[1]CHEN Zhipeng,ZHANG Yong,GAO Hairong,et al.Integrated energy multivariate load forecasting combining federated learning with LSTM in privacy-protected and low-data environments[J].CAAI Transactions on Intelligent Systems,2024,19(3):565-574.[doi:10.11992/tis.202208049]
Copy

Integrated energy multivariate load forecasting combining federated learning with LSTM in privacy-protected and low-data environments

CAAI Transactions on Intelligent Systems[ISSN 1673-4785/CN 23-1538/TP] Volume: 19 Number of periods: 2024 3 Page number: 565-574 Column: 学术论文—机器学习 Public date: 2024-05-05
Title:
Integrated energy multivariate load forecasting combining federated learning with LSTM in privacy-protected and low-data environments
Author(s):
CHEN Zhipeng1;  ZHANG Yong1;  GAO Hairong2;  SUN Xiaoyan1;  HU Hejuan1
1. School of Information and Control Engineering, China University of Mining and Technology, Xuzhou 221116, China;
2. Department of Traffic Engineering, Shandong Transport Vocational College, Weifang 261206, China
Keywords:
multivariate load forecasting; integrated energy system; federated learning; privacy preserving; neural network; few data; time series data prediction; dot product protocol
CLC:
TP18
DOI:
10.11992/tis.202208049
Abstract:
For an integrated energy system with insufficient energy consumption data, a high-precision multivariate load forecasting model can be established using data from similar systems. However, due to the limitations of data security and other factors, many systems are unwilling to share data. Federated learning provides an important idea to deal with the problem of multivariate energy load forecast based on a small amount of data under privacy protection. However, the existing methods still exhibit deficiencies, such as low accuracy in identifying similar parties. In this view, a few-data multitask learning based on shared dot product confidentiality under federated learning (MT-SDP-FL) is proposed, combining federated learning and long short-term memory (LSTM). A similar party identification method using a shared vector dot product confidentiality protocol is proposed to select the most similar parties from many available integrated energy systems. Then, the parameter sharing federated learning algorithm is used to jointly train the selected participants, combining the LSTM and fine-tuning technology to establish the multivariate load prediction model for each participant. The proposed method is applied to several energy systems, and the experimental outcomes show that the proposed method can achieve high-precision multi-source load forecasting results in the circumstance of sparse data.
References:
[1] 贾宏杰, 王丹, 徐宪东, 等. 区域综合能源系统若干问题研究[J]. 电力系统自动化, 2015, 39(7): 198–207
JIA Hongjie, WANG Dan, XU Xiandong, et al. Research on some key problems related to integrated energy systems[J]. Automation of electric power systems, 2015, 39(7): 198–207
[2] INFIELD D G, HILL D C. Optimal smoothing for trend removal in short term electricity demand forecasting[J]. IEEE transactions on power systems, 1998, 13(3): 1115–1120.
[3] LI Yiyan, HAN Dong, YAN Zheng. Long-term system load forecasting based on data-driven linear clustering method[J]. Journal of modern power systems and clean energy, 2018, 6(2): 306–316.
[4] JIN Min, ZHOU Xiang, ZHANG Z M, et al. Short-term power load forecasting using grey correlation contest modeling[J]. Expert systems with applications, 2012, 39(1): 773–779.
[5] 高亚静, 孙永健, 杨文海, 等. 基于新型人体舒适度的气象敏感负荷短期预测研究[J]. 中国电机工程学报, 2017, 37(7): 1946–1955
GAO Yajing, SUN Yongjian, YANG Wenhai, et al. Weather-sensitive load’s short-term forecasting research based on new human body amenity indicator[J]. Proceedings of the CSEE, 2017, 37(7): 1946–1955
[6] PANDEY A S, SINGH D, SINHA S K. Intelligent hybrid wavelet models for short-term load forecasting[J]. IEEE transactions on power systems, 2010, 25(3): 1266–1273.
[7] 曾鸣, 吕春泉, 田廓, 等. 基于细菌群落趋药性优化的最小二乘支持向量机短期负荷预测方法[J]. 中国电机工程学报, 2011, 31(34): 93–99,11
ZENG Ming, LYU Chunquan, TIAN Kuo, et al. Least squares-support vector machine load forecasting approach optimized by bacterial colony chemotaxis method[J]. Proceedings of the CSEE, 2011, 31(34): 93–99,11
[8] 张素香, 赵丙镇, 王风雨, 等. 海量数据下的电力负荷短期预测[J]. 中国电机工程学报, 2015, 35(1): 37–42
ZHANG Suxiang, ZHAO Bingzhen, WANG Fengyu, et al. Short-term power load forecasting based on big data[J]. Proceedings of the CSEE, 2015, 35(1): 37–42
[9] SAFTA C, CHEN R L Y, NAJM H N, et al. Efficient uncertainty quantification in stochastic economic dispatch[J]. IEEE transactions on power systems, 2017, 32(4): 2535–2546.
[10] CHENG Lilin, ZANG Haixiang, DING Tao, et al. Multi-meteorological-factor-based graph modeling for photovoltaic power forecasting[J]. IEEE transactions on sustainable energy, 2021, 12(3): 1593–1603.
[11] HE Feifei, ZHOU Jianzhong, FENG Zhongkai, et al. A hybrid short-term load forecasting model based on variational mode decomposition and long short-term memory networks considering relevant factors with Bayesian optimization algorithm[J]. Applied energy, 2019, 237: 103–116.
[12] POTO?NIK P, ?KERL P, GOVEKAR E. Machine-learning-based multi-step heat demand forecasting in a district heating system[J]. Energy and buildings, 2021, 233: 110673.
[13] SIMON Haykin. 神经网络原理[M]. 叶世伟, 史忠植, 译. 北京: 机械工业出版社, 2004.
[14] CAO Jian, LI Zhi, LI Jian. Financial time series forecasting model based on CEEMDAN and LSTM[J]. Physica A:statistical mechanics and its applications, 2019, 519: 127–139.
[15] MUZAFFAR S, AFSHARI A. Short-term load forecasts using LSTM networks[J]. Energy procedia, 2019, 158: 2922–2927.
[16] 李鹏, 何帅, 韩鹏飞, 等. 基于长短期记忆的实时电价条件下智能电网短期负荷预测[J]. 电网技术, 2018, 42(12): 4045–4052
LI Peng, HE Shuai, HAN Pengfei, et al. Short-term load forecasting of smart grid based on long-short-term memory recurrent neural networks in condition of real-time electricity price[J]. Power system technology, 2018, 42(12): 4045–4052
[17] SOMU N, RAMAN M R G, RAMAMRITHAM K. A deep learning framework for building energy consumption forecast[J]. Renewable and sustainable energy reviews, 2021, 137: 110591.
[18] MCMAHAN H B, MOORE E, RAMAGE D, et al. Communication-efficient learning of deep networks from decentralized data[EB/OL]. (2016–02–17)[2022–08–30]. https://arxiv.org/abs/1602.05629.
[19] 杨强. 联邦学习: 人工智能的最后一公里[J]. 智能系统学报, 2020, 15(1): 183–186
YANG Qiang. Federated learning: the last on kilometer of artificial intelligence[J]. CAAI transactions on intelligent systems, 2020, 15(1): 183–186
[20] BRISIMI T S, CHEN Ruidi, MELA T, et al. Federated learning of predictive models from federated Electronic Health Records[J]. International journal of medical informatics, 2018, 112: 59–67.
[21] SUBRAMANYA T, RIGGIO R. Centralized and federated learning for predictive VNF autoscaling in multi-domain 5G networks and beyond[J]. IEEE transactions on network and service management, 2021, 18(1): 63–78.
[22] WU Xing, LIANG Zhaowang, WANG Jianjia. FedMed: a federated learning framework for language modeling[J]. Sensors, 2020, 20(14): 4048.
[23] HUA Gaofeng, ZHU Li, WU Jinsong, et al. Blockchain-based federated learning for intelligent control in heavy haul railway[J]. IEEE access, 2020, 8: 176830–176839.
[24] SAVI M, OLIVADESE F. Short-term energy consumption forecasting at the edge: a federated learning approach[J]. IEEE access, 2021, 9: 95949–95969.
[25] CHEN Yujing, NING Yue, CHAI Zheng, et al. Federated multi-task learning with hierarchical attention for sensor data analytics[C]//2020 International Joint Conference on Neural Networks. Glasgow: IEEE, 2020: 1–8.
[26] 王鑫, 周泽宝, 余芸, 等. 一种面向电能量数据的联邦学习可靠性激励机制[J]. 计算机科学, 2022, 49(3): 31–38
WANG Xin, ZHOU Zebao, YU Yun, et al. Reliable incentive mechanism for federated learning of electric metering data[J]. Computer science, 2022, 49(3): 31–38
[27] HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural computation, 1997, 9(8): 1735–1780.
[28] 刘旭红. 高效安全向量计算及其推广[J]. 软件学报, 2021, 32(11): 3628–3645
LIU Xuhong. Efficient secure vector computation and its extension[J]. Journal of software, 2021, 32(11): 3628–3645
[29] 燕忠毅, 曾艳, 赵乃良, 等. 一种基于二阶导数解决联邦学习中数据不平衡问题的方法: CN113691594A[P]. 2021–11–23.
YAN Zhongyi, ZENG Yan, ZHAO Nailiang, et al. Method for solving data imbalance problem in federated learning based on second derivative: CN113691594A[P]. 2021–11–23.
[30] 刘叶. 数据不平衡的联邦学习方法研究[D]. 北京: 北京邮电大学, 2021.
LIU Ye. Research on federatedlearning methods for unbalanced data[D]. Beijing: Beijing University of Posts and Telecommunications, 2021.
[31] 孙庆凯, 王小君, 张义志, 等. 基于LSTM和多任务学习的综合能源系统多元负荷预测[J]. 电力系统自动化, 2021, 45(5): 63–70
SUN Qingkai, WANG Xiaojun, ZHANG Yizhi, et al. Multiple load prediction of integrated energy system based on long short-term memory and multi-task learning[J]. Automation of electric power systems, 2021, 45(5): 63–70
[32] PENG Chao, TAO Yifan, CHEN Zhipeng, et al. Multi-source transfer learning guided ensemble LSTM for building multi-load forecasting[J]. Expert systems with applications, 2022, 202: 117194.
Similar References:

Memo

-

Last Update: 1900-01-01

Copyright © CAAI Transactions on Intelligent Systems