To accurately estimate the remaining useful life of rolling bearings, which are essential components in mechanical devices, would greatly enhance their reliability and efficiency. A novel prediction approach that integrates digital twin technology with time series forecasting, employing the Informer model, is introduced. By utilizing digital twin technology to establish a virtual model of the bearing, we can monitor and simulate the physical condition of the bearing. Then the Informer is utilized to complete the remaining useful life projection on the information provided by the processed digital twin technology. The experiment demonstrates that the method can accurately estimate the remaining useful life of the bearing, which provides strong support for realizing intelligent maintenance and fault prevention of equipment.
Aiming at the problem that the traditional ORB algorithm is difficult to realize the real-time and accurate grasp of the object by the robotic arm equipped with the binocular stereo vision,a robotic arm rec-ognition and grasping method based on an improved ORB algorithm was proposed.Firstly,the improved ORB al-gorithm was used to identify the target,and then the target was positioned according to the binocular stereo vi-sion optical axis parallel model.Then,the mathematical model of the robotic arm was established by using the standard D-H parameter method,and the inverse kinematics was solved.Finally,the obtained drive angles of each joint of the robotic arm were transmitted to the control end of the system.The end effector of the robotic arm was driven to complete the grasp of the target.The results show that compared with the traditional ORB algo-rithm,the improved ORB algorithm in this study can improve the speed and accuracy of target recognition and positioning,and effectively improve the real-time and accurate grasping of the robotic arm.
In order to solve the problem that the current tool wear defects are difficult to be collected by the visual inspection system, an Otsu threshold segmentation algorithm based on particle swarm optimization is proposed.to detect tool wear The algorithm improved update strategy for inertia coefficients which effectively expanding the search scope of the algorithm and shortens the running time of the algorithm. By adding a perturbation equation to the particle swarm, solved the problem of traditional particle swarm optimization algorithms easily falling into local optima. Finally, an experimental platform is built to verify the effectiveness of the algorithm. This inspection method can achieve the identification of tool damage areas and the measurement of tool damage amount, and has advantages such as high recognition accuracy and fast running speed compared to traditional Otsu algorithm, Canny algorithm ,local threshold segmentation and so on. The research results have certain reference value for the actual tool defect detection system.
The proposed method for tire specification character recognition based on the YOLOv5 network aimed to address the low efficiency and accuracy of the current character recognition methods. The approach involved making three major modifications to the YOLOv5 network to improve its generalization ability, computation speed, and optimization. The first modification involved changing the coupled head in YOLOv5 to a decoupled head, which could improve the network's generalization ability. The second modification proposed incorporating the C3-Faster module, which would replace some of the C3 modules in YOLOv5's backbone and head and improve the network's computation speed. Finally, the third modification proposed replacing YOLOv5's CIoU loss function with the WIoU loss function to optimize the network. Comparative experiments were conducted to validate the effectiveness of the proposed modifications. The C3-Faster module and the WIoU loss function were found to be effective, reducing the training time of the improved network and increasing the mAP by 3.7 percentage points in the ablation experiment. The experimental results demonstrated the effectiveness of the proposed method in improving the accuracy of tire specification character recognition and meeting practical application requirements. Overall, the proposed method showed promising results for improving the efficiency and accuracy of automotive tire specification character recognition, which has potential applications in various industries, including automotive manufacturing and tire production.
In order to recognize the fault type of reducer abnormal vibration and reduce the cost of inspection and maintenance, an intelligent diagnosis model is developed. In the case of insufficient historical abnormal vibration data, a fault tree of the reducer is established by combing the historical fault data. It is then mapped to the Bayesian network structure. The expectation maximization (EM) algorithm is selected as the parameter learning method to determine the probability distribution of the node variables. After processing real-time vibration data, the model integrates the abnormal vibration feature discrimination mechanism and hierarchical Gibbs sampling algorithm to carry out fault probability inference. Compared with other models, the proposed model has achieved great improvement in the accuracy of diagnosis results and distinguishing normal and abnormal data. The model is integrated into the intelligent operation and maintenance system of belt conveyor for engineering verification.
Machine tool condition monitoring is of great significance for machine tool health management and workpiece processing quality assurance.In view of the problems of existing tool wear prediction models such as long training time,slow convergence speed and weak generalization ability,this paper proposes a distributed one-dimensional convolutional neural network to predict tool wear.The residual connection and channel attention modules are sequentially stacked as the feature extraction module,and cross-validation was used to select the appropriate number of network layers.Since the feature information extracted by dif-ferent sensors may be redundant,a weight difference strategy was used to improve the effectiveness and comprehensiveness of feature extraction.In addition,considering that the distribution of the training set and the test set may be different,which affects the generalization performance of the model,a domain adapta-tion method is introduced to improve the performance of the model in unknown data sets.In order to verify the effect of the model,experiments were carried out using the PHM 2010 milling cutter wear data set.The experimental results show that the average RMSE and average MAE of the model on the three tools C1,C4,and C6 are 6.97 and 6.29,respectively,which is more than 12%improved compared with TCN,TDConv-LSTM and other models.
The concomitant vibration and deformation produced by propeller blades in single-sided machining seriously affect the surface machining precision. Double-sided symmetrical machining can improve system rigidity through mutual shoring on both sides which abates the concomitant vibration and deformation. However, the actual double-sided symmetrical machining cannot be applied to blade machining due to its shape complexity. The double-sided collaborative machining method combining symmetrical machining and staggered machining is devised in this paper, and its tool path planning algorithm is investigated. Firstly, the algorithm achieves smooth fitting and correspondence of bilateral cutter position points through double-curve interpolation and position data alignment. Secondly, the blade surface is divided into four regions by two partition parameters: tip region, internal region, variable region, and edge region. Then, the conversion between symmetrical machining and staggered machining is completed through the Sigmoid deformation curve in the variable region. Finally, the feasibility and superiority of double-sided collaborative machining are verified through machining experiments.
Abstract Connecting rod bolt is the key component of the diesel engine crank connecting rod mechanism, and the phenomenon of stress concentration at the thread easily occurs in the working process of the crank connecting rod, due to the characteristics of the threaded connection. In order to improve the quality of the threaded connection of the diesel engine connecting rod, the connecting rod’s big end threaded connection structure is numerically simulated based on ANSYS finite element analysis software through the combination of theoretical analysis and numerical simulation. The stress distribution of different pitches, threaded connection length and bolt preload are analyzed, and the effects of different factors on the strength of threaded connection structure are obtained. The research results have important engineering application value for improving the reliability of the threaded connection and ensuring the safe operation of the diesel engine connecting rod mechanism.
Effective integration of process planning and shop scheduling is the main way to solve the problems of low equipment utilization and poor production efficiency when process planning and shop scheduling are optimized separately. In order to achieve green and low-carbon manufacturing, this paper firstly takes the minimization of total carbon emissions, the maximum time of completion and delay time in the manufacturing process as the optimization goals. On account of the principle of nonlinear process planning, a multi-objective integrated process planning is established. Then, the NSGA-II genetic algorithm is used to solve the problem, and the Pareto optimal solution set is obtained by optimization; and the TOPSIS decision method based on the entropy weight method is proposed to select the optimal process route and scheduling scheme from the Pareto optimal solution set. Finally, combined with cases, the validity of the model and the solution process is verified and also provides a basis for the decision of scheduling scheme in the actual production process.
Aiming at the problem that the initial fault signal of rolling bearings is weak and the fault characteristic is difficult to extract, this study proposes a rolling bearing fault diagnosis method based on variational modal decomposition (VMD) for adaptive parameter optimization based on the improved sparrow search algorithm (SSA) and the extreme learning machine (ELM) with multi-layer feature vector fusion. Firstly, the optimization step size of SSA is adaptively changed according to the fittness function value and the number of iterations. Secondly, the improved SSA optimizes the important parameters (decomposition number K and penalty factor α) of the VMD algorithm, and the fittness function adopts the minimum envelope entropy. Thirdly, the intrinsic mode function (IMF) component with the smallest envelope spectral entropy after SSA-VMD decomposition is extracted as the optimal component, and its eigenvalue is calculated. Finally, through the screening of coefficients of the variation method, the root mean square value and peak value are constructed as the two-dimensional eigenvalue vector of the first layer, and the sample entropy, kurtosis and root mean square are constructed as the three-dimensional eigenvalue vector of the second layer, which are respectively sent to the limit learning machine ELM for the training and classification of rolling bearing faults.The experiment results show that the proposed algorithm has good fault diagnosis performance,ultimately achieving a classification accuracy of 98.25% and an actual diagnostic accuracy of 93.36%.
To realize permanent magnet synchronous motor (PMSM) in the full speed domain without speed sensor operation, a hybrid control method combining I/F startup and extended Kalman filter (EKF) is proposed in this paper. This method employs I/F startup to transition at low speed, effectively resolving the issue that the position estimation method based on the back electromotive force (EMF) model fails at zero speed and low speed, and converts to EKF for speed closed-loop vector control at medium and high speed. Moreover, a new feedback regulation mechanism as a solution to the problem of smooth switching between the two methods is proposed. First, the power angle is determined based on the relationship between the given I/F frequency and the estimated EKF position angle. Using the information of power angle, the damping torque of the system is increased to reduce velocity fluctuations during I/F startup. In addition, the balance point of current and position error angle is adjusted using the closed-loop information of position error angle to reduce the torque abrupt change before and after switching, thereby making the motor switching process to EKF speed closed-loop control more stable. Finally, simulation results are used to verify the effectiveness of the proposed scheme.
针对航空发动机叶片修复过程中,损伤叶片在进行激光熔覆后需进行铣削加工来达到叶片表面具有较好的表面粗糙度的效果,采用3D增材制造后的TC4材料开展机理性实验.为使试验所选用TC4材料表面的加工质量和加工效率得以提升,确定TC4试件表面粗糙度为研究对象,借助Design-Expert软件设计正交试验,并在数控车床JDPVM600_A13SH上开展硬质合金刀具铣削试验,探究铣削参数(主轴转速、进给速度、切削深度)对表面粗糙度的影响.采用多元回归的方法建立TC4材料表面粗糙度与铣削参数之间的关系模型,进而以TC4材料表面粗糙度和加工效率作为优化目标构建多 目标优化模型,并对试验铣削参数采用粒子群算法(PSO)进行优化.结果表明,采用粒子群算法优化后的铣削参数进行加工,TC4材料表面的粗糙度有所降低,加工效率也有所提升.
多相机测量系统会因较小的重叠视野与环境干扰而导致外参标定失败,因此提出了一种基于多编码块组合型标定板的多相机外参标定方法.首先,设计了一种平面标定板,它由多个具有独立编码信息的编码块组合而成;其次,研究了编码块特征点的提取与管理方法;最后,通过重建特征点并优化重投影误差,实现了各个相机外参的全局标定.实验结果表明,该方法能够适应各相机无法同时观测完整标定板的情况,并且有标定后的测量误差为0.014 mm,标准差为0.112 mm.因此,提出的方法能够实现多相机测量系统全局外参的高精度快速标定.
针对元件分拣作业效率低下、传统的二维约束矩阵自动化程度低的问题,结合机器视觉和遗传算法,提出一种最短分拣路径优化策略.首先,对视觉系统进行标定解算出机器人-像素坐标系位姿变换关系;其次,运用深度学习识别出目标元件,以开启多线程的方法对3类目标框进行图像处理得到定位结果;最后,创建三维动态约束矩阵,运用贪心策略和遗传算法组合算法对分拣路径进行优化,探究不同的元件放置和元件组成对路径优化的影响.实验结果表明,优化后的分拣路径较随机路径缩短了2.12％ ～15.2％,分拣效率提高了0.52％ ～3.06％.实验结果揭示的相关规律对分拣作业的实际应用具有重要的现实意义.
针对实际工况下旋转机械故障样本稀缺,传统深度学习方法在小样本情况下出现欠拟合的问题,提出一种基于孪生网络(siamese network)的旋转机械故障诊断方法.首先,对有限的故障样本进行交叉配对,构造相同类别与不同类别的故障输入样本,实现对数据样本量的大幅扩容;然后,针对小样本问题构造了包含两个子模型的孪生网络,其中子模型由卷积神经网络与自注意力机制构建而成,通过计算样本之间的欧式距离得到相似度参数;最后,将相似度参数输入自建损失函数和故障分类器实现故障识别.实验证明,所以提出的网络能够依靠少量样本完成故障诊断,在样本数量相同的情况下,故障识别准确率显著高于其他深度学习模型.
为提高磨抛机器人在未知干扰和建模不确定性下的动态柔顺性能,实现高精度的自动磨抛,提出了一种主动柔顺模糊自适应变阻抗控制方法.以阻抗控制为基础,引入快速终端滑模进行运动控制,以提升控制系统鲁棒性与响应速度,基于指数趋近率,采用连续的双曲正切函数代替符号函数建立带有抑制抖振功能的控制律;在此基础上,采用模糊控制调节阻抗参数来提高系统响应时间和减少超调量,设计了模糊自适应变阻抗模块,定义了模糊规则.通过李雅普诺夫函数验证了控制器稳定性.进行了Simulink仿真对比实验,结果表明,相较于PID,所提方法在响应速度方面提高了37.2%,超调量降低了20.5%,累计误差方面也有明显改善,有利于提高磨抛表面的一致性和磨抛精度.
为了减小数控进给轴运动过程中产生的阿贝误差和余弦误差等几何误差对位置测量精度的影响,提出了一种基于多路激光组合测量的误差辨识与补偿方法.利用三路激光干涉仪的空间坐标关系和实际测量值,映射出数控进给轴理想运动轴线上的虚拟测量值,以补偿阿贝误差,并建立了进给轴倾斜角度解算模型,进而补偿余弦误差,提高进给轴定位精度.为保证位置测量结果的可靠性,设计了环境参数补偿实验,在减小环境因素影响的前提下,验证了误差辨识与补偿方法的有效性.实验结果表明,该方法有效补偿了运动过程中的阿贝误差以及余弦误差,提高了数控进给轴的定位精度.
为提高NURBS曲线插补的运动平滑性,提出一种基于四阶S曲线加减速的NURBS曲线插补算法.该算法通过使用四阶S曲线加减速求解NURBS曲线节点上的速度,首先分析最大速度、最大加速度和最大加加速度的可达性,同时考虑非对称四阶S曲线加减速模型,并对较短弧长的采用二分法寻找实际最大速度,规划出31 种速度曲线类型,然后根据NURBS曲线参数和约束条件生成对应的速度曲线.仿真结果表明,与三阶S曲线加减速控制的NURBS曲线插补算法相比,所提算法的插补精度更高,加速度曲线光滑且加加速度曲线连续无突变,降低其运动过程中的柔性冲击.
针对轴承表面缺陷小、几何形状多变以及低对比度的特点,提出了一种改进的Faster R-CNN算法,对轴承表面缺陷进行检测.首先,以ResNet-50 结合特征金字塔网络对轴承表面缺陷进行特征提取;其次,在改进的特征提取网络中引入可变形卷积,通过卷积学习偏移量自适应调整感受野,提高了缺陷的提取能力;最后,针对ROI Pooling因二次量化而导致的区域不匹配问题,采用基于双线插值的ROI Align改进ROI Pooling.实验结果表明,在采集的轴承表面缺陷数据集上,改进的Faster R-CNN平均精度均值为97.6%,与改进前相比,提高了11.76%,可以实现对轴承表面各类缺陷更为准确的检测,具有较强的实用性.
针对当前带钢在表面缺陷检测过程中存在检测算法精度有待提高等问题,提出了 一种基于改进YOLOv5算法的带钢表面缺陷检测模型.首先,在检测端构建新的检测层,提高网络对不同尺寸目标的检测;其次,在主干网络结构中引入注意力模块,进一步加强网络提取特征的能力;然后,通过BiFPN_Add来增强深浅层特征信息的融合;最后,构建新的CNeB模块来取代各检测层对应的C3模块,进而增强网络对特征的提取.实验结果表明,改进后的算法在NEU-DET数据集上均值平均精度达到了 80.9％,较原有的算法提升了 4.5％,同时检测速度与原模型保持基本不变,性能优于目前其他主流的检测方法.
