Identification of Switching Operation Based on LSTM and MoE

doi:10.16182/j.issn1004731x.joss.21-0288

Abstract

Abstract:

Aiming at the individual differences of different personnel in the same operation and differences of the same person in the same operation at different times, a switching operation recognition model(MoE-LSTM) based on Mixture of experts model (MOE) and long short-term memory network(LSTM) is proposed. Based on MoE, LSTM is integrated to learn the feature distribution of different sources data. The acceleration data is collected to build the switching operation dataset and the action sequence is segmented and aligned based on sliding window. The action sequence is input to MoE-LSTM, and the temporal dependencies of different actions are independently learned by different LSTMs. The gating network selects the output of LSTM that classifies the current input better as the action recognition result. The result of model learning is that for action data from different time and space, different LSTMs perform better in a certain feature area than other LSTMs. The experiments on the switching operation dataset demonstrate superior performance of the proposed method compared to other existing action recognition algorithms.

Key words: switching operation, long short-term memory network(LSTM), mixture of experts model (MOE), neural network

CLC Number:

TP391.9

Xiaoqing Zhang, Wanfang Xiao, Yingjie Guo, Bowen Liu, Xuesen Han, Jingwei Ma, Gao Gao, He Huang, Shihong Xia. Identification of Switching Operation Based on LSTM and MoE[J]. Journal of System Simulation, 2022, 34(8): 1899-1907.

Figures/Tables 10

Fig. 1

Fig. 2

Table 1

Fig. 3

Fig. 4

Table 2

Table 3

Table 4

Fig. 6

References 17

1	杜洋. 2014~2018年电力安全事故总量统计与分析[EB/OL].[2021-03-14]. .
	Du Yang. Statistics and Analysis of the Total Number of Power Safety Accidents from 2014 to 2018[EB/OL]. [2021-03-14]. .
2	刘培贞, 贾玉祥, 夏时洪. 一种面向电力运维作业的LSTM动作识别方法[J]. 系统仿真学报, 2019, 31(12): 2837-2844.
	Liu Peizhen, Jia Yuxiang, Xia Shihong. An LSTM-Based Motion Recognition Method for Power Operation and Maintenance[J]. Journal of System Simulation, 2019, 31(12): 2837-2844.
3	张晓青, 贾玉祥, 夏时洪. 基于MoE的倒闸操作过程识别研究[C]//2020中国仿真大会. 北京: 中国仿真学会, 2020: 30-38.
	Zhang Xiaoqing, Jia Yuxiang, Xia Shihong. Research on Recognition of Switching Operation Based on MoE[C]//2020 China Simulation Conference. Beijing: China Simulation Federation, 2020: 30-38.
4	Dang L M, Min K, Wang H, et al. Sensor-Based and Vision-Based Human Activity Recognition: A Comprehensive Survey[J]. Pattern Recognition(S0031-3203), 2020, 108: 107561.
5	Gong J, Cui L, Xiao K, et al. MPD-Model: A Distributed Multipreference-Driven Data Fusion Model and its Application in a WSNs-Based Healthcare Monitoring System[J]. International Journal of Distributed Sensor Networks(S1550-1477), 2012, 8(12): 602358.
6	Incel O D. Analysis of Movement, Orientation and Rotation-Based Sensing for Phone Placement Recognition[J]. Sensors(S1424-8220), 2015, 15(10): 25474-25506.
7	徐川龙, 顾勤龙, 姚明海. 一种基于三维加速度传感器的人体行为识别方法[J]. 计算机系统应用, 2013, 22(6): 132-135.
	Xu Chuanlong, Gu Qinlong, Yao Minghai. Activity Recognition Method Based on Three-Dimensional Accelerometer[J]. Computer System Application, 2013, 22(6): 132-135.
8	Xiao Q, Song R. Action Recognition Based on Hierarchical Dynamic Bayesian Network[J]. Multimedia Tools and Applications(S1380-7501), 2018, 77(6): 6955-6968.
9	Sok P, Xiao T, Azeze Y, et al. Activity Recognition for Incomplete Spinal Cord Injury Subjects Using Hidden Markov Models[J]. IEEE Sensors Journal(S1530-437X), 2018, 18(15): 6369-6374.
10	Abidine B M, Fergani L, Fergani B, et al. The Joint Use of Sequence Features Combination and Modified Weighted SVM for Improving Daily Activity Recognition[J]. Pattern Analysis and Applications(S1433-7541), 2018, 21(1): 119-138.
11	Chen Z, Wang Y. Infrared-Ultrasonic Sensor Fusion for Support Vector Machine-Based Fall Detection[J]. Journal of Intelligent Material Systems and Structures(S1045-389X), 2018, 29(9): 2027-2039.
12	Murad A, Pyun J Y. Deep Recurrent Neural Networks for Human Activity Recognition[J]. Sensors(S1424-8220), 2017, 17(11): 2556.
13	Xia K, Huang J, Wang H. LSTM-CNN Architecture for Human Activity Recognition[J]. IEEE Access(S2169-3536), 2020, 8: 56855-56866.
14	Moya Rueda F, Grzeszick R, Fink G A, et al. Convolutional Neural Networks for Human Activity Recognition Using Body-Worn Sensors[J]. Informatics(S2227-9709), 2018, 5(2): 26.
15	Dong X, Yu Z, Cao W, et al. A Survey on Ensemble Learning[J]. Frontiers of Computer Science(S2095-2228), 2020, 14(2): 241-258.
16	Jacobs R A, Jordan M I, Nowlan S J, et al. Adaptive Mixtures of Local Experts[J]. Neural Computation(S0899-7667), 1991, 3(1): 79-87.
17	俞斌峰, 季海波. 稀疏贝叶斯混合专家模型及其在光谱数据标定中的应用[J]. 自动化学报, 2016, 42(4): 566-579.
	Yu Binfeng, Ji Haibo. Sparse Bayesian Mixture of Experts and Its Application to Spectral Multivariate Calibration[J]. Acta Automatica Sinica, 2016, 42(4): 566-579.

动作类型	原数据集^[3]	扩充数据集
合计	100 499	117 802
合闸接地	1 608	2 243
分闸接地	1 829	2 533
关柜门	13 254	14 258
开柜门	10 732	11 565
合闸	7 404	7 815
分闸	7 164	7 596
远方	6 197	6 552
就地	6 375	6 756
打开五防编码锁	10 670	11 438
关闭五防编码锁	7 933	8 423
摇动断路器	27 333	38 623

Expert网络数目	验证集准确率/%
Expert网络数目	改进的MoE模型	MoE-LSTM模型
2	89.89	94.62
3	91.36	94.77
4	91.61	95.46
5	90.69	95.75
6	91.04	95.50
7	91.69	95.44
8	92.86	95.68
9	92.06	95.58

动作类别	精确率	召回率	F₁分数
关柜门	95.26	93.50	94.37
关闭五防编码锁	94.58	92.58	93.57
分闸	87.59	99.22	93.04
分闸接地	92.51	88.20	90.30
合闸	100	100	100
合闸接地	78.23	77.60	77.91
就地	74.77	91.95	82.47
远方	99.00	100	99.50
开柜门	91.05	93.90	92.45
打开五防编码锁	98.50	94.13	96.27
摇动断路器	93.89	96.51	95.18

性能指标	整体准确率	平均召回率	平均精确率	平均F₁值
MoE-LSTM	93.49	93.42	91.40	92.28
改进的MoE	93.11	92.09	90.44	91.08
LSTM	91.45	90.88	88.82	89.70
KNN	86.74	86.07	83.06	84.01
RF	91.02	92.24	89.33	90.46
SVM	92.49	91.19	89.69	89.79

[1]	Yecai Guo, Qingwei Wang. Modulation Recognition Algorithm Based on Truncated Migration and Parallel ResNet [J]. Journal of System Simulation, 2022, 34(9): 2009-2018.
[2]	Weiguo Tong, Shichao Zeng, Lifeng Zhang, Zhe Hou, Jiayue Guo. Electrical Resistance Tomography and Flow Pattern Identification Method Based on Deep Residual Neural Network [J]. Journal of System Simulation, 2022, 34(9): 2028-2036.
[3]	Lifeng Zhang, Yu Miao. A High Resolution Reconstruction Method of Temperature Distribution in Acoustic Tomography [J]. Journal of System Simulation, 2022, 34(9): 2065-2073.
[4]	Xinghua Liu, Chen Geng, Shenghan Xie, Jiaqiang Tian, Hui Cao. Day Ahead Thermal-photovoltaic Economic Dispatch Considering Uncertainty of Photovoltaic Power Generation [J]. Journal of System Simulation, 2022, 34(8): 1874-1884.
[5]	Sen Zhang, Mengyan Zhang, Jingping Shao, Jiexin Pu. Multi-UAVs 3D Path Planning Method Based on Random Strategy Search [J]. Journal of System Simulation, 2022, 34(6): 1286-1295.
[6]	Lifeng Zhang, Huiren Wang. Image Reconstruction of Electrical Capacitance Tomography Based on Convolutional Neural Network and Finite Element Simulation [J]. Journal of System Simulation, 2022, 34(4): 712-718.
[7]	Sijin Zhou, Dicheng Chen, Geng Tu, Dazhi Jiang. Multi-modality Affective Computing Model Based on Personality and Memory Mechanism [J]. Journal of System Simulation, 2022, 34(4): 745-758.
[8]	Xiaoan Sun, Xiaoli Luan, Fei Liu. Electronic Solid Waste Prediction Based on Intelligent Optimization Grey Model [J]. Journal of System Simulation, 2022, 34(3): 536-542.
[9]	Juan Wei, Lei You, Yangyong Guo, Zhihai Tang. Multi-floor Evacuation Model Based on Wavelet Neural Network [J]. Journal of System Simulation, 2022, 34(2): 269-277.
[10]	Bangqian Ao, Sha Yang, Jinqing Linghu, zhenhuan Ye. Design of Fatigue Driving Detection System Based on Cascaded Neural Network [J]. Journal of System Simulation, 2022, 34(2): 323-333.
[11]	Wei Tengfei, Pan Tinglong. Short-term Power Load Forecasting Based on LSTM Neural Network Optimized by Improved PSO [J]. Journal of System Simulation, 2021, 33(8): 1866-1874.
[12]	Zhao Ying, Lu Yao, Zhang Jian, Liang Qidi, Long Wei. Multi-view Human Action Recognition Based on Deep Neural Network [J]. Journal of System Simulation, 2021, 33(5): 1019-1030.
[13]	Liu Xiaolin, Wu Jingyi. Design and Simulation for Compensatory Controller of Aircraft Rudder Electro-hydraulic Loading System [J]. Journal of System Simulation, 2021, 33(5): 1031-1038.
[14]	Yu Guangyu, Dong Xueping, Wang Xiangmin, Gan Min. Decomposition Furnace Outlet Temperature Prediction Based on ElasticNet and LSTM [J]. Journal of System Simulation, 2021, 33(5): 1078-1085.
[15]	Fang Qun, Chen Xukai, He Xin, Zhu Yujun, Liu Yiyang, Fang Yangyang, Li Heju. Study on Relay Selection Algorithm Based on SWIPT Wireless Cooperative Network [J]. Journal of System Simulation, 2021, 33(4): 801-808.