• Journal of Inorganic Materials
  • Vol. 36, Issue 1, 61 (2021)
Yanran MENG1、2、3, Xinger WANG1、2、3、4, Jian YANG1、2、3、*, Han XU1、2、3, and Feng YUE1、2
Author Affiliations
  • 1School of Naval Architecture, Ocean and Civil Engineering, Shanghai Jiao Tong University, Shanghai 200240, China
  • 2State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, Shanghai 200240, China
  • 3Shanghai Key Laboratory for Digital Maintenance of Buildings and Infrastructure, School of Naval Architecture, Ocean and Civil Engineering, Shanghai Jiao Tong University, Shanghai 200240, China
  • 4Key Laboratory of Impact and Safety Engineering, Ministry of Education, Ningbo University, Ningbo 315211, China
  • show less
    DOI: 10.15541/jim20200187 Cite this Article
    Yanran MENG, Xinger WANG, Jian YANG, Han XU, Feng YUE. Research on Machine Learning Based Model for Predicting the Impact Status of Laminated Glass[J]. Journal of Inorganic Materials, 2021, 36(1): 61 Copy Citation Text show less
    Diagram of impact tests equipment
    1. Diagram of impact tests equipment
    ROC curves of single input parameter (a) Outer layer; (b) Inner layer
    2. ROC curves of single input parameter (a) Outer layer; (b) Inner layer
    ROC curves of integrated input parameters (a) Outer layer; (b) Inner layer
    3. ROC curves of integrated input parameters (a) Outer layer; (b) Inner layer
    The network structure of KELM
    4. The network structure of KELM
    WOA optimization process
    5. WOA optimization process
    WOA-KELM flow chart
    6. WOA-KELM flow chart
    WOA optimization process
    7. WOA optimization process
    WOA-KELM glass failure status prediction results
    8. WOA-KELM glass failure status prediction results
    SVM glass failure status prediction results
    9. SVM glass failure status prediction results
    LSSVM glass failure status prediction results
    10. LSSVM glass failure status prediction results
    IDMaterialMake-up (o/m/i)Dimensional of glass/mmSupport conditionQuantity
    P01FTG/PVB/FTG8/1.52/81000 × 1000Edge clamped12
    P02FTG/PVB/FTG8/1.52/81000 × 1000Bolted connection6
    P03FTG/PVB/FTG8/1.52/81500 × 1500Edge clamped3
    P04FTG/PVB/FTG8/1.52/81500 × 1500Bolted connection3
    P05HSG/PVB/HSG8/1.52/81000 × 1000Bolted connection3
    P06FTG/PVB/FTG8/0.76/81000 × 1000Bolted connection3
    P07FTG/PVB/FTG8/3.04/81000 × 1000Bolted connection3
    P08FTG/PVB/FTG8/1.52/101000 × 1000Bolted connection3
    P09FTG/PVB/FTG6/1.52/101000 × 1000Bolted connection3
    P10FTG/PVB/HSG8/1.52/81000 × 1000Bolted connection3
    P11HSG/PVB/FTG8/1.52/81000 × 1000Bolted connection3
    S01ANG/SGP/FTG8/3/81000 × 1000Edge clamped3
    S02FTG/SGP/ANG8/3/81000 × 1000Edge clamped3
    S03FTG/SGP/FTG8/3/81000 × 1000Bolted connection6
    S04FTG/SGP/FTG8/3/81500 × 1500Bolted connection3
    S05HSG/SGP/FTG8/3/81000 × 1000Bolted connection3
    S06FTG/SGP/FTG8/5/81000 × 1000Bolted connection3
    Table 1.

    夹层玻璃试件配置情况

    Material parameterFTGHSGANGPVBSGP
    Density/(kg·m-3)-2500.0-1100.00950.00
    Elasticity modulus/GPa-70.0-0.150.30
    Poission ratio-0.2-0.450.45
    Mean failure strength/MPa157.4104.042.0--
    Table 2.

    试验数据库材料参数

    nInput parameterOuter layer state AUC (Ao n)Inner layer state AUC (Ai n)
    1Thickness of interlayer0.6050.571
    2Thickness of outer layer0.5160.537
    3Thickness of inner layer0.5110.546
    4Type of interlayer0.5730.576
    5Type of outer layer0.5730.515
    6Type of inner layer0.5630.559
    7Side length0.5160.507
    8Boundary condition0.5410.573
    9Peak kinetic energy0.6540.675
    10State of outer layer0.8730.513
    11State of inner layer0.4720.714
    12Multiple input0.9160.842
    Table 3.

    破坏状态预测模型AUC值

    ItemDetailed settings
    Hardware
    CPUQuad-core intel core i7-4850HQ
    Frequency2.3 GHz
    RAM16GB 1600 MHz DDR3
    Hard drive500 GB
    Operating systemMacOS
    Table 4.

    仿真环境

    ModelComputing time/msTrainingaccuracy/%Testingaccuracy/%
    WOA-KELM10.6293.8088.45
    SVM367.8792.8087.00
    LSSVM65.2889.2085.56
    Table 5.

    玻璃破坏状态预测结果

    Yanran MENG, Xinger WANG, Jian YANG, Han XU, Feng YUE. Research on Machine Learning Based Model for Predicting the Impact Status of Laminated Glass[J]. Journal of Inorganic Materials, 2021, 36(1): 61
    Download Citation