 等. 考虑多灵活性资源协调调度的配电网新能源消纳策略[J]. 电工技术学报, 2022, 37(7): 1820-1835.
Jiang Yunpeng, Ren Zhouyang, Li Qiuyan, et al. An accommodation strategy for renewable energy in distribution network considering coordinated dispatching of multi-flexible resources[J]. Transactions of China Electrotechnical Society, 2022, 37(7): 1820-1835.
[4] 赵志鹏, 于志辉, 程春田, 等. 水风光综合基地多风险量化及长期多目标协调优化调度方法[J]. 电力系统自动化, 2024, 48(22): 118-130.
Zhao Zhipeng, Yu Zhihui, Cheng Chuntian, et al. Multi-risk quantification and long-term multi-objective coordinative optimal dispatch method for hydro-wind-solar integrated energy base[J]. Automation of Electric Power Systems, 2024, 48(22): 118-130.
[5] 郭怿, 明波, 黄强, 等. 考虑输电功率平稳性的水-风-光-储多能互补日前鲁棒优化调度[J]. 电工技术学报, 2023, 38(9): 2350-2363.
Guo Yi, Ming Bo, Huang Qiang, et al.Day-ahead robust optimal scheduling of hydro-wind-PV-storage complementary system considering the steadiness of power delivery[J]. Transactions of China Electrotechnical Society, 2023, 38(9): 2350-2363.
[5] 郭怿, 明波, 黄强, 等. 考虑输电功率平稳性的水-风-光-储多能互补日前鲁棒优化调度[J]. 电工技术学报, 2023, 38(9): 2350-2363.
Guo Yi, Ming Bo, Huang Qiang, et al.Day-ahead robust optimal scheduling of hydro-wind-PV-storage complementary system considering the steadiness of power delivery[J]. Transactions of China Electrotechnical Society, 2023, 38(9): 2350-2363.
[6] 王开艳, 罗先觉, 贾嵘, 等. 充分发挥多能互补作用的风蓄水火协调短期优化调度方法[J]. 电网技术, 2020, 44(10): 3631-3641.
Wang Kaiyan, Luo Xianjue, Jia Rong, et al. Short-term coordinated scheduling of wind-pumped-hydro-thermal power system with multi-energy complementarities[J]. Power System Technology, 2020, 44(10): 3631-3641.
[7] 孙惠, 翟海保, 吴鑫. 源网荷储多元协调控制系统的研究及应用[J]. 电工技术学报, 2021, 36(15): 3264-3271.
Sun Hui, Zhai Haibao, Wu Xin. Research and application of multi-energy coordinated control of generation, network, load and storage[J]. Transactions of China Electrotechnical Society, 2021, 36(15): 3264-3271.
[8] 李建林, 郭兆东, 马速良, 等. 新型电力系统下“源网荷储” 架构与评估体系综述[J]. 高电压技术, 2022, 48(11): 4330-4342.
Li Jianlin, Guo Zhaodong, Ma Suliang, et al. Overview of the ‘‘source-grid-load-storage'' architecture and evaluation system under the new power system[J]. High Voltage Engineering, 2022, 48(11): 4330-4342.
[9] 韩丽, 王冲, 于晓娇, 等. 考虑风电爬坡灵活调节的碳捕集电厂低碳经济调度[J]. 电工技术学报, 2024, 39(7): 2033-2045.
Han Li, Wang Chong, Yu Xiaojiao, et al. Low-carbon and economic dispatch considering the carbon capture power plants with flexible adjustment of wind power ramp[J]. Transactions of China Electrotechnical Society, 2024, 39(7): 2033-2045.
[10] 袁文腾, 陈亮, 王春波, 等. 基于氨储能技术的电转氨耦合风-光-火综合能源系统双层优化调度[J]. 中国电机工程学报, 2023, 43(18): 6992-7003.
Yuan Wenteng, Chen Liang, Wang Chunbo, et al. Bi-level optimal scheduling of power-to-ammonia coupling wind-photovoltaic-thermal integrated energy system based on ammonia energy storage technology[J]. Proceedings of the CSEE, 2023, 43(18): 6992-7003.
[11] 李军徽, 张嘉辉, 李翠萍, 等. 参与调峰的储能系统配置方案及经济性分析[J]. 电工技术学报, 2021, 36(19): 4148-4160.
Li Junhui, Zhang Jiahui, Li Cuiping, et al. Configuration scheme and economic analysis of energy storage system participating in grid peak shaving[J]. Transactions of China Electrotechnical Society, 2021, 36(19): 4148-4160.
[12] 谭乔凤, 聂状, 闻昕, 等. 大规模风光接入下梯级水电站调度方式研究[J]. 水力发电学报, 2022, 41(9): 44-55.
Tan Qiaofeng, Nie Zhuang, Wen Xin, et al. Operation modes of cascade hydropower stations considering large-scale integration of wind and photovoltaic power[J]. Journal of Hydroelectric Engineering, 2022, 41(9): 44-55.
[13] Li Huanhuan, Zhang Runfan, Mahmud M A, et al. A novel coordinated optimization strategy for high utilization of renewable energy sources and reduction of coal costs and emissions in hybrid hydro-thermal-wind power systems[J]. Applied Energy, 2022, 320: 119019.
[14] 杨茂, 王达, 王小海, 等. 基于数据物理混合驱动的超短期风电功率预测模型[J]. 高电压技术, 2024, 50(11): 5132-5141.
Yang Mao, Wang Da, Wang Xiaohai, et al. Ultra-short term wind power prediction method based on data physics hybrid driven model[J]. High Voltage Engineering, 2024, 50(11): 5132-5141.
[15] 邹金, 朱继忠, 赖旭, 等. 基于时空自回归移动平均模型的风电出力序列模拟[J]. 电力系统自动化, 2019, 43(3): 101-107.
Zou Jin, Zhu Jizhong, Lai Xu, et al. Simulation of wind power output series based on space-time auto-regressive moving average model[J]. Automation of Electric Power Systems, 2019, 43(3): 101-107.
[16] 彭曙蓉, 陈慧霞, 孙万通, 等. 基于改进LSTM的光伏发电功率预测方法研究[J]. 太阳能学报, 2024, 45(11): 296-302.
Peng Shurong, Chen Huixia, Sun Wantong, et al. Research on photovoitaic power prediction method based on improved LSTM[J]. Acta Energiae Solaris Sinica, 2024, 45(11): 296-302.
[17] Harbola S, Coors V. One dimensional convolutional neural network architectures for wind prediction[J]. Energy Conversion and Management, 2019, 195: 70-75.
[18] 钟吴君, 李培强, 涂春鸣. 基于EEMD-CBAM-Bi LSTM的牵引负荷超短期预测[J]. 电工技术学报, 2024, 39(21): 6850-6864.
Zhong Wujun, Li Peiqiang, Tu Chunming. Traction load ultra-short-term forecasting framework based on EEMD-CBAM-BiLSTM[J]. Transactions of China Electrotechnical Society, 2024, 39(21): 6850-6864.
[19] 张孝顺, 李锦诚, 郭正勋. 大模型辅助的大型海上风电场集电系统拓扑优化[J]. 高电压技术, 2024, 50(7): 2894-2905.
Zhang Xiaoshun, Li Jincheng, Guo Zhengxun. Topology optimization of large-scale offshore wind farm collector systems based on large language models[J]. High Voltage Engineering, 2024, 50(7): 2894-2905.
[20] 曹祎, 张莉, 郭静, 等. 基于大语言模型的低碳电力市场发展应用前景[J]. 智慧电力, 2024, 52(2): 8-16.
Cao Yi, Zhang Li, Guo Jing, et al. Prospects for development of low-carbon electricity markets based on large language models[J]. Smart Power, 2024, 52(2): 8-16.
[21] 李莉, 时榕良, 郭旭, 等. 融合大模型与图神经网络的电力设备缺陷诊断[J]. 计算机科学与探索, 2024, 18(10): 2643-2655.
Li Li, Shi Rongliang, Guo Xu, et al. Diagnosis of power system defects by large language models and graph neural networks[J]. Journal of Frontiers of Computer Science and Technology, 2024, 18(10): 2643-2655.
[22] Bonadia R S, Trindade F C L, Freitas W, et al. On the potential of ChatGPT to generate distribution systems for load flow studies using OpenDSS[J]. IEEE Transactions on Power Systems, 2023, 38(6): 5965-5968.
[23] Gao Mingyang, Zhou Suyang, Gu Wei, et al. LFPLM: a general and flexible load forecasting framework based on pre-trained language model[J]. arXiv preprint arXiv:2406. 11336, 2024.
[24] Hu E, Shen Yelong, Wallis P, et al.Lora: low-rank adaptation of large language models[J]. arXiv preprint arXiv:2106.09685, 2021.
[25] 陈帝伊, 张猛, 刘泳, 等. 考虑避振条件的水风光互补发电系统运行经济性评估[J]. 水利学报, 2024, 55(4): 403-415.
Chen Diyi, Zhang Meng, Liu Yong, et al.Operational economic evaluation of hydro-wind-photovoltaic power generation system considering the vibration avoidance strategy[J]. Journal of Hydraulic Engineering, 2024, 55(4): 403-415.
[26] Wang Kaiyan, Zhu Hengtao, Dang Jian, et al. Short-term optimal scheduling of wind-photovoltaic-hydropower-thermal-pumped hydro storage coupled system based on a novel multi-objective priority stratification method[J]. Energy, 2024, 309: 133190.
[27] Wu Xingyu, Wu Shenghao, Wu Jibin, et al. Evolutionary computation in the era of large language model: survey and roadmap[J]. IEEE Transactions on Evolutionary Computation, 2025, 29(2): 534-554.
[28] 董俊, 束洪春, 刘瑞, 等. 大语言模型赋能场景生成和双层优化的多农业园区供电-灌溉-蓄水耦合运行[J]. 高电压技术, 2024, 50(7): 2906-2917.
Dong Jun, Shu Hongchun, Liu Rui, et al. Large language models empowering scenario generation and dual-layer optimization for coupled operations of power supply, irrigation, and water storage in multiple agricultural parks[J]. High Voltage Engineering, 2024, 50(7): 2906-2917.
[29] 王开艳, 杜浩东, 贾嵘, 等. 基于相似日聚类和QR-CNN-BiLSTM模型的光伏功率短期区间概率预测[J]. 高电压技术, 2022, 48(11): 4372-4388.
Wang Kaiyan, Du Haodong, Jia Rong, et al. Short-term interval probability prediction of photovoltaic power based on similar daily clustering and QR-CNN-BiLSTM model[J]. High Voltage Engineering, 2022, 48(11): 4372-4388.
[30] Li Xudong, Yang Weijia, Zhao Zhigao, et al. Advantage of priority regulation of pumped storage for carbon-emission-oriented co-scheduling of hybrid energy system[J]. Journal of Energy Storage, 2023, 58: 106400.