数据与模型驱动的高速公路服务区交通自洽能源系统状态估计方法

石志鹏; 金雨哲; 柯吉; 王飚; 张懿璞

doi:10.3963/j.jssn.1674-4861.2024.05.016

数据与模型驱动的高速公路服务区交通自洽能源系统状态估计方法

doi: 10.3963/j.jssn.1674-4861.2024.05.016

石志鹏^1,,
金雨哲²,
柯吉²,
王飚²,
张懿璞^2, ,

1.
长安大学电子与控制工程学院西安 710064
2.
长安大学能源与电气工程学院西安 710064

基金项目:

国家重点研发计划项目 2021YFB1600200

中央高校基本科研业务费专项项目 300102383202

详细信息

作者简介:
石志鹏（2000—），硕士研究生. 研究方向：交通能源融合. E-mail: 2022232067@chd.edu.cn

通讯作者:
张懿璞（1985—），博士，教授. 研究方向：交通能源融合. E-mail: zyipu@chd.edu.cn

中图分类号: U495
计量
- 文章访问数: 55
- HTML全文浏览量: 13
- PDF下载量: 6
- 被引次数: 0
出版历程
- 收稿日期: 2024-01-31

A Method for Data and Model Driven Estimation of Traffic Self-Consistent Energy System States in Highway Service Areas

SHI Zhipeng^1
,,
JIN Yuzhe²,
KE Ji²,
WANG Biao²,
ZHANG Yipu^{2
, ,}

1.
School of Electronic and Control Engineering, Chang'an University, Xi'an 710064, China
2.
School of Energy and Electronic Engineering, Chang'an University, Xi'an 710064, China

摘要

摘要: 构建高速公路服务区交通自洽能源系统是实现交通与能源融合的关键技术，而对其进行系统状态估计是近年来的研究重点。考虑到高速公路服务区交通自洽能源系统的复杂性和多样性，单一的数据或模型驱动方法难以全面、准确地估计系统实时状态。因此，研究了1种数据与模型驱动的复合方法，旨在实现更高效的系统状态估计。在数据驱动方面，尽管基于深度学习的光伏功率预测模型性能优越，但通常忽视输入特征间的互相依赖机制。为此，建立了基于自注意力机制（self-attention, SA）的时间卷积-双向长短期记忆网络（time convolution-bidirectional long short-term memory network，TCN-BiLSTM-SA），用于预测系统光伏出力情况。SA重新分配TCN-BiLSTM输入特征的权重，从而提升时空信息提取的有效性。在模型驱动方面，考虑高速公路路网车流量分布，建立了高速公路电动汽车出行轨迹概率模型；基于蒙特卡洛模拟法抽取初始和充电时电池容量，综合考虑车主用车习惯、环境温度等多种不确定性因素，来预测电动汽车充电负荷时空分布。通过利用新疆某高速公路服务区交通自洽能源系统的实际数据进行仿真验证，结果表明：在光伏预测上，所提模型在平均绝对误差、均方根误差和决定系数这3个指标上，相较于最佳模型分别提高了25.3%、16.7%、0.7%；在负荷预测上，所提模型有效预测高速公路电动汽车的充电负荷时空分布；在系统状态估计上，所提方法的精度达到了89.1%。
- 交通与能源融合 /
- 系统状态估计 /
- TCN-BiLSTM-SA模型 /
- 数据与模型驱动 /
- 蒙特卡洛法
Abstract: The development construction of a self-consistent energy system for highway service areas is essential technology for integrating transportation and energy. A key focus for recent research is system state estimation. Given the complexity and diversity of these energy systems, relying soley on either data-driven or model-driven approaches often leads to challenges in accurately and comprehensively estimating real-time system states. Therefore, this study explores a hybrid method combining data-driven and model-driven approaches for more efficient system state estimation. In terms of data-driven methods, although deep learning-based photovoltaic power forecasting models demonstrate superior performance, they fail to account for the interdependencies among input features. To address this issue, we developed a Time Convolution-Bidirectional Long Short-Term Memory network (TCN-BiLSTM-SA) based on a Self-Attention mechanism (SA) to predict the photovoltaic output of the system. The SA module adjusts redistributes the weights of the TCN-BiLSTM input features, improving the extraction of spatiotemporal information. On the model-driven side, considering the traffic flow distribution of highway networks, we established a probability model for electric vehicle travel trajectories. By utilizing Monte Carlo simulations, the initial and charging battery capacities are extracted while accounting for various uncertainties such as driver behavior and environmental temperature, thereby approximating accurate results to predict the spatiotemporal distribution of electric vehicle charging loads. Simulation validation using actual data from a self-consistent energy system at a highway service area in Xinjiang indicated that, in terms of photovoltaic forecasting, the proposed model improved the mean absolute error, root mean square error, and coefficient of determination by 25.3%, 16.7%, and 0.7%, respectively, compared to the best model. Furthermore, the proposed model effectively predicted the spatiotemporal distribution of electric vehicle charging loads on highways, achieving a system state estimation accuracy of 89.1%.
- integration of transportation and energy /
- system state estimation /
- TCN-BiLSTM-SA model /
- data- and model-driven /
- Monte Carlo

HTML全文

图 1 所提模型图

Figure 1. Proposedmodel diagram

下载: 全尺寸图片幻灯片

图 2 不同典型日日均车流量对比

Figure 2. Comparison of average daily traffic volume for different typical days

下载: 全尺寸图片幻灯片

图 3 EV充电负荷计算流程

Figure 3. EV charging load calculation process.

下载: 全尺寸图片幻灯片

图 4 所选高速路段的路网拓扑

Figure 4. The road network topology of the selected expressway section.

下载: 全尺寸图片幻灯片

图 5 各模型光伏出力预测效果对比

Figure 5. Comparison of Photovoltaic Output Prediction Effects Among Different Models

下载: 全尺寸图片幻灯片

图 6 添加自注意力前后模型预测结果对比

Figure 6. Comparsion of model predictions before and after adding self-attention

下载: 全尺寸图片幻灯片

图 7 典型日车辆出行时间概率分布

Figure 7. Probability distribution of vehicle travel time on typical days

下载: 全尺寸图片幻灯片

图 8 冬季-夏季典型日气温分布

Figure 8. Temperature distribution on typical days in summer and winter

下载: 全尺寸图片幻灯片

图 9 夏季典型日充电负荷分布

Figure 9. Charging load distribution on typical days in summer

下载: 全尺寸图片幻灯片

图 10 冬季典型日充电负荷分布

Figure 10. Charging load distribution on typical days in winter

下载: 全尺寸图片幻灯片

图 11 系统储能容量变化曲线

Figure 11. Change curve of system energy storage capacity

下载: 全尺寸图片幻灯片

表 1 系统运行状态划分

Table 1. Classification of system operation status

系统运行状态	表现形式
能量不足	$\left\{\begin{array}{l} P_{\text {load }}(t)-P_{\mathrm{pv}}(t)>P_{\mathrm{grid} \max }(t)+P_{f \max }(t)(\text { 并网 }) \\ P_{\text {load }}(t)-P_{\mathrm{pv}}(t)>P_{\mathrm{dg} \max }(t)+P_{f \max }(t)(\text { 离网 }) \end{array}\right.$
能量严重不足	$\left\{\begin{array}{l} P_{\text {load }}(t)-P_{\mathrm{pv}}(t)>P_{\text {grid max }}(t)+P_{f \text { max }}(t)(\text { 并网 }) \\ P_{\text {load }}(t)-P_{\mathrm{pv}}(t)>P_{\mathrm{dg} \text { max }}(t)+P_{f \text { max }}(t)(\text { 离网 }) \end{array}\right.$
能量最优利用	$P_{\text {load }}(t)-P_{\mathrm{pv}}(t)=0$
能量充足	$\left\{\begin{array}{l} 0<P_{\text {load }}(t)-P_{\mathrm{pv}}(t) \leqslant P_{\text {grid max }}(t)+P_{f \text { max }}(t) \text { (并网) } \\ 0<P_{\text {load }}(t)-P_{\mathrm{pv}}(t) \leqslant P_{\mathrm{dg} \max }(t)+P_{f \text { max }}(t)(\text { 离网 }) \\ P_{f \text { max }}(t)=\left[S O C(t-1)-S O C_{\min }\right]\cdot E_b \end{array}\right.$
能量充足	$\left\{\begin{array}{l} 0<P_{\mathrm{pv}}(t)-P_{\text {load }}(t) \leqslant P_{c \max }(t) \\ P_{c \max }(t)=\left[S O C_{\max }-\operatorname{SOC}(t-1)\right] \cdot E_b \end{array}\right.$
能量过剩	$P_{\mathrm{pv}}(t)-P_{\text {1oad }}(t)>P_{c \max }(t)$

下载: 导出CSV

表 2 各模型在光伏数据集上预测效果对比

Table 2. Comparison of prediction effects of various models in PV datasets

模型类别	评价指标
模型类别	MAE	RMSE	R²
本文方法	9.203 1	11.730 2	0.889 3
TCN-BiLSTM	9.456 6	11.897 5	0.881 5
BiLSTM	11.251 0	13.238 1	0.816 5
CNN	14.358 0	23.850 0	0.751 1
RF	16.781 1	29.004 5	0.749 2

下载: 导出CSV

表 3 系统状态估计结果

Table 3. Results of system state estimation

系统状态类别	系统状态总数	验证集状态数	相应时刻正确估计的状态数	精度/%
能量不足	1 783	325	304	93.5
能量严重不足	0	0
能量最优利用	73	15	9	60
能量充足	1 402	232	211	90.9
能量过剩	438	52	32	61.5
总计	3 696	624	556	89.1

下载: 导出CSV

参考文献(35)

[1]	杨勇平, 武平, 程鹏, 等. 我国陆路交通能源系统发展战略研究[J]. 中国工程科学, 2022, 24(3): 153-162. YANGY P, WU P, CHENG P, etal. Development strategy for energy system ofland transport in China[J]. Engineering Sciences, 2022, 24(3): 153-162. (in Chinese)
[2]	蒋海峰, 黄森炎, 刘志强, 等. 公路交通设施与清洁能源融合创新技术与发展[J]. 公路交通科技, 2020, 37(增刊2): 1-5. JIANG H F, HUANG S Y, LIU Z Q, et a-l. Innovative technology and development of integration of highway trafficfacilities and clean energy[J]. Journal of Highway and Transportation Research and Development, 2020, 37(S2): 1-5. (in Chinese)
[3]	李瑞杰. 面向绿色低碳的韧性公路交通-能源融合发展技术和展望[J]. 交通节能与环保, 2024, 20(5): 1-7. LI R J. Development technology and prospect for green and low-carbon resilient road transportenergy Integration[J]. Transport Energy Conservation & Environmental, 2024, 20 (5): 1-7. (in Chinese).
[4]	贾利民, 师瑞峰, 吉莉, 等. 我国道路交通与能源融合发展战略研究[J]. 中国工程科学, 2022, 24(3): 163-172. JIA L M, SHI R F, JI L, et al. Road transportation and energy integration strategy in China[J]. Strategic Study of CAE, 2022, 24(3): 163-172. (in Chinese).
[5]	毛宁, 于琦, 刘杰, 等. 交能融合宏观策略与技术路径[J]. 中国公路, 2022(22): 28-31. MAO N, YU Q, LIU J, et al. Macro strategies and technical pathways for the integration of transportation and energy[J]. China Highway, 2022(22): 28-31. (in Chinese).
[6]	桑博, 张涛, 刘亚杰, 等. 多微电网能量管理系统研究综述[J]. 中国电机工程学报, 2020, 40(10): 3077-3093. SANG B, ZHANG T, LIU Y J, et al. Energy management system research of multi-microgrid: a review[J]. Proceedings of the CSEE, 2020, 40(10): 3077-3093. (in Chinese)
[7]	刘继文, 侯强. 基于文献计量的微电网博弈调度研究热点分析[J]. 电气技术, 2023, 24(6): 13-20. doi: 10.3969/j.issn.1673-3800.2023.06.003 LIU J W, HOU Q, Hotspots analysis of research and development trend of microgrid game dispatching based on information visualization[J]. Electrical Engineering, 2023, 24(6): 13-20. (in Chinese) doi: 10.3969/j.issn.1673-3800.2023.06.003
[8]	汪诗怡, 艾芊. 含微电网的电力系统状态估计研究[J]. 电器与能效管理技术, 2014(16): 33-38. WANG S Y, AI Q, Analysis on power system state estimation with microgrid[J]. Low Voltage Apparatus, 2014(16): 33-38. (in Chinese)
[9]	RANA M, LI L. Kalman filter based microgrid state estimation using the internet of things communication network[C]. 12^th International Conference on Information Technology-New Generations, Nevada USA: IEEE, 2015.
[10]	SUN H, FENG G, NIKOVSKI D, et al. Dynamic state estimation based on unscented Kalman filter and very short-term load and distributed generation forecasting[C]. International Conference on Power System Technology. Wollongong, NSW, Australia: IEEE, 2016.
[11]	YANG S, TAN Y, TAN Q, et al. State estimation of electromagnetic scanning micro-mirrors using unscented Kalman filtering method[C]. International Conference on Realtime Computing and Robotics(RCAR), Datong, China: IEEE, 2023.
[12]	刘晓莉, 曾祥晖, 黄翊阳, 等. 联合粒子滤波和卷积神经网络的电力系统状态估计方法[J]. 电网技术, 2020, 44(9): 3361-3367. LIU X L, ZENG X H, HUANG Y Y et al. State estimation based on particle filtering and convolutional neural networks for power systems[J]. Power System Technology, 2020, 44 (9): 3361-3367. (in Chinese)
[13]	MBUNGU N T, ISMAIL A A, ALSHAHI M, et al. Control and estimation techniques applied to smart microgrids: a review[J]. Renewable and Sustainable Energy Reviews, 2023: 113251.
[14]	黄小庆, 陈颉, 陈永新, 等. 大数据背景下的充电站负荷预测方法[J]. 电力系统自动化, 2016, 40(12): 68-74. HUANG X Q, CHEN J, CHEN Y X, et al. Charging station load forecasting method in the context of big data[J]. Automation of Electric Power System, 2016, 40(12): 68-74. (in Chinese).
[15]	张明辉, 周亚同, 孔晓然. 基于深度混合储备池计算模型的短期电力负荷预测[J]. 电网技术, 2022, 46(12): 4751-4762. ZHANG M H, ZHOU Y T, KONG X R. Short-term electrical load prediction based on deep hybrid reservoir calculation model[J]. Power System Technology, 2022, 46(12): 4751-4762. (in Chinese)
[16]	锁军, 李龙, 贺瀚青, 等. 考虑交通路况的电动汽车充电负荷预测[J]. 电网与清洁能源, 2022, 38(10): 141-147. SUO J, LI L, HE H Q, et al. Load forecasting of electric vehicle charging considering traffic conditions[J]. Power System and Clean Energy, 2022, 38(10): 141-147. (in Chinese)
[17]	罗浩成, 胡泽春, 张洪财. 环境温度对电动汽车充电负荷的影响分析[J]. 电力建设, 2015, 36(7): 69-74. LUO H C, HU Z C, ZHANG H C. Effect analysis of ambient temperature on electric vehicle charging load[J]. Electric Power Construction, 2015, 36(7): 69-74. (in Chinese)
[18]	葛少云, 冯亮, 刘洪, 等. 考虑电量分布及行驶里程的高速公路充电站规划[J]. 电力自动化设备, 2013, 33(7): 111-116. GE S Y, FENG L, LIU H, et al. Planning of charging stations on highway considering power distribution and driving mileage[J]. EIectric Power Automation Equipment, 2013, 33(7): 111-116. (in Chinese)
[19]	MERAL M E, DINCER F. A review of the factors affecting operation and efficiency of photovoltaic based electricity generation systems[J]. Renewable and Sustainable Energy Reviews, 2011, 15(5): 2176-2184.
[20]	周浩杰, 杨建卫, 王尊, 等. 基于LSTM光伏发电功率超短期预测模型研究[J]. 电源技术, 2023, 47(6): 785-789. ZHOU H J, YANG J W, WANG Z, et al. Research on ultrashort-term prediction model of photovoltaic power generation based on LSTM[J]. Chinese Journal of Power Sources, 2023, 47(6): 785-789. (in Chinese).
[21]	陈禹帆, 温蜜, 张凯, 等. 基于相似日匹配及TCN-Attention的短期光伏出力预测[J]. 电测与仪表2022, 59(10): 108-116. CHEN Y F, WEN M, ZHANG K, et al. Short-term photovoltaic output forecasting based on similar day matching and TCN-Attention. [J]. Electrial Measurement and Instrumentation, 2022, 59(10): 108-116. (in Chinese).
[22]	LI L L, WEN S Y, TSENG M L, et al. Renewable energy prediction: a novel short-term prediction model of photovoltaic output power[J]. Journal of Cleaner Production, 2019, 228: 359-375.
[23]	焦卫东, 杨蓓. MDAT-Net: 1种融合MSC和时空注意力的TCN航迹预测方法[J]. 北京航空航天大学学报, 2024. JIAO W D, YANG B. MDAT-Net: A TCN trajectory prediction method fusing MSC and spatio-temporal attention[J]. Journal of Beijing University of Aeronautics and Astronautics, 2024. (in Chinese).
[24]	张淑清, 李君, 姜安琦, 等. 基于FPA-VMD和BiLSTM神经网络的新型两阶段短期电力负荷预测[ J]. 电网技术, 2022, 46(8): 3269. ZHANG S Q, LI J, JIANG A Q, et al. A novel two-stage model based on FPA-VMD and BiLSTM neural network for shortterm power load forecasting[J]. Power System Technology, 2022, 46(8): 3269. (in Chinese)
[25]	刘清梅, 万明, 严利鑫, 等. 基于集合经验模态分解降噪和优化LSTM的道路交通事故预测[J]. 交通信息与安全, 2023, 41(5): 12-23. LIU Q M, WAN M, ZHANG Z Q, et al. A method for predicting traffic accidents based on an ensemble empirical mode decomposition and an optimized LSTM model[J]. Journal of Transport Information and Safety, 2023, 41(5): 12-23. (in Chinese).
[26]	付国忠, 杜华, 张志强, 等. 基于注意力机制和CNN-BiLSTM模型的滚动轴承剩余寿命预测[J]. 核动力工程, 2023, 44(增刊2): 33-38. FU G Z, DU H, ZHANG Z Q, et al. Remaining useful life prediction of rolling bearings based on attention mechanism and CNN-BiLS-TM[J]. Nuclear Power Engineering, 2023, 44 (S2): 33-38. (in Chinese).
[27]	葛少云, 朱林伟, 刘洪, 等. 基于动态交通仿真的高速公路电动汽车充电站规划[J]. 电工技术学报, 2018, 33(13): 2991-3001. GE S Y, ZHU L W, LIU H, et al. Optimal deployment of electric vehicle charging stations on the highway based on dynamic traffic simulation[J]. Transactions Of China Electrotec-hnical Society, 2018, 33(13): 2991-3001. (in Chinese)
[28]	刘珏, 穆云飞, 董晓红, 等. 考虑交通时滞的配电-交通耦合系统协同优化运行策略[J/OL]. 电力系统及其自动化学报[2024-10-08]. LIU Y, MU Y F, DONG X H, et al. A collaborative optimization strategy for power-traffic coupling system considering time delay[J/OL]. Proceedings of the CSU-EPSA[2024-10-08]. (in Chinese)
[29]	BAE S, KWASINSKI A. Spatial and temporal model of electric vehicle charging demand[J]. IEEE Transactions on Smart Grid, 2011, 3(1): 394-403.
[30]	葛少云, 冯亮, 刘洪, 等. 考虑电量分布及行驶里程的高速公路充电站规划[J]. 电力自动化设备, 2013, 33(7): 111-116. GE S Y, FENG L, LIU H, et al. Plug-and-play technology applied in microgrid and utilitygrid[J]. Electric Power Automation Equipment, 2013, 33(7): 111-116. (in Chinese)
[31]	ARIAS M B, KIM M, BAE S. Prediction of electric vehicle charging-power demand in realistic urban traffic networks[J]. Applied Energy, 2017, 195: 738-753.
[32]	龙雪梅, 杨军, 吴赋章, 等. 考虑路网-电网交互和用户心理的电动汽车充电负荷预测[J]. 电力系统自动化, 2020, 44 (14): 86-93. LONG X M, YANG J, WU F Z, et al. Prediction of electric vehicle charging load considering interaction between road network and power grid and user's psychology[J]. Automation of Electric Power Systems, 2020, 44(14): 86-93. (in Chinese).
[33]	MA H Y, WANG J H, ZHENG T, et al. A model for electric vehicle charging load forecasting based on simulated driving path[C]. 4^th Asia Conference on Power and Electrical Engineering, Hangzhou, China: ACPEE, 2019.
[34]	MEGAN. A Electric range for the nissan leaf & chevrolet volt in coldweather[OL]. (2013-12-16). [2015-05-20]. http://www.Fleetcarma.com/nissan-leaf-chevrolet-volt-cold-weaer-range-loss.
[35]	王海玲, 张美霞, 杨秀. 基于气温影响的电动汽车充电需求预测[J]. 电测与仪表, 2017, 54(23): 123-128. WANG H L, ZHANG M X, YANG X. Electric vehicle charging demand forecasting based on influence of weather and temperature[J]. Electrical Measurement & Instrumentation, 2017, 54(23): 123-128. (in Chinese)