Table of Content

25 March 2025, Volume 36 Issue 03

Previous Issue   

Visual Detection of Subsurface Corrosions in Ferromagnetic Metal Plates Using Pulsed Eddy Current Based on Dual-sensor Differential Mechanism

WANG Jin1, LI Yong1, SU Bingjie1, GAO Wenlong1, XIANG Yi1, 2, LIU Zhengshuai1, CHEN Zhenmao1

2025, 36(03):  381-390.  DOI: 10.3969/j.issn.1004-132X.2025.03.001

Asbtract ( )   PDF (17686KB) ( )  

References | Related Articles | Metrics

A pulse eddy current visual detection method was proposed based on a dual-sensor differential mechanism for subsurface corrosions in ferromagnetic metal plates. In this method, a dual-sensor differential probe was used to pick up the pulsed eddy current testing signals, and the slope of the logarithmic curve along the falling edge of the testing signals and the peak value of the normalized differential signals were extracted as signal features, which were used for visual detection of defects of different sizes. Through simulation and experimental research, the correlation laws between the proposed signal features and the sizes of defects were established, and it is verified that the dual-sensor differential probe has the advantages of reducing noise interference and improving the detection sensitivity compared with traditional single-sensor probes. In addition, a method was proposed to fuse the two signal features, and the results indicate that the defect images using fused signal features have a higher image signal-to-noise ratio. The research method proposed herein provides an effective and reliable solution for the visual detection of subsurface corrosions in ferromagnetic metal plates.

Experimental Study of Machining Micro Holes in Cf/SiC Composite Materials Using Brazed Diamond Milling Head

HONG Mengjie, CHEN Zhuo, ZHOU Jiayi, WANG Haixu, WANG Jianyu, HUANG Guoqin

2025, 36(03):  391-397.  DOI: 10.3969/j.issn.1004-132X.2025.03.002

Asbtract ( )   PDF (14837KB) ( )  

References | Related Articles | Metrics

Cf/SiC ceramic matrix composites were typical high hardness and brittleness ones difficult to machine that combined anisotropy and heterogeneous structures. It was difficult to machine micro holes with a pore size of less than 5 mm, and always was a hot research topic in the industry. Three types of small diameter diamond milling heads was used for hole machining experiments, the preparation method of milling heads and the influences of end geometry on drilling forces, service life, micro hole defects, and tool surface quality were studied. The results show that the micro hole defects produced by the core brazing milling head are the least and the service life is 3.6 times that of the flat bottomed electroplating milling head and 2 times that of the flat bottomed brazing milling head,the core taking brazing milling head has an end face core taking structure, which may reduce drilling force and significantly improve the service life of the tool and the surface quality of the machined workpieces effectively.

Rapid Robust Design of Multibody Weapon Systems Considering Uncertainty

MA Yuanzhuo1, LI Chenxu1, WANG Youyu2, ZHANG Zhiyong3, DING Anai4, LI Hongshuang4

2025, 36(03):  398-406.  DOI: 10.3969/j.issn.1004-132X.2025.03.003

Asbtract ( )   PDF (4954KB) ( )  

References | Related Articles | Metrics

The robust design of multibody weapon systems considering uncertain parameters had the problems of low efficiency in multibody system dynamics modeling, analysis and robust design solutions. Therefore, based on the probability theory and mathematical statistics, the uncertainties were formulated and characterized within multibody weapon systems, including quick identification and representation of random parameters, and the efficient simulation of load stochastic processes. Using naval artillery as an illustrative example, the multibody system transfer matrix method was quickly employed to ascertain the vibration quality of the artillery muzzles. The results were compared with ADAMS simulation outcomes. By harnessing PCE-Kriging(PCK) surrogate model, multi-objective subset simulation optimization method, and a single-cycle robust solution strategy, the original double-cycle robust design optimization problem was transformed into a single-cycle optimization problem and solved efficiently. Based on uncertain input, the dynamics model of multibody system transfer matrix method of a naval artillery and the PCK surrogate model, the robustness of the naval artillery model was finally designed using double-cycle and single-cycle robust design optimization methods. The performance of single-cycle robust design method was verified by comparing the optimization results. 

Tip-trajectory Following Algorithm for Snake-arm Robots Based on Dynamics Denoising Model

ZHANG Zhigang1, JIN Yongli1, WANG Dongyin1, FU Zhijun1, QIN Guodong2

2025, 36(03):  407-413.  DOI: 10.3969/j.issn.1004-132X.2025.03.004

Asbtract ( )   PDF (4037KB) ( )  

References | Related Articles | Metrics

 Based on dynamics model smoothing formulation of equivalent multi-link system, a tip-trajectory following algorithm was proposed for motion planning of hyper-redundant SARs in narrow and confined spaces. Regarding SAR as a multi-link system of joint points constrained to move on collision-free trajectory curve of tip, and the dynamics equation of rigid-flexible coupling system was derived using virtual work principle. The length conditions of robot arm segments were guaranteed by axial stiffness of connecting rods, and the influences of high-frequency components were filtered out from the model level by employing model denoising method, so that the conventional explicit algorithm might be used to calculate the solution of equation of equivalent multi-link system efficiently. The proposed algorithm may specify the velocity law of tip or base of snake-arms, and realize the motion planning of SAR along tip-trajectory. Due to the strictly limited joint points of SAR moving on feasible trajectory curve, obstacle avoidance motion planning may be achieved in narrow and deep cavity environments. After solving typical numerical examples in planar and spatial cases, the results show that the proposed algorithm has sufficient computational accuracy and efficiency for the motion planning of SARs.

Research on Smoothness of Mode Switch of Independent Metering Systems Considering Time-delay Factors

CHEN Junxiang1, 2, JIANG Hongda1, 2, KONG Xiangdong1, 2, AI Chao1, 2

2025, 36(03):  414-425.  DOI: 10.3969/j.issn.1004-132X.2025.03.005

Asbtract ( )   PDF (12319KB) ( )  

References | Related Articles | Metrics

The independent valve of the load port realized the decoupling control of the inlet and outlet orifices in the hydraulic systems, which might support a variety of control modes. However, due to the dynamic response and controller delay during switching, the system mode was easily asynchronous with the controller mode, causing pressure shock and instability problems. A stable switching strategy was proposed based on independent metering system. The strategy introduced the idea of "slow switching", considering system time-delay and combining Lyapunov stability theory, a boundary value solution method was proposed for the dynamic residence time of an asynchronous switching system based on mode correlation average residence time. The method obtains a shorter residence time than that of the traditional average residence time method. The stability control problem was solved under asynchronous switching of linear switching systems, which ensures the stability of switching instantaneously, and reduces the influences of residence time on system responses. Based on the excavation actions of bucket pole of rescue vehicles, the control strategy simulation verifies that under the conditions of time-delay, the pressure shock of hydraulic systems before and after mode switch is reduced by more than 80%, and the amplitude of the actuator speed oscillation is reduced by nearly 20%, which effectively improves the operation stability of the actuator of the rescue vehicles. 

High Energy Density Hydraulic Energy Storage Method Based on Composite Principle of Gas Liquid Dissolution and Gas Compression

WANG Weiping, ZHOU Xinyi, LU Shun

2025, 36(03):  426-434，443.  DOI: 10.3969/j.issn.1004-132X.2025.03.006

Asbtract ( )   PDF (9849KB) ( )  

References | Related Articles | Metrics

To address the issues of low energy density in traditional hydraulic accumulators, a high-energy-density hydraulic energy storage method was proposed based on the combined principle of gas-liquid dissolution and gas compression. This method involved adding a quantitative amount of liquid solvent to the gas section of the hydraulic accumulator, enabling the energy storage processes to encompass both of gas compression and gas-liquid dissolution. Through gas-liquid dissolution, a portion of the energy was stored in the form of internal energy within the liquid phase, thereby reducing the increase in internal energy of the gas. The energy density expression for the composite energy storage method was derived and comparatively analyzed. Numerical relationship between the gas polytropic index and factors such as gas compression rate was established. Finally, a combination of carbon dioxide and water was selected as the energy storage medium for simulation and experimental validation. Both of simulation and experimental results indicate that, under conditions of a pre-charged gas pressure of 5 MPa, an energy storage duration of 5 s, and a water injection volume of 1 L, the energy density improvement rate reaches 9%.

Experimental Optimization of Microwave Sintering Parameters for Si₃N₄-based Ceramic Materials

XU Weiwei1, ZHANG Zuoxuan1, ZHU Songqing1, YIN Zengbin2

2025, 36(03):  435-443.  DOI: 10.3969/j.issn.1004-132X.2025.03.007

Asbtract ( )   PDF (10978KB) ( )  

References | Related Articles | Metrics

Microwave sintering technology was widely used in preparation of ceramic materials. However, the silicon nitride based ceramic materials sintered by microwave had problems such as poor performance consistency and large power fluctuations during heating processes. Therefore, the distribution characteristics of electric field and temperature field were compared in microwave sintering chambers with or without loads and different sample placement methods according to simulation. The influences of microwave power on the distribution of electric field and temperature field were analyzed, and the principle of microwave sintering power selection was proposed. The influences of heating rate and matching power on mechanics properties of ceramics in different temperature ranges were compared through experimental results. The results show that adding loads and arranging samples tightly in microwave field are beneficial for improving the uniformity of the electric field. The experimental results indicate that the Si3N4-based ceramic materials with the best mechanics properties may be obtained by selecting heating rates of 80 ℃/min, 50 ℃/min and 25 ℃/min at 0~800 ℃, 800~1400 ℃ and 1400~1650 ℃ respectively, and matching powers of 1600 W, 1800 W and 2200 W are selected for each temperature range. The Vickers hardness reaches(18.278±0.233)GPa, and the fracture toughness reaches(8.588±0.165)MPa·m1/2. Compared with the uniform heating processes, the hardness increases by 14.8%, the toughness increases by 22.5%, and the consistency of ceramic material properties may be improved effectively with the regulation strategy herein.

Influences of Size of “Reverse L-Container” on Effect of Vertical Vibration Superfinishing Rolling of Cylindrical Rollers

LIANG Zhenhua1, 3, LI Xiuhong1, 3, WANG Xingfu1, 3, LI Wenhui2, 3, YANG Shengqiang1, 3, LIANG Zhiqiang1, 3

2025, 36(03):  444-454.  DOI: 10.3969/j.issn.1004-132X.2025.03.008

Asbtract ( )   PDF (10253KB) ( )  

References | Related Articles | Metrics

Through simulation and experiments, the influences of container sizes on the machining effectiveness of cylindrical rollers in vertical vibration superfinishing rolling processes were investigated. The relation among the dimensional parameters and the movement characteristics of rollers and the mechanical behavior of surface particles was analyzed with the angle α and size d of “reverse L-container” as variables. With the increase of α, the height of the roller active region decreased gradually, and the deviation degree between the axis and the horizontal plane decreases first and then increased. The normal contact force of the outer diameter surface increased first and then decreased, while the tangential relative velocity decreased gradually. The normal contact force decreased first and then increased, while the tangential relative velocity increased first and then decreased. With the increase of d, the height of the rollers active region gradually increased, the deviation degree between the axis and the horizontal plane first decreased and then increased, the normal contact force of the outer diameter surface first increased and then decreased, the normal contact force of the end face first decreased and then increased, and the tangential relative velocity of the outer diameter surface and the end face increased gradually. The experimental results agree with the simulation results. The better container size:α=10°, d=50 mm, after processing, the outer diameter surface roughness value is reduced from 0.105 μm to 0.036 μm, the end surface roughness value is reduced from 0.150 μm to 0.057 μm, the surface profile is flat relatively, and the scratches are removed basically.


Research on Bending Characteristics of Cannula Flexible Needles Based on Dimensional Analysis and Equivalent Stiffness Method

ZHAO Yanjiang1, 2, ZHANG He1, QIAN Cheng1, ZHANG Yongde1

2025, 36(03):  455-462.  DOI: 10.3969/j.issn.1004-132X.2025.03.009

Asbtract ( )   PDF (5734KB) ( )  

References | Related Articles | Metrics

Aiming at the problems of precise insertion control of cannula flexible needles, the bending characteristics of the needles were studied. The kinematics model of the cannula flexible needles was established by means of the screw theory and the exponential product equation. The bending characteristic analysis method of the cannula flexible needles was proposed based on the dimensional analysis theory and the equivalent stiffness approach to identify the parameters affecting the bending deformations of the needles, and the variable-curvature characteristic function was determined by function regression method. The variable-curvature characteristic experiments were carried out, and the systematic error values are within 2 mm, which meets the accuracy requirements of general clinical surgery and the accuracy of the proposed the variable-curvature characteristic modeling for the cannula flexible needles is proved.

Effects of Reciprocating Motions on Surface Formation Mechanism of Axial Sleeve Typed Lapping of Spherical Rollers

DING Jiahao1, 2, CHEN Guang1, 2, SU Yongxiang1, 2, SUN Yongquan1, 2, REN Chengzu1, 2

2025, 36(03):  463-474,482.  DOI: 10.3969/j.issn.1004-132X.2025.03.010

Asbtract ( )   PDF (11819KB) ( )  

References | Related Articles | Metrics

 In order to reveal the influence laws of reciprocating motions of lapping tools on geometrical accuracy of spherical rollers in new axial sleeve type spherical roller lapping processes, a spherical roller kinematics model was established for simulation calculation, and a material removal model was set up based on Prestons equation, which was used to deduce material removal of spherical roller at each angle of circumferential direction and distribution of texture of rolled surfaces, and then carried out the tests of spherical roller lapping. The results show that the errors between relative sliding rate and simulated value are as 8.33% ~10.94%. The errors of the simulated values for percentage of polar deviation and sliding rate correction are as 11.2%~17.1%, and the roundness errors of ground spherical rollers reach a minimum value of 0.78 μm. The rolling surface texture direction angle is differentially distributed and positively correlated with the surface roughness. The test results validate the accuracy of kinematics simulation analysis, which is instructive for improving the roundness error and surface roughness of the new process.

Analysis of Transmission Characteristics of Multi-stage Denatured Pascal Worm Gears

TAO Dehua1, WANG Ying2, CAO Zhenxin1, LI Ning1, HU Liguang1, XUAN Zhongyi1, JIANG Hongkui1

2025, 36(03):  475-482.  DOI: 10.3969/j.issn.1004-132X.2025.03.011

Asbtract ( )   PDF (4898KB) ( )  

References | Related Articles | Metrics

 In order to obtain the Pascal worm gear mechanisms with different transmission characteristics, a multi-stage denatured Pascal worm gear mechanism was constructed based on 2 stage denatured Pascal worm gear generation mechanism. Pascal worm curve was divided into N segments, and different denatured coefficients were introduced in each segment, and the multi-stage denatured Pascal worm lines were obtained. The mathematical equation of the curve was constructed based on geometric graphics. The multi-stage denatured Pascal worm lines were applied to the gear mechanisms to obtain a non-circular gear mechanism for variable speed transmission, and a mathematical model of this type of gear mechanisms was constructed. A multi-stage denatured Pascal worm line gear mechanism visual design platform was compiled for facilitating the optimization of the design, the effects of denatured coefficient on Pascal worm gear were analyzed, and the selection criteria of denatured coefficient were summarized. The transmission characteristics of 3 stage denatured Pascal worm gear mechanisms were analyzed, and multi-stage denatured Pascal worm gear was applied on the presses. Results show that different transmission characteristics may be obtained by introducing different denaturation coefficients in different sections, and the larger the number of segments, the stronger the adjustability, and the more types of transmission characteristics may be obtained.

Effects of Occupant Posture on Injury Risk in Frontal Crash AEB for Six-year-old Children

WANG Peng1, PAN Di1, 2, YE Meiting1, YE Fan1, HAN Yong1, 2

2025, 36(03):  483-492.  DOI: 10.3969/j.issn.1004-132X.2025.03.012

Asbtract ( )   PDF (13724KB) ( )  

References | Related Articles | Metrics

To clarify the differences in head and thoracic-abdominal injuries among child occupants in different seating positions during frontal car collisions with and without the intervention of AEB, the driving speeds of urban expressways(90km/h) were combined with NHTSA-NCAP test conditions to establish a full-vehicle 100% overlap rigid barrier frontal collision simulation model. A CRS with a backrest was selected. The THUMS 6YO human finite element model was used to simulate three seating positions:reference(RF), head forward-out of position(HF-OP), and pre-submarining(PSB). The kinematics responses and the risks of head and thoracic-abdominal injuries of child occupants in the three positions were analyzed and compared with and without the intervention of AEB. The results show that, under AEB conditions, the HIC15 and the chest 3 ms resultant acceleration of the children in RF sitting position decrease by 43.6% and 24.6% respectively. In PSB position, the head 3 ms resultant acceleration is reduced by 19.2%, and the chest compression is reduced by 18.1%. In HF-OP position, the peak accelerations of the head and chest are reduced by 28% and 25%, respectively. In all three sitting positions, AEB intervention significantly reduces the risk of head and chest injuries in child occupants.

Nonlinear Hysteresis Control and Experimental Study of Low-frequency Large Displacement Multi-mode Piezoelectric Linear Actuators

LI Chong, BAI Xin, TONG Yujian, FANG Jiwen

2025, 36(03):  493-503,514.  DOI: 10.3969/j.issn.1004-132X.2025.03.013

Asbtract ( )   PDF (10778KB) ( )  

References | Related Articles | Metrics

A piezoelectric linear actuator featuring low frequency, large displacement, and high-load characteristics was proposed, capable of operating in three modes to accommodate a variety of operational conditions. The working principle of the actuators and the displacement amplification under different loads were analyzed. For the hysteresis nonlinear phenomenon of the piezoelectric actuators, an improved hysteresis nonlinearity model was developed, which was based on PI and enhanced through a genetic algorithm optimized backpropagation neural network(GA-BP). A GA-BP feedforward fuzzy self-tuning PID compound control system was established, and the hysteresis nonlinear compensation experiments of the actuators were carried out. Results show that the maximum output displacement of the actuators is as 558.3 μm when two piezoelectric stacks are working simultaneously, and the maximum relative error of trajectory tracking is as 0.0573 μm under sinusoidal signals. The implementation of the improved control strategy effectively enhances the performance of the systems, resulting in a trajectory tracking control accuracy of up to 97%, a reduction of the delay time to less than 2 ms, and the ability to realize the control response within 60 ms. This strategy accelerates the systems response speed, and reduces the steady-state errors to less than 0.09%, demonstrating the capability for rapid responses and minimal steady-state errors.

A New Method for Path-engulfment Topology Optimization

LIN Jiahui, LI Zhuangzhuang, LI Xuelin, LI Jun

2025, 36(03):  504-514.  DOI: 10.3969/j.issn.1004-132X.2025.03.014

Asbtract ( )   PDF (11812KB) ( )  

References | Related Articles | Metrics

 To address the difficulty in establishing and solving mathematical models in optimization processes of complex engineering structures, a topology optimization method was established combining load transfer path theory and engulfment algorithm. Using load-transfer path theory to solve the major and minor transfer paths of the structures, an engulfment algorithm was proposed to optimally allocate materials according to the principle of allocating more materials to the regions with relatively high-stress and fewer materials to the regions with relatively low-stress in the structures. Taking cantilever beam and thin plate structure as examples, the path-engulfment method herein was compared to variable-density method with simulations and experiments. Results show that the method herein improves the stiffness and strength of the structures, and reduces the mass of the structures, which proves the accuracy and effectiveness of the method in structure optimization. The method does not need to establish or solve complex mathematical models, and takes into account smooth boundary processing, the optimized structure may be directly used for manufacturing, which is especially suitable for engineering applications.

Design and Optimization of Negative Poissons Ratio Energy Absorption Structures for Anti-shock and Absorption Hydraulic Support Lattices

SHEN Jiaxing1, 2, DONG Jianxiu2, FAN Zhonghai2, YU Yinghua2

2025, 36(03):  515-524.  DOI: 10.3969/j.issn.1004-132X.2025.03.015

Asbtract ( )   PDF (13215KB) ( )  

References | Related Articles | Metrics

In order to improve the energy absorption and supporting characteristics of energy absorption hydraulic supports, an energy absorption structure of lattice material with negative Poisson ratio was designed. Specific energy absorption performance and support force per unit mass performance of straight-edged concave star-shaped cells and concave hexagonal cells with circular cross section and rectangular cross section were investigated using simulation analysis methods. It was determined that the comprehensive performance of the cells with rectangular cross-section stepped edge concave stars  was the best. Using Workbench software based on parametric analysis, the influence rules of cell body height, cell body width, cell body thickness and cell body shrinkage of materials with negative Poisson ratio on energy absorption and support characteristics of the energy absorption structure were studied, and the neural network prediction model of energy absorption and support forces was established. The multi-objective parameter optimization mathematical model of energy absorption structures was established. The optimization mathematical model of energy absorption structures was solved based on fmincon function, and three local optimal solutions were determined. Through comparison and verification analyses, the optimal structural parameters of energy absorption structures were determined as follows. The height of the cell body is as 40 mm, the width of the cell body ia as 9.2 mm, the thickness of the cell body is as 10 mm, and the shrinkage of the cell body is as 5.2 mm.The maximum absorption energy of the optimal parameter energy absorption structures is as 1083.36 kJ, the average supporting force is as 2162.56 kN, and the supporting force fluctuation coefficient is as 1.123,the standard deviation of supporting force is as 28.58 kN. The support size is moderate and the supporting force is stable.

Research on Non-homogeneous Boundary Condition Imposing Techniques for Isogeometric Topology Optimization

WANG Shuting1, XIE Qingtian1, YANG Aodi1, LI Xiaobing2, XIONG Tifan1, XIE Xianda2

2025, 36(03):  525-535.  DOI: 10.3969/j.issn.1004-132X.2025.03.016

Asbtract ( )   PDF (8188KB) ( )  

References | Related Articles | Metrics

The lack of interpolation property of non-uniform rational B-splines(NURBS) basis functions led to the failure of imposing non-homogeneous boundary conditions directly at control points by isogeometric topology optimization method. Therefore, two types of non-homogeneous boundary condition imposing methods were introduced to isogeometric analysis herein, which improved the accuracy of imposing boundary conditions through two perspectives, namely strong imposition method and weak imposition method. Then, the numerical results verify the feasibility of the above methods in applying non-homogeneous boundary condition for isogeometric analysis. Finally, the above methods were embedded into isogeometric topology optimization, and an innovative non-homogeneous boundary condition imposing method was proposed with adaptive penalty factor, which improves the solution accuracy of isogeometric topology optimization, and further demonstrates the effectiveness of the proposed method.

A Method of Metro Bogie Frame Load Identification for Broadband Excitations

WU Yangmin, REN Zunsong, YANG Guangxue, JIN Xincan, XU Ning, YANG Chao

2025, 36(03):  536-545.  DOI: 10.3969/j.issn.1004-132X.2025.03.017

Asbtract ( )   PDF (9016KB) ( )  

References | Related Articles | Metrics

The inversion of transfer matrix in dynamic load identification of engineering structures would produce errors. The load decoupling calibration method was used to diagonalize the transfer matrix to avoid errors caused by the inversion of transfer matrix. Thus, a high-precision force measurement frame with a measurement error of less than 5% was developed based on strain measurement principle, and a multi-source load decoupling identification method for metro bogie frames was established. Aiming at the problems that the frames were affected by wheel-rail excitation, resulting in a wide load band, causing resonance when the excitation frequency was close to elastic modal frequency of the frames, the load amplitude amplification effect near the resonance frequency was directly quantified in frequency domain through a second-order damping system, and the damping ratio was identified by least squares complex frequency domain method. The resonance range was determined by half-power bandwidth method and the calibration critical frequency was obtained. The load amplitude amplification effect was quantified and the load optimization factor was obtained. The effectiveness of the processed load was reflected to a certain extent based on vibration acceleration responses in resonance areas. In terms of time domain, frequency domain and damage, the load prediction stresses are close to the measured ones, and the ratio of predicted damage to actual damage is generally between 1.0 and 1.6, which meets the requirements of frame fatigue assessment. The results have important engineering significance for understanding the real load characteristics of metro bogie frames and exploring resonance fatigue problems.

Bearing Fault Data Generation Method Based on WLT-ACGAN

JIAO Huachao, SUN Wenlei, WANG Hongwei

2025, 36(03):  546-557.  DOI: 10.3969/j.issn.1004-132X.2025.03.018

Asbtract ( )   PDF (7836KB) ( )  

References | Related Articles | Metrics

Using data generation method to generate high-quality data which made time-domain and frequency-domain features consistent with the real signals of bearing faults, and constructing balanced dataset, were of great significance for the establishment of an efficient diagnostic model of bearing faults in the case of data imbalance. In order to address the limitations of the existing data generation methods, which focused on a single feature in time or frequency domains, WLT-ACGAN was proposed herein. Firstly, a WLT network was constructed with a multi-layer neural network based on the principle of wavelet transform. The wavelet transform and inverse transform were simulated, and the mapping relationship between time-domain signal and frequency-domain signal was established. Secondly, the WLT network was embedded into ACGAN model as the primary component of model generator. Finally, two discriminators were constructed with different functions, enabling the improved ACGAN to learn time-domain and frequency-domain feature information of authentic bearing vibration signals concurrently. Experimental results show that the bearing vibration signals generated by WLT-ACGAN model exhibit consistent time-domain and frequency-domain features with those of the actual bearing vibration signals. Furthermore, the fault diagnostic model constructed with the balanced dataset augmented by the generated signals exhibits a high degree of accuracy when the data are imbalanced.

Multi-channel Flatness Parallel Prediction Method for Cold Rolled Strips

DUAN Bowei1, WANG Dongcheng1, 2, XU Yanghuan1, LIU Hongmin1

2025, 36(03):  558-569.  DOI: 10.3969/j.issn.1004-132X.2025.03.019

Asbtract ( )   PDF (9950KB) ( )  

References | Related Articles | Metrics

An ensemble learning method was employed to develop a high-accuracy, strong-generalization cold rolled strip flatness prediction approach. Firstly, a model training dataset was constructed based on industrial big data, which had large data scale and high degree of flatness diversity. The time lag compensation was employed to eliminate time asynchrony among data based on the relative position between rolling mill and flatness meter. The isolation forest algorithm of data mining technology was used to clean outliers in data, improving the quality of training data and model performance. Subsequently, an architecture of multi-channel flatness parallel prediction was constructed based on XGBoost algorithm. This architecture was trained using processed production dataset to obtain the cold rolled strip flatness prediction model(CCFD_M). Lastly, based on CCFD_M, a flatness channel optimization algorithm was proposed to eliminate the issue of “pseudo-flatness” in prediction results, and the practical version of cold rolled strip flatness prediction model(CCFD_OM) was obtained. After verification on test set, the prediction error indictors mean absolute error(MAE) and root mean square error(RMSE) of model CCFD_OM reach 0.4044I and 0.6816I, respectively. And the fitting performance indictor R2 reaches 0.83, which may meet the practical production requirements.

Model-free Trajectory Tracking Control of Soft Robots Based on LQR and UKF

GUAN Shengchuang1, LIU Yujun2, YANG Qinghao1, LIU Zhaobing1

2025, 36(03):  570-575,583.  DOI: 10.3969/j.issn.1004-132X.2025.03.020

Asbtract ( )   PDF (5460KB) ( )  

References | Related Articles | Metrics

A novel nonlinear estimation and control strategy for controlling the dynamic performance of a 2D pneumatic soft robot was proposed to address the problems of accurate modelling and control of soft robots. Firstly, a linear model of the 2D pneumatic soft robot was established using a Koopman operator-based approach. Then, the UKF was proposed for sensor data filtering and system state estimation, while the LQR was used for optimal control of trajectory tracking. Simulation and experimental results consistently show that the strategy herein performs better than other two strategies  in terms of trajectory tracking.

Abnormal Detection of Injection Molding Products Based on Semi-supervised Learning Dual-model Structure

CHEN Yu1, XIANG Wei1, 2, LIN Wenwen1, GONG Chuan1, ZHANG Huaizhi1, YU Renhao1

2025, 36(03):  576-583.  DOI: 10.3969/j.issn.1004-132X.2025.03.021

Asbtract ( )   PDF (1735KB) ( )  

References | Related Articles | Metrics

The imbalance in quality data distribution and the ambiguity of classification boundaries were identified as the factors limiting the performance of traditional classifiers, which hindered the efficient implementation of intelligent production decision-making in enterprises. Therefore, a dual-model-based deep generative model anomaly detection method was proposed. Qualified products were classified into two levels, excellent and sub-optimal, based on size data distribution, and two deep generative models were trained separately. A weighted ensemble strategy was designed considering characteristics of data distribution, and anomaly scores were calculated to determine product qualification. Two dual-model structures were developed based on VAE and WGAN. Experimental results indicate that, compared to single-model structures, the dual-model-based VAE and WGAN improve classification accuracy on the test sets by 4.5% and 6%, respectively.

Mesoscale Numerical Simulation of SLM Processes for Al-Mn-Sc-Zr

LI Xinyu1, 2, 3, ZHOU Yonghua1, 2, 3

2025, 36(03):  584-592，603.  DOI: 10.3969/j.issn.1004-132X.2025.03.022

Asbtract ( )   PDF (13297KB) ( )  

References | Related Articles | Metrics

The microstructure and defects of parts produced by SLM were closely related to the quality of SLM single/double-track forming, and experimental methods were difficult to explain complex physical phenomena of track forming at mesoscale. Al-Mn-Sc-Zr alloy was used as forming material to establish a mesoscale numerical model for SLM single/double-track. SLM single/double-track forming experiments were conducted, and the accuracy of model was verified. Numerical model was employed to reveal the basic laws of SLM molten pool evolution. The variation of molten pool characteristics such as single-track morphology, molten pool temperature, and molten pool size was investigated with laser power ranging in 200~400 W. The best quality of single-track is achieved at a laser power of 360 W and a scanning speed of 1.0 m/s. The overlapping behavior and the evolution law of double-track at scanning speed of 1.0 m/s, the laser power of 360 W and the scanning distance of 60 μm, 80 μm and 120 μm were investigated. Results show good molding quality and high molding efficiency are reached with a scanning distance of 80 μm. Suitable VED is screened in combination with multi-track SLM experiments. In the VED range of 107~139 J/mm3, the multi-track is continuous and smooth. With the help of the model, key processing parameters may be initially predicted, and reduce the cost of processing parameter optimization.

Study on Forming and Mechanics Properties of 17-4PH Stainless Steel Parts Fabricated by MFFF Technique

JIANG Shijie1, 2, XU Zizhao1, LI Shuguang1, WANG Fei1, HUANG Xuzhen3

2025, 36(03):  593-603.  DOI: 10.3969/j.issn.1004-132X.2025.03.023

Asbtract ( )   PDF (11355KB) ( )  

References | Related Articles | Metrics

Three types of self-made 17-4PH metal/polymer composite filaments were developed, which were polylactic acid(PLA) single-component, PLA/styrene(SEBS) bi-component and PLA/high-density polyethylene(HDPE)/ethylene vinyl acetate(EVA) three-component composite filament. The ordinary fused filament fabrication(FFF) machine was then used to build green parts, and the feasibility of one-step thermal debinding/sintering was studied using a box sintering furnace. The formability of the self-made composite filaments was further explored. Finally, MFFF sintered parts were formed with Ultrafuse 17-4PH composite filament through the shaping-debinding-sintering(SDS) processes. The density, shrinkage and static mechanics properties were studied. Results indicate that the one-step thermal debinding/sintering process is prone to generate defects such as oxidation and non-shrinkage in the samples, while the samples processed with two-step debinding(i.e. catalytic debinding and thermal debinding) and then sintering have a good forming quality. The average density of Ultrafuse 17-4PH sintered parts is as 7.2 g/cm3, and the shrinkage in X, Y and Z direction are as 15%, 18% and 20% separately. The maximum tensile strength and bending strength may reach 600 MPa and 1200 MPa, respectively. The hardness is about 230 HV. 

Research on Fuzzy Model Predictive Control Method for High Speed Intelligent Vehicles Based on Variable Universe

HE Yang, LI Gang, YU Xiaonan

2025, 36(03):  604-613.  DOI: 10.3969/j.issn.1004-132X.2025.03.024

Asbtract ( )   PDF (6214KB) ( )  

References | Related Articles | Metrics

In order to improve the ability of trajectory tracking and driving stability of high speed intelligent vehicles, a variable universe fuzzy model predictive control method(VUFMPC) was proposed. Based on the traditional method of trajectory tracking model predictive control (MPC) of intelligent vehicles, a fuzzy model predictive controller(FMPC) was established by taking the output errors and the rate of change as inputs, and the adjustment factors of error weight and control increment as outputs. For the universe inability to adaptively adjust, variable universe fuzzy control method (VUFC) was introduced to adaptively adjust the universe of FMPC based on output errors. Finally, this method was verified through hand-in-loop experiments. The experimentd results show that compared to MPC and FMPC, the maximum tracking error is reduced by 78.8% and 53.6%, the average tracking error is reduced by 38.1% and 31.6%, the optimization quantity of lateral speedy is in 52.3%~50.7% and 33.5%~30.9% respectively. VUFMPC reduces the tracking errors and makes driving more stable for a high speed intelligent vehicles.

An Improved Ant Colony Algorithm for Cast Plan Considering Online-width-adjustment Technology of Crystallizers

LI Yiren1, 2, WANG Bailin1, 2, YUAN Shuaipeng1, 2, LI Tieke1, 2, WANG Yang1, 2

2025, 36(03):  614-622，633.  DOI: 10.3969/j.issn.1004-132X.2025.03.025

Asbtract ( )   PDF (2976KB) ( )  

References | Related Articles | Metrics

 Aiming at the problems of cast planning in steelmaking-continuous casting scheduling, integrating the online width adjustment technology for the mold，with the objective of minimizing the weighted cost(including the furnace switching cost between heats, downtime cost for the continuous caster, and the online width adjustment cost for the mold), the traditional model was improved, resulting in the development of a new cast planning model. Meanwhile, inspired by the optimization approach used in the vehicle routing problem, an improved ant colony algorithm was proposed. The algorithm included node sorting and selection rules, and the reward-punishment factors and quantile parameters were incorporated to enhance the search breadth and adaptability. Finally, the effectivenesses of the proposed model and algorithm were validated by using real production data. 

Research on Comprehensive Motion Model of Dynamic Seats Characterized by Dynamic Dimensions

WANG Jin1, ZHANG Xuwei1, CHEN Jianping2, ZHANG Lili2

2025, 36(03):  623-633.  DOI: 10.3969/j.issn.1004-132X.2025.03.026

Asbtract ( )   PDF (5755KB) ( )  

References | Related Articles | Metrics

Dynamic seats were a crucial means for intervening in prolonged sitting discomfort, and the comprehensive motion model of dynamic seats played a significant role in determining the effectiveness of intervention. To enhance the efficacy of dynamic seating interventions, a study of comprehensive motion model of dynamic seats characterized by dynamic dimensions was conducted. Firstly, a method for constructing dynamic seating dimensions was developed, involving the use of adjustable seats to collect user-specific comfortable seat dimensions. Secondly, based on the differences in comfortable seat dimensions among users with different body types, the variation trend of measured dimensions of comfortable seats for different body types over time was fitted, thereby constructing dynamic dimensions. Then, taking passenger seats as examples, a comprehensive motion model of dynamic seats was constructed using seven interrelated seat support performance dimensions. Finally, an intra-subject experiment was conducted to evaluate the effectiveness of dynamic dimensions in reducing sedentary discomfort. The results reveal that seat height, backrest inclination, lumbar support height, lumbar support thickness, and neck support thickness all include a set of dynamic dimensions, while seat pan inclination includes three sets of dynamic dimensions, and neck support height includes two sets. These dynamic dimensions significantly reduce discomfort in key body parts and overall discomfort.