融合多头注意力机制的货车载重估计模型

doi:10.13229/j.cnki.jdxbgxb.20230297

摘要/Abstract

摘要：

为了提高实时货车载重估计的便捷性和精准度，辅助开展大范围低等级公路网的货车运输过程中载重的实时监管，本文从货车动静态信息的交互效应出发，提出了一种融合多头注意力机制的货车载重估计模型（Mix-MAN）。首先，在模型中引入了多头注意力机制，增强网络对运动学时序特征的提取能力；其次，利用堆叠自编码器捕获货车的静态特征；最后，设计了一种特征融合结构，融合提取动态特征和静态特征，建立输入特征与货车载重之间的非线性映射关系，得到货车载重估计结果。试验结果表明：与未考虑货车静态信息的MAN模型对比，Mix-MAN的平均绝对值误差减小了6%，均方根误差减小了5%，平均绝对百分比误差减小了0.5%。本文模型可为我国公路货物运输监管、道路养护等方面提供技术支持。

关键词: 交通运输系统工程, 载重估计, 多头注意力, 特征融合, 堆叠自编码器

Abstract:

In order to improve the real-time truck load estimation of convenience and accuracy， and to help to develop real-time monitoring for truck loads in a large-scale low-grade highway networks， this paper proposed a weight estimation model（Mix-MAN） for trucks integrating the multi-head attention mechanism based on the interaction effect of dynamic and static information of trucks. Firstly， multi-head attention was introduced into the model to enhance the network's ability to extract kinematic time series features； secondly， a stacked auto-encoder was used to capture the static features of trucks； finally， a feature fusion structure was designed to extract dynamic features and static features， establish the nonlinear mapping relationship between input features and weight estimation， and then to obtain the final weight estimation result of trucks. The experimental results show that compared with the MAN model without considering static information of truck， the mean absolute value error of Mix-MAN is reduced by 6%， root mean square error is reduced by 5%， mean absolute percentage error is reduced by 0.5%. The model in this paper can provide technical support for the cargo transport supervision of highway and road maintenance.

Key words: engineering of communication and transportation, load estimation, multi-head attention, feature fusion, stacked auto-encoders

中图分类号:

U495

顾明臣,熊慧媛,刘增军,罗清玉,刘宏. 融合多头注意力机制的货车载重估计模型[J]. 吉林大学学报(工学版), 2024, 54(10): 2771-2780.

Ming-chen GU,Hui-yuan XIONG,Zeng-jun LIU,Qing-yu LUO,Hong LIU. Weight estimation model for trucks integrating multi-head attention mechanism[J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(10): 2771-2780.

图/表 10

图1

图2

图3

图4

表1

图5

表2

图6

表3

图7

参考文献 23

1	徐银,刘星材,刘海旭, 等.公路超限超载运输治理的立法对策研究[J].综合运输, 2023, 45(4): 33-37.
	Xu Yin, Liu Xing-cai, Liu Hai-xu, et al. Research on legislative countermeasures for governing over-limited and over-load transportation of highway[J]. China Transportation Review, 2023, 45(4): 33-37.
2	陈一锴, 王富超, 王凯, 等. 基于二元 Logistic 的公路货运超载关键影响因素识别[J]. 重庆交通大学学报: 自然科学版, 2018, 37(5): 92-96.
	Chen Yi-kai, Wang Fu-chao, Wang Kai, et al. Identification of key factors affecting highway freight overloading based on binary logistic regression[J]. Journal of Chongqing Jiaotong University(Natural Science Edition), 2018, 37(5): 92-96.
3	李彬,肖润谋,闫晟煜,等.中国高速公路运输态势[J].交通运输工程学报,2020, 20(4): 184-193.
	Li Bin, Xiao Run-mou, Yan Cheng-yu, et al. Transportation trend of chinese expressway[J]. Journal of Traffic and Transportation Engineering, 2020,20(4):184-193.
4	祝志文,黄炎,向泽.货运繁重公路的车辆荷载谱和疲劳车辆模型[J].交通运输工程学报,2017,17(3):13-24.
	Zhu Zhi-wen, Huang Yan, Xiang Ze. Vehicle loading spectrum and fatigue truck models of heavy cargo highway[J]. Journal of Traffic and Transportation Engineering, 2017, 17(3): 13-24.
5	韩玫枝.车载称重系统的设计与应用[D].南京: 南京理工大学自动化学院,2017.
	Han Mei-zhi. Design and application of vehicle weighing system[D].Nanjing:School of Automation, Nanjing University of Science and Technology,2017.
6	韩春阳,宿旸,裴欣,等.基于深度学习的货车载重实时估计方法研究[J].中国公路学报, 2022, 35(3): 295-306.
	Han Chun-yang, Su Yang, Pei Xin, et al. Real-time weight estimation for trucks based on deep learning method[J]. China Journal of Highway and Transport, 2022, 35(3): 295-306.
7	Li B Y, Zhang J W, Du H P, et al. Two-layer structure based adaptive estimation for vehicle mass and road slope under longitudinal motion[J]. Measurement, 2017, 95: 439-455.
8	张从力,史记征,陈增江.一种车载静态称重系统设计[J].传感器与微系统,2013,32(1): 99-101.
	Zhang Cong-li, Shi Ji-zheng, Chen Zeng-jiang.Design of an onboard static weighing system[J]. Transducer and Microsystem Technologies, 2013, 32(1): 99-101.
9	Mohan S, Kumar P. An empirical study on low-cost, portable vehicle's weight estimat-ion solution using smartphone's acceleration d-ata for developing countries[C]∥ Proceedings of the 7th International Conference on VehicleTechnology and Intelligent Transport Systems, Berlin, Germany, 2021: 44-55.
10	Sun Y, Li L, Yan B J, et al. A hybrid algorithm combining EKF and RLS in synchronous estimation of road grade and vehicle' mass for a hybrid electric bus[J]. Mechanical Systems and Signal Processing, 2016, 68: 416-430.
11	Nguyen P X, Akiyama T, Ohashi H, et al. Vehicle´s weight estimation using smartphone´s acceleration data to control overloading[J]. International Journal of Intelligent Transportation Systems Research, 2018, 16(3): 151-162.
12	孔烜,张杰,王腾义,等.基于图像识别轮胎变形的非接触式车辆称重方法[J].中国公路学报,2022,35(8):186-193.
	Kong Xuan, Zhang Jie, Wang Teng-yi, et al. Non-contact vehicle weighing method based on tire deformation using image recognition[J]. China Journal of Highway and Transport, 2022, 35(8): 186-193.
13	Barati K, Shen X S, Li N, et al. Automatic mass estimation of construction vehicles by modeling operational and engine data[J]. Journal of Construction Engineering and Management, 2022, 148(3): No.4021208.
14	Korayem A H, Khajepour A, Fidan B. Trailer mass estimation using system model-based and machine learning approaches[J]. IEEE Transactions on Vehicular Technology, 2020, 69(11): 12536-12546.
15	Jia Z X, Fu K Y, Lin M X. Tire-pavement contact-aware weight estimation formulti-sensor WIM systems[J]. Sensors, 2019, 19(9): 2027-2040.
16	Yu Y, Cai C S, Liu Y M. Probabil-istic vehicle weight estimation using physics constrained generative adversarial network[J]. Computer-Aided Civil and Infrastructure Engineering, 2021, 36(6): 781-799.
17	郝彬,毕玉峰,李金海.济菏高速公路货车载重特征分析[J].公路,2015,60(4): 174-179.
	Hao Bin, Bi Yu-feng, Li Jin-hai. Analysis of truck load characteristics on Ji-he highway[J].Highway, 2015, 60(4): 174-179.
18	周大可,张超,杨欣.基于多尺度特征融合及双重注意力机制的自监督三维人脸重建[J].吉林大学学报:工学版, 2022, 52(10): 2428-2437.
	Zhou Da-ke, Zhang Chao, Yang Xin. Self-supervised 3D face reconstruction based on multi-scale feature fusion and dual attention mechanism[J]. Journal of Jilin University (Engineering and Technology Edition), 2022, 52(10): 2428-2437.
19	Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[C]∥ Proceedings of the International Conference on Neural Information Processing Systems, Long Beach, USA, 2017, 30: 6000-6010.
20	陈巧红,李妃玉,孙麒.基于LSTM与衰减自注意力的答案选择模型[J].浙江大学学报: 工学版,2022,56(12): 2436-2444.
	Chen Qiao-hong, Li Fei-yu, Sun Qi. Answer selection model based on LSTM and decay s-elf-attention[J]. Journal of Zhejiang University (Engineering Science),2022, 56(12): 2436-2444.
21	He K M, Zhang X Y, Ren S Q, et al. Deep residual learning for image recognition[C]∥ Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, USA, 2016: 770-778.
22	Huang G, Liu Z, van der Maaten L, et al. Densely connected convolutional networks[C]∥ Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA, 2017: 4700-4708.
23	张彭,霍恩泽,王英平,等.自动化公路货物运输量统计方法[J].中国公路学报,2021, 34(12): 302-312.
	Zhang Peng, Huo En-ze, Wang Ying-ping, et al. Automated statistical methods for highway freight volume[J]. China Journal of Highway and Transport, 2021, 34(12): 302-312.

相关文章 15

[1]	张娜,陈峰,王剑坡,朱亚迪. 基于时空序列相似性的城轨乘客出行模式识别[J]. 吉林大学学报(工学版), 2024, 54(9): 2588-2599.
[2]	周锡浈,宫贺,李敦敦,季彦婕,严杰. 建成环境对路内停车泊位使用率的非线性影响模型[J]. 吉林大学学报(工学版), 2024, 54(9): 2520-2530.
[3]	严利鑫,曾涛,贺宜,郭军华,胡鑫辉. 共驾型智能车辆人机接管行为序列编码与解析[J]. 吉林大学学报(工学版), 2024, 54(9): 2547-2556.
[4]	曲昭伟,李霖,陈永恒,吴场建. 长区间掉头车辆特性分析及其安全评价[J]. 吉林大学学报(工学版), 2024, 54(8): 2206-2213.
[5]	何永明,权聪,魏堃,冯佳,万亚楠,陈世升. 超高速公路车路协同路侧单元感知融合方法[J]. 吉林大学学报(工学版), 2024, 54(7): 1923-1934.
[6]	程国柱,盛林,王浩宇,冯天军. 考虑右转车二次冲突的信号交叉口行人过街安全评价方法[J]. 吉林大学学报(工学版), 2024, 54(7): 1903-1912.
[7]	秦雅琴,钱正富,谢济铭. 协同换道避障模型和轨迹数据驱动的车辆协同避障策略[J]. 吉林大学学报(工学版), 2024, 54(5): 1311-1322.
[8]	张明业,杨敏,黎彧,黄世玉,李清韵. 考虑有序充电策略的多车型电动公交调度优化[J]. 吉林大学学报(工学版), 2024, 54(5): 1293-1301.
[9]	马潇驰,陆建. 基于基因表达式编程的高架道路事故实时预测[J]. 吉林大学学报(工学版), 2024, 54(3): 719-726.
[10]	严利鑫,冯进培,郭军华,龚毅轲. 不同险态情景下共驾型智能车辆接管行为特征分析[J]. 吉林大学学报(工学版), 2024, 54(3): 683-691.
[11]	曲大义,张可琨,顾原,王韬,宋慧,戴守晨. 自动驾驶车辆换道决策行为分析及分子动力学建模[J]. 吉林大学学报(工学版), 2024, 54(3): 700-710.
[12]	涂辉招,王万锦,乔鹏,郭静秋,鹿畅,吴海飞. 自动驾驶卡车路测安全员接管干预行为解析[J]. 吉林大学学报(工学版), 2024, 54(3): 727-740.
[13]	王春华,李恩泽,肖敏. 多特征融合和孪生注意力网络的高分辨率遥感图像目标检测[J]. 吉林大学学报(工学版), 2024, 54(1): 240-250.
[14]	张健,李青扬,李丹,姜夏,雷艳红,季亚平. 基于深度强化学习的自动驾驶车辆专用道汇入引导[J]. 吉林大学学报(工学版), 2023, 53(9): 2508-2518.
[15]	郑植,袁佩,金轩慧,魏思斯,耿波. 桥墩复合材料柔性防撞护舷试验[J]. 吉林大学学报(工学版), 2023, 53(9): 2581-2590.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

数据集	数据类型	主要字段
高速公路收费入口-治超站称重数据	车辆称重数据	车牌、入口时间、收费车型、车长、轴数、核定质量、车辆总重等
交调轴载站称重数据	车辆称重数据	车牌、经过时间、轴型、轴数、轴重、车辆总重等
货车轨迹数据	驾驶轨迹数据	车辆ID、时间、速度（km/h）、车辆转向角等

类别	变量名称	变量类型	取值范围
车辆运动学时序特征	速度	连续变量	0~120 km/h
车辆运动学时序特征	加速度	连续变量	—
车辆静态特征	收费车型	离散变量	11~16
	轴数	离散变量	2~6
	车长	连续变量	—
	入口时间	连续变量	2021.8.1 00：00 ~2021.9.1 00：00

模型	评价指标
模型	MAE	RMSE	MAPE/%
GRU	1.51	1.56	3.6
Mix-GRU	1.52	1.49	3.2
LSTM	1.48	1.52	3.5
Mix-LSTM	1.47	1.50	3.2
MAN	1.42	1.50	3.1
Mix-MAN	1.34	1.42	2.6