To improve the accuracy of three-dimensional quantitative evaluation for corrosion thinning, pitting, and boundary-discontinuity defects in steel components of forestry equipment under service conditions involving high humidity, mud-water erosion, and cyclic temperature variations, an improved discretized magnetic dipole forward model and a two-stage inversion-based quantification framework are proposed. In the forward modeling stage, a second-order response-driven adaptive mesh refinement strategy and a three-point surface-element quadrature scheme are introduced to enhance the characterization of pit edges and discontinuity zones while controlling the number of nodes. In the inversion stage, principal component analysis (PCA) is first employed to reduce the dimensionality of the defect description and constrain the search space, while the Newton–Raphson optimizer is used to optimize the hyperparameters of the radial basis function (RBF) model to obtain robust initial estimates. In the second stage, the covariance matrix adaptation evolution strategy is adopted to further refine the solution in the low-dimensional space by coupling it with the forward model. Simulation results demonstrate that the proposed forward model achieves accuracy close to that of the finite element model while significantly reducing computational time, and that the inversion framework provides more stable reconstructions for defects with diverse morphologies. The proposed method can be embedded into intelligent forestry condition monitoring and predictive maintenance systems, thereby providing a quantitative basis for risk assessment, maintenance prioritization, and safety management of forestry equipment.
在采样平面计算模型输出与FEM数据之间的均方根误差(root mean square error,),并记录细化后的节点数与单次前向耗时从而量化改进模型的精度与效率。在扫描试验中,FEM仅在相同条件下预先计算一次,得到有限元计算结果作为参考,随后在扫描中复用该参考场进行误差评估。(式中记为RMSE)定义公式为
YangJin, YanZhe, LeiYongjie,et al.Construction of an integrated sky-air-ground intelligent forestry machinery system in the context of low-altitude economy[J].Forestry Machinery & Woodworking Equipment,2025,53(10):61-71,80.
YanShiyun, ZhangHuibin, JiHaokai,et al.Research on wood defect detection based on improved YOLOv8 model[J].Forest Engineering,2025,41(4):750-760.
[5]
MaktoubianJ, TaskhiriM S, TurnerP.Intelligent predictive maintenance (IPdM) in forestry:a review of challenges and opportunities[J].Forests,2021,12(11):1495.
YangShuo, KanJiangming, ZhaoXixuan.Forest visibility prediction method based on APO-RBF neural network[J].Journal of Forestry Engineering,2026,11(2):100-107.
[8]
CuiC, ZhangQ, ZhangD,et al.Monitoring and detection of steel bridge diseases:a review[J].Journal of Traffic and Transportation Engineering (English Edition),2024,11(2):188-208.
HuXiaoxuan, JiaYanmin, XuJunming,et al.Study on flexural behavior of connecting steel-concrete composite beams based on PBL[J].Forest Engineering,2025,41(5):1073-1081.
HanWenhua, YuanYe, ZhangYuxiang,et al.Research on finite element forward model for magnetic flux leakage testing of arbitrary shape defects[J].Fire Control & Command Control,2018,43(1):80-84.
WangYufei, HanWenhua.3D irregular defect reconstruction method based on coordinate offset magnetic dipole-Newton-Raphson method[J].Chinese Journal of Scientific Instrument,2024,45(6):328-336.
[15]
ZhongY, WangG, YuX,et al.Defect size quantification for pipeline dynamic MFL detection via 3-D dynamic magnetic dipole forward model[J].IEEE Transactions on Instrumentation and Measurement,2025,74:6001910.
[16]
WangY, GuoJ, HuT,et al.Machine learning-based dipole modeling of 3-D magnetic flux leakage caused by outer-surface linear cracks for pipeline in-line inspection[J].IEEE Transactions on Instrumentation and Measurement,2024,73:2531614.
[17]
LiS, ZhangJ, ZhangX,et al.A 3D reconstruction method for complex defects based on MFD and multidirectional MFL[C]//2024 IEEE International Instrumentation and Measurement Technology Conference (I2MTC).May 20-23,2024,Glasgow,United Kingdom.IEEE,2024.
MiaoLiheng, PanFeng, PengLisha,et al.Iterative optimization method of defect depth profile based on depth characteristics of MFL signal[J].Proceedings of the Chinese Society for Electrical Engineering,2022,42(8):3077-3086.
[20]
HouD, LuS, YiG,et al.A target-focusing optimization method for 3-D profile reconstruction of defects using MFL measurements[J].IEEE Transactions on Instrumentation and Measurement,2023,72:1-11.
ShiQing, ZhangGuoshan, BeiGang,et al.Detection and quantification of pipeline magnetic flux leakage defects based on random forest[J].Journal of Electronic Measurement and Instrumentation,2025,39(3):199-207.
YuZhiqi, LiuHaoyuan, HeLuyao,et al.Three dimensional reconstruction method of pipeline defects based on CNN-Transformer[J].Modern Information Technology,2024,8(20):48-55.
ZhaoDongsheng, YangLijian, GengHao,et al.MFL data augmentation method for pipeline defect detection[J].Instrument Technique and Sensor,2023(12):111-116.
[27]
ZhangS, LiH, ZhaoC.Defect-depth-field algorithm for simulating magnetic flux leakage signals based on discrete magnetic dipole model[J].NDT & E International,2023,139:102939.
ZhuHongxiu, LiuHuan, LiHongyuan,et al.Quantitative analysis method for pipeline defects based on optimized RBF neural network[J].Instrument Technique and Sensor,2016(2):83-86.
WangXiaohong, XinShouying, ZhangWei,et al.Study on tree species identification of planted forests based on PCA-BO-CNN model[J].Forest Engineering,2025,41(2):298-311.
[32]
SowmyaR, PremkumarM, JangirP.Newton-raphson-based optimizer:a new population-based metaheuristic algorithm for continuous optimization problems[J].Engineering Applications of Artificial Intelligence,2024,128:107532.