Pub Date : 2025-03-06DOI: 10.1016/j.ymssp.2025.112480
Lorenzo Mazzanti , Daniel De Gregoriis , Thijs Willems , Simon Vanpaemel , Mathijs Vivet , Frank Naets
This contribution introduces the Generalized Augmented MANifold Differential Algebraic Extended Kalman Filter (GAMANDA-EKF), a novel Kalman filter-based methodology for state-input-parameter estimation for structures modelled as multibody systems described by differential algebraic equations. The proposed Kalman filter allows for exact equality and inequality constraint satisfaction and consistent error covariance propagation, without requiring a reformulation of the system equations. In addition to the enforcement of the equality and inequality constraints on the a-posteriori estimated system state with a constrained optimization approach, the estimation error covariance matrix is projected on the constraint manifold as well. This results in increased robustness and stability. Numerical and experimental validation cases using a slider-crank system, employing camera-based position tracking as reference measurements for the estimation, demonstrate the effectiveness of the proposed approach in estimating parameters such as connection stiffnesses and slider friction forces across diverse dynamic scenarios. Furthermore, this work highlights how the enforcement of inequality constraints mitigates estimation instability resulting from suboptimal filter tuning, providing increased robustness to the estimation process.
{"title":"Robust vision-based estimation of structural parameters using Kalman filtering","authors":"Lorenzo Mazzanti , Daniel De Gregoriis , Thijs Willems , Simon Vanpaemel , Mathijs Vivet , Frank Naets","doi":"10.1016/j.ymssp.2025.112480","DOIUrl":"10.1016/j.ymssp.2025.112480","url":null,"abstract":"<div><div>This contribution introduces the Generalized Augmented MANifold Differential Algebraic Extended Kalman Filter (GAMANDA-EKF), a novel Kalman filter-based methodology for state-input-parameter estimation for structures modelled as multibody systems described by differential algebraic equations. The proposed Kalman filter allows for exact equality and inequality constraint satisfaction and consistent error covariance propagation, without requiring a reformulation of the system equations. In addition to the enforcement of the equality and inequality constraints on the a-posteriori estimated system state with a constrained optimization approach, the estimation error covariance matrix is projected on the constraint manifold as well. This results in increased robustness and stability. Numerical and experimental validation cases using a slider-crank system, employing camera-based position tracking as reference measurements for the estimation, demonstrate the effectiveness of the proposed approach in estimating parameters such as connection stiffnesses and slider friction forces across diverse dynamic scenarios. Furthermore, this work highlights how the enforcement of inequality constraints mitigates estimation instability resulting from suboptimal filter tuning, providing increased robustness to the estimation process.</div></div>","PeriodicalId":51124,"journal":{"name":"Mechanical Systems and Signal Processing","volume":"229 ","pages":"Article 112480"},"PeriodicalIF":7.9,"publicationDate":"2025-03-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143548715","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-06DOI: 10.1016/j.ymssp.2025.112516
Binghang Xiao, Jianzhe Huang, Zhongliang Jing
For a discontinuous dynamical system with multi-degree-of-freedom, the coefficients of the established model may be coupled and the response can be also non-smooth, which bring difficulties to estimating the system under fast convergence requirement. In this paper, a nonlinear observer based on super-twisting sliding mode method is designed to evaluate the difference between the prediction of the model and the actual system dynamics. The pseudoinverse and Newton-Raphson iteration are firstly integrated to achieve the fast and reliable estimation of the coefficients of the model including the coupled ones. The event-triggered mechanism is specifically designed to detect the singularity of the dynamics for such a discontinuous system, such that the measured response which is used for the proposed algorithm can be divided into different segments. The experimental validation is conducted, which shows that the proposed method can gain more accurate estimations than pseudoinvese-based comparison method. The switching time between free motion and collision, as well as the collision force can also be predicted accurately based on the model with the estimated coefficients.
{"title":"A fast system estimation algorithm for a discontinuous dynamical model with coefficients coupling","authors":"Binghang Xiao, Jianzhe Huang, Zhongliang Jing","doi":"10.1016/j.ymssp.2025.112516","DOIUrl":"10.1016/j.ymssp.2025.112516","url":null,"abstract":"<div><div>For a discontinuous dynamical system with multi-degree-of-freedom, the coefficients of the established model may be coupled and the response can be also non-smooth, which bring difficulties to estimating the system under fast convergence requirement. In this paper, a nonlinear observer based on super-twisting sliding mode method is designed to evaluate the difference between the prediction of the model and the actual system dynamics. The pseudoinverse and Newton-Raphson iteration are firstly integrated to achieve the fast and reliable estimation of the coefficients of the model including the coupled ones. The event-triggered mechanism is specifically designed to detect the singularity of the dynamics for such a discontinuous system, such that the measured response which is used for the proposed algorithm can be divided into different segments. The experimental validation is conducted, which shows that the proposed method can gain more accurate estimations than pseudoinvese-based comparison method. The switching time between free motion and collision, as well as the collision force can also be predicted accurately based on the model with the estimated coefficients.</div></div>","PeriodicalId":51124,"journal":{"name":"Mechanical Systems and Signal Processing","volume":"229 ","pages":"Article 112516"},"PeriodicalIF":7.9,"publicationDate":"2025-03-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143548716","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-06DOI: 10.1016/j.ymssp.2025.112530
Zeng-hui Liu , Jian-bing Chen , Yong-bo Peng
Atmospheric boundary layer turbulent wind field is a typical 3-directional (3-D) random field with three turbulence components. Reasonable description and simulation of the 3-D turbulent wind field lay a significant foundation for the wind-induced response analysis and wind-resistance design of large wind turbines (WTs). In this study, a novel simulation method of the 3-D turbulent wind field for large WTs based on the stochastic harmonic function, physical turbulent spectrum and the rotational sampling scheme is proposed. The theoretical context of the physical spectral tensor based on the rapid distortion theory (RDT) of the uniform shear turbulent field is first introduced. Then the stochastic harmonic function (SHF) representation method is proposed to express the 3-D turbulence components as a summation of a series of harmonic components with random wavenumbers and phase angles, and the rotational sampling scheme is employed to reduce the number of spatial discretization points of the wind field. To further relieve the computing burden of the harmonic superposition, a two-step acceptance-rejection (A-R) scheme is proposed to reduce the wavenumber terms in the simulation formula, and the evolution phase model (EPM) is employed to reduce the dimension of the random phase angles. The proposed method is then validated through the numerical simulation of 3-D turbulent wind fields of a 5 MW WT, and the advantages of the proposed method in both efficiency and accuracy are revealed through comparisons with the conventional spectral representation approaches.
{"title":"An efficient stochastic harmonic function approach for the simulation of 3-directional wind field of large wind turbines based on physical turbulent spectral model","authors":"Zeng-hui Liu , Jian-bing Chen , Yong-bo Peng","doi":"10.1016/j.ymssp.2025.112530","DOIUrl":"10.1016/j.ymssp.2025.112530","url":null,"abstract":"<div><div>Atmospheric boundary layer turbulent wind field is a typical 3-directional (3-D) random field with three turbulence components. Reasonable description and simulation of the 3-D turbulent wind field lay a significant foundation for the wind-induced response analysis and wind-resistance design of large wind turbines (WTs). In this study, a novel simulation method of the 3-D turbulent wind field for large WTs based on the stochastic harmonic function, physical turbulent spectrum and the rotational sampling scheme is proposed. The theoretical context of the physical spectral tensor based on the rapid distortion theory (RDT) of the uniform shear turbulent field is first introduced. Then the stochastic harmonic function (SHF) representation method is proposed to express the 3-D turbulence components as a summation of a series of harmonic components with random wavenumbers and phase angles, and the rotational sampling scheme is employed to reduce the number of spatial discretization points of the wind field. To further relieve the computing burden of the harmonic superposition, a two-step acceptance-rejection (A-R) scheme is proposed to reduce the wavenumber terms in the simulation formula, and the evolution phase model (EPM) is employed to reduce the dimension of the random phase angles. The proposed method is then validated through the numerical simulation of 3-D turbulent wind fields of a 5 MW WT, and the advantages of the proposed method in both efficiency and accuracy are revealed through comparisons with the conventional spectral representation approaches.</div></div>","PeriodicalId":51124,"journal":{"name":"Mechanical Systems and Signal Processing","volume":"229 ","pages":"Article 112530"},"PeriodicalIF":7.9,"publicationDate":"2025-03-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143548666","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-06DOI: 10.1016/j.ymssp.2025.112551
Kunhong Chen , Hongguang Liu , Jun Zhang , Wanhua Zhao
Surface roughness plays a crucial role in maintaining the productivity and quality of milling operations. Cutting force signals have direct relationships with the surface roughness; however, they are difficult to measure in the practical milling process. Spindle vibration signals are much more easily accessible during milling and widely utilized for force and surface roughness monitoring. However, some vibration frequency contents related to surface roughness are distorted due to the dynamic characteristics of the machine tool, which results in poor monitoring performance. Meanwhile, the mechanism of existing studies that compensate for these distortions remains unclear, and most existing studies apply multi-sensor technology, which makes the monitoring system costly. To bridge these gaps, this paper proposes a causal inference model that learns the dynamic behavior of the spindle system to mitigate signal distortions, alongside developing lightweight models for roughness monitoring. First, a novel causal inference neural network model was proposed to analyze the dynamic behavior of spindle system components during vibration transmission, and a wide-bandwidth force reconstruction model based on single-channel vibration signals was established. Second, to reduce the complexity of the roughness monitoring model, this study investigated the relationship between the surface profile and cutting force, proposing an adaptive method called improved decorrelation EMD (IDEMD) to extract surface roughness-related force components automatically. To enhance the algorithm’s efficiency for real-time component extraction, a cutting force component extraction neural network (FCENN) is proposed to learn the IDEMD process. Third, lightweight surface roughness models were established using a small number of features extracted from the extracted force components. Experimental results under variable cutting conditions demonstrated that the force reconstruction accuracy could reach 0.989, and the surface roughness monitoring accuracy could reach 0.994 in the test set. The proposed method improved the interpretability of data-driven methods by incorporating domain knowledge from signal transmission, surface formation, and signal decomposition, met real-time monitoring requirements, reduced the monitoring model’s complexity, and was less costly and suitable for industrial applications.
{"title":"Causal inference dynamic modeling for real-time surface roughness monitoring in the milling process","authors":"Kunhong Chen , Hongguang Liu , Jun Zhang , Wanhua Zhao","doi":"10.1016/j.ymssp.2025.112551","DOIUrl":"10.1016/j.ymssp.2025.112551","url":null,"abstract":"<div><div>Surface roughness plays a crucial role in maintaining the productivity and quality of milling operations. Cutting force signals have direct relationships with the surface roughness; however, they are difficult to measure in the practical milling process. Spindle vibration signals are much more easily accessible during milling and widely utilized for force and surface roughness monitoring. However, some vibration frequency contents<!--> <!-->related to surface roughness are distorted due to the dynamic characteristics of the machine tool, which results in poor monitoring performance. Meanwhile, the mechanism of existing studies that compensate for<!--> <!-->these distortions remains unclear, and most existing studies apply multi-sensor technology, which makes the monitoring system costly. To bridge these gaps, this paper proposes a causal inference model that learns the dynamic behavior of the spindle system to mitigate signal distortions, alongside developing lightweight models for roughness monitoring. First, a novel causal inference neural network model was proposed to analyze the dynamic behavior of spindle system components during vibration transmission, and a wide-bandwidth force reconstruction model based on single-channel vibration signals was established. Second, to reduce the complexity of the roughness monitoring model, this study investigated the relationship between the surface profile and cutting force, proposing an adaptive method called improved decorrelation EMD (IDEMD) to extract surface roughness-related force components automatically. To enhance the algorithm’s efficiency for real-time component extraction, a cutting force component extraction neural network (FCENN) is proposed to learn the IDEMD process. Third, lightweight surface roughness models were established using a small number of features extracted from the extracted force components. Experimental results under variable cutting conditions demonstrated that the force reconstruction accuracy could reach 0.989, and the surface roughness monitoring accuracy could reach 0.994 in the test set. The proposed method improved the interpretability of data-driven methods by incorporating domain knowledge from signal transmission, surface formation, and signal decomposition, met real-time monitoring requirements, reduced the monitoring model’s complexity, and was less costly and suitable for industrial applications.</div></div>","PeriodicalId":51124,"journal":{"name":"Mechanical Systems and Signal Processing","volume":"229 ","pages":"Article 112551"},"PeriodicalIF":7.9,"publicationDate":"2025-03-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143548667","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-05DOI: 10.1016/j.ymssp.2025.112517
Hongbo Wang , Yuting Qiao , Huan Liu , Yaguo Lei , Yanxin Zhang , Junyi Cao
Due to variable operational loads and uncertain model parameters induced by long-time joint wear, accurate and timely collision detection remains an open issue in the field of industrial robots. It is well known that model-based methods are popular for collision detection. However, traditional model-based methods will fail to detect collision because of time-varied model parameters caused by changing joint lubrication and working load. Therefore, a dynamic parameter adaptive collision detection method is proposed to improve the accuracy and robust of identifying collision in changing dynamic environments. The modeling error and the dynamic disturbance are considered to establish the parameter updating mechanism for reducing the momentum residual. Moreover, the dynamic parameters are adaptively updated to adjust the residuals during the contact-free situations of the robot. Consequently, a smaller stable range for the residual threshold is obtained to increase the performance of collision detection. Finally, experimental measurements of 1- and 7-degree-of-freedom (DoF) robots are performed to analyze the torques and momentum residuals under different conditions. The momentum residual of the 1-DoF manipulator decreases from 25 kg∙m/s to 9 kg∙m/s, and the joint momentum residual of the 7-DoF robot reduces from 8.2 kg∙m/s to 1.6 kg∙m/s. Compared to other methods, the proposed method has the lowest threshold deviation of 0.27 kg∙m/s. The results demonstrate that the proposed method can accurately detect collision under condition of variable load and model parameters.
{"title":"Parameter adaptive detection method of robot collisions under dynamic disturbance","authors":"Hongbo Wang , Yuting Qiao , Huan Liu , Yaguo Lei , Yanxin Zhang , Junyi Cao","doi":"10.1016/j.ymssp.2025.112517","DOIUrl":"10.1016/j.ymssp.2025.112517","url":null,"abstract":"<div><div>Due to variable operational loads and uncertain model parameters induced by long-time joint wear, accurate and timely collision detection remains an open issue in the field of industrial robots. It is well known that model-based methods are popular for collision detection. However, traditional model-based methods will fail to detect collision because of time-varied model parameters caused by changing joint lubrication and working load. Therefore, a dynamic parameter adaptive collision detection method is proposed to improve the accuracy and robust of identifying collision in changing dynamic environments. The modeling error and the dynamic disturbance are considered to establish the parameter updating mechanism for reducing the momentum residual. Moreover, the dynamic parameters are adaptively updated to adjust the residuals during the contact-free situations of the robot. Consequently, a smaller stable range for the residual threshold is obtained to increase the performance of collision detection. Finally, experimental measurements of 1- and 7-degree-of-freedom (DoF) robots are performed to analyze the torques and momentum residuals under different conditions. The momentum residual of the 1-DoF manipulator decreases from 25 kg∙m/s to 9 kg∙m/s, and the joint momentum residual of the 7-DoF robot reduces from 8.2 kg∙m/s to 1.6 kg∙m/s. Compared to other methods, the proposed method has the lowest threshold deviation of 0.27 kg∙m/s. The results demonstrate that the proposed method can accurately detect collision under condition of variable load and model parameters.</div></div>","PeriodicalId":51124,"journal":{"name":"Mechanical Systems and Signal Processing","volume":"229 ","pages":"Article 112517"},"PeriodicalIF":7.9,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143548665","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-05DOI: 10.1016/j.ymssp.2025.112493
Z. Li, R. Zhu, T. Verwimp, H. Wen, K. Gryllias
Estimating the Remaining Useful Life (RUL) of Rolling Element Bearings (REBs) is a very challenging task, complicating the optimal scheduling of shutdowns and maintenance operations in industry. In recent years, a number of prognostic methodologies have been proposed, mainly categorized into 3 groups: physical model-based, AI-based, and statistical-based methodologies. Statistical based methodologies, including the Kalman filter and its variants, are commonly used in RUL estimation due to their explainability and efficiency. However, their reliance on specific Health Indicators (HIs) often restricts their generalization capabilities for different scenarios. Additionally they often face accuracy and robustness issues due to the nonlinearity of the degradation trends in rotating machines. To overcome these limitations, this study introduces a prognostic methodology based on the Adaptive Kernel Kalman Filter (AKKF), which integrates HI extraction, anomaly and failure thresholds setting, parameters estimation in the data and kernel space, and uncertainty quantification. The proposed methodology is applied to three bearing degradation datasets: an in-house dataset captured on the KU Leuven gearbox prognostics test rig, and two publicly available datasets from the University of Ferrara (UNIFE) and Xi’an Jiaotong University (XJTU). The performance of the methodology is evaluated using different metrics and is compared with the Extended Kalman Filter (EKF). The results from the aforementioned three datasets indicate that the AKKF-based methodology is promising to be used for the highly nonlinear degradation trends of REBs, achieving good results. Moreover, several issues, including the HI selection, the degradation model selection and the filter selection are discussed in detail.
{"title":"Estimation of remaining useful life of rolling element bearings based on the Adaptive Kernel Kalman filter","authors":"Z. Li, R. Zhu, T. Verwimp, H. Wen, K. Gryllias","doi":"10.1016/j.ymssp.2025.112493","DOIUrl":"10.1016/j.ymssp.2025.112493","url":null,"abstract":"<div><div>Estimating the Remaining Useful Life (RUL) of Rolling Element Bearings (REBs) is a very challenging task, complicating the optimal scheduling of shutdowns and maintenance operations in industry. In recent years, a number of prognostic methodologies have been proposed, mainly categorized into 3 groups: physical model-based, AI-based, and statistical-based methodologies. Statistical based methodologies, including the Kalman filter and its variants, are commonly used in RUL estimation due to their explainability and efficiency. However, their reliance on specific Health Indicators (HIs) often restricts their generalization capabilities for different scenarios. Additionally they often face accuracy and robustness issues due to the nonlinearity of the degradation trends in rotating machines. To overcome these limitations, this study introduces a prognostic methodology based on the Adaptive Kernel Kalman Filter (AKKF), which integrates HI extraction, anomaly and failure thresholds setting, parameters estimation in the data and kernel space, and uncertainty quantification. The proposed methodology is applied to three bearing degradation datasets: an in-house dataset captured on the KU Leuven gearbox prognostics test rig, and two publicly available datasets from the University of Ferrara (UNIFE) and Xi’an Jiaotong University (XJTU). The performance of the methodology is evaluated using different metrics and is compared with the Extended Kalman Filter (EKF). The results from the aforementioned three datasets indicate that the AKKF-based methodology is promising to be used for the highly nonlinear degradation trends of REBs, achieving good results. Moreover, several issues, including the HI selection, the degradation model selection and the filter selection are discussed in detail.</div></div>","PeriodicalId":51124,"journal":{"name":"Mechanical Systems and Signal Processing","volume":"229 ","pages":"Article 112493"},"PeriodicalIF":7.9,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143548664","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-04DOI: 10.1016/j.ymssp.2025.112543
Wenkang Huang , Zifang Bian , Minghao Pan , Bohao Xiao , Ang Li , Haifeng Hu , Yongmin Yang , Fengjiao Guan
In modern aviation engines and turbomachinery, the vibration characteristics of blades are crucial for the safety and performance of machines. Blade tip timing (BTT) technology serves as a key method for monitoring and analyzing blade vibrations, enabling engineers to promptly identify potential faults and anomalies. However, traditional BTT methods often face challenges such as under-sampled data acquisition and improper sensor layout, which compromise the precision and robustness of frequency identification. To address these limitations, a novel displacement and velocity-based blade tip timing (D-V-BTT) method is proposed. Initially, a full pulse waveform-based method is applied to capture the vibration displacement and velocity of the blade as it passes each sensor. Subsequently, a new under-sampled sparse reconstruction method integrating displacement and velocity information is established. A non-convex regularization algorithm is used for reconstructing under-sampled vibration signals at a constant speed, enabling accurate frequency identification of blade tip vibration without prior conditions. Numerical simulations and experimental validations demonstrate the accuracy and validity of the proposed D-V-BTT method. It is demonstrated that the quantity of sensors can be reduced by the D-V-BTT method without decreasing the frequency discrimination accuracy after integrating the displacement and velocity information. Additionally, the D-V-BTT method can reduce the sensitivity to sensor layout.
{"title":"A novel sparse reconstruction method for under-sampled blade tip timing signals: Integrating vibration displacement and velocity","authors":"Wenkang Huang , Zifang Bian , Minghao Pan , Bohao Xiao , Ang Li , Haifeng Hu , Yongmin Yang , Fengjiao Guan","doi":"10.1016/j.ymssp.2025.112543","DOIUrl":"10.1016/j.ymssp.2025.112543","url":null,"abstract":"<div><div>In modern aviation engines and turbomachinery, the vibration characteristics of blades are crucial for the safety and performance of machines. Blade tip timing (BTT) technology serves as a key method for monitoring and analyzing blade vibrations, enabling engineers to promptly identify potential faults and anomalies. However, traditional BTT methods often face challenges such as under-sampled data acquisition and improper sensor layout, which compromise the precision and robustness of frequency identification. To address these limitations, a novel displacement and velocity-based blade tip timing (D-V-BTT) method is proposed. Initially, a full pulse waveform-based method is applied to capture the vibration displacement and velocity of the blade as it passes each sensor. Subsequently, a new under-sampled sparse reconstruction method integrating displacement and velocity information is established. A non-convex regularization algorithm is used for reconstructing under-sampled vibration signals at a constant speed, enabling accurate frequency identification of blade tip vibration without prior conditions. Numerical simulations and experimental validations demonstrate the accuracy and validity of the proposed D-V-BTT method. It is demonstrated that the quantity of sensors can be reduced by the D-V-BTT method without decreasing the frequency discrimination accuracy after integrating the displacement and velocity information. Additionally, the D-V-BTT method can reduce the sensitivity to sensor layout.</div></div>","PeriodicalId":51124,"journal":{"name":"Mechanical Systems and Signal Processing","volume":"229 ","pages":"Article 112543"},"PeriodicalIF":7.9,"publicationDate":"2025-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143548662","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-04DOI: 10.1016/j.ymssp.2025.112499
Jiduo Zhang , Robert Heinemann , Otto Jan Bakker , Siqi Li , Xiaoyu Xiao , Yixian Ding
The determination of minimum sufficient condition of machining signals in representing events is vital to not only the procedure of signal acquisition, transfer, and storage, but also the design, training and deployment of deep learning and its integrated system. This paper proposed a deep learning-based approach to accurately identify the crucial process incidence in drilling of hybrid stacks. The influence of three sampling elements: duration, frequency and phase on both signal fidelity and the performance of deep learning is investigated. Based on the property of translation-invariance of convolutional neural network (CNN) and statistical probability, a theory establishing continuous classification equivalent accuracy is proposed and proved through experimentation, which helps strike a balance between immediacy and accuracy in deep learning application. Both minimum sufficient duration (MSD) and frequency (MSF) are found to have a close relation with harmonics driven by machining operation and are determined by spindle rate in the drilling process. Through investigation into phase, the condition and property for minimum sufficient unit are examined. The work establishes the machine signal’s feasibility boundary for deep learning models, supporting a safe, compact, and lossless application of deep learning in industry especially where real-time and low latency is valued.
{"title":"Minimum sufficient signal condition of identifying process incidence in stacked drilling through deep learning","authors":"Jiduo Zhang , Robert Heinemann , Otto Jan Bakker , Siqi Li , Xiaoyu Xiao , Yixian Ding","doi":"10.1016/j.ymssp.2025.112499","DOIUrl":"10.1016/j.ymssp.2025.112499","url":null,"abstract":"<div><div>The determination of minimum sufficient condition of machining signals in representing events is vital to not only the procedure of signal acquisition, transfer, and storage, but also the design, training and deployment of deep learning and its integrated system. This paper proposed a deep learning-based approach to accurately identify the crucial process incidence in drilling of hybrid stacks. The influence of three sampling elements: duration, frequency and phase on both signal fidelity and the performance of deep learning is investigated. Based on the property of translation-invariance of convolutional neural network (CNN) and statistical probability, a theory establishing continuous classification equivalent accuracy is proposed and proved through experimentation, which helps strike a balance between immediacy and accuracy in deep learning application. Both minimum sufficient duration (MSD) and frequency (MSF) are found to have a close relation with harmonics driven by machining operation and are determined by spindle rate in the drilling process. Through investigation into phase, the condition and property for minimum sufficient unit are examined. The work establishes the machine signal’s feasibility boundary for deep learning models, supporting a safe, compact, and lossless application of deep learning in industry especially where real-time and low latency is valued.</div></div>","PeriodicalId":51124,"journal":{"name":"Mechanical Systems and Signal Processing","volume":"229 ","pages":"Article 112499"},"PeriodicalIF":7.9,"publicationDate":"2025-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143548660","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-04DOI: 10.1016/j.ymssp.2025.112541
Pengcheng Xu , Yaguo Lei , Zidong Wang , Naipeng Li , Xiao Cai , Ke Feng
Self-data-driven remaining useful life (RUL) prediction of machinery has gained significant attention in engineering due to its potential to reduce dependency on extensive training data during the prediction process. However, existing self-data-driven methods typically necessitate the storage and reuse of entire historical degradation data for updating prediction models, which limits their applicability in online and real-time scenarios. To overcome this limitation, this paper proposes a self-data-driven method for online RUL prediction using a recursive update strategy. Unlike conventional methods, the proposed recursive update strategy updates model parameters and selects the optimal model based on the latest monitoring data. A model base consisting of various degradation functions is initially established. During online prediction, model parameters are continuously updated in real-time using a sequential Bayesian algorithm, while the optimal degradation model is automatically identified using the proposed recursive Bayesian information criterion (RBIC). The effectiveness of this approach is validated through both simulation and experimental case studies. The results demonstrate that the proposed method not only achieves higher prediction accuracy but also requires less computation time compared to existing methods in online applications.
{"title":"A self-data-driven approach for online remaining useful life prediction of machinery using a recursive update strategy","authors":"Pengcheng Xu , Yaguo Lei , Zidong Wang , Naipeng Li , Xiao Cai , Ke Feng","doi":"10.1016/j.ymssp.2025.112541","DOIUrl":"10.1016/j.ymssp.2025.112541","url":null,"abstract":"<div><div>Self-data-driven remaining useful life (RUL) prediction of machinery has gained significant attention in engineering due to its potential to reduce dependency on extensive training data during the prediction process. However, existing self-data-driven methods typically necessitate the storage and reuse of entire historical degradation data for updating prediction models, which limits their applicability in online and real-time scenarios. To overcome this limitation, this paper proposes a self-data-driven method for online RUL prediction using a recursive update strategy. Unlike conventional methods, the proposed recursive update strategy updates model parameters and selects the optimal model based on the latest monitoring data. A model base consisting of various degradation functions is initially established. During online prediction, model parameters are continuously updated in real-time using a sequential Bayesian algorithm, while the optimal degradation model is automatically identified using the proposed recursive Bayesian information criterion (RBIC). The effectiveness of this approach is validated through both simulation and experimental case studies. The results demonstrate that the proposed method not only achieves higher prediction accuracy but also requires less computation time compared to existing methods in online applications.</div></div>","PeriodicalId":51124,"journal":{"name":"Mechanical Systems and Signal Processing","volume":"229 ","pages":"Article 112541"},"PeriodicalIF":7.9,"publicationDate":"2025-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143548663","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-04DOI: 10.1016/j.ymssp.2025.112518
Shihao Zhu , Yongbo Zhang , Zhonghan Li , Junling Wang , Shangwu Yuan
The robot manipulator of the lunar rover is a crucial device in lunar exploration missions, capable of performing sampling, experimental operations, and environmental analysis. To meet different task requirements, optimal trajectory planning is essential, and this planning relies on an accurate kinematic model. In the harsh environment of space, the kinematic parameters of the robot manipulator can change due to noise and structural damage, affecting the accuracy of trajectory planning. To address this, an energy-time-jerk optimal trajectory planning method for the robot manipulator with real-time parameter monitoring is proposed. The Sequential Quadratic Programming (SQP) algorithm is utilized for trajectory planning. Building on this, a new algorithm that combines the Extended Kalman Filter (EKF) and SQP algorithm (EKF-SQP) is introduced. Simulation results demonstrate that the proposed algorithm significantly improves the accuracy of the robot manipulator’s trajectory planning. Compared to existing methods, the integration of real-time parameter identification and compensation enhances precision by effectively reducing position errors of the end joint. By continuously updating the kinematic parameters in real-time, the algorithm ensures that the trajectory is dynamically re-planned, allowing the robot manipulator to reach the target position with higher accuracy.
{"title":"Multi-objective optimal trajectory planning method for robot manipulator with real-time parameters monitoring","authors":"Shihao Zhu , Yongbo Zhang , Zhonghan Li , Junling Wang , Shangwu Yuan","doi":"10.1016/j.ymssp.2025.112518","DOIUrl":"10.1016/j.ymssp.2025.112518","url":null,"abstract":"<div><div>The robot manipulator of the lunar rover is a crucial device in lunar exploration missions, capable of performing sampling, experimental operations, and environmental analysis. To meet different task requirements, optimal trajectory planning is essential, and this planning relies on an accurate kinematic model. In the harsh environment of space, the kinematic parameters of the robot manipulator can change due to noise and structural damage, affecting the accuracy of trajectory planning. To address this, an energy-time-jerk optimal trajectory planning method for the robot manipulator with real-time parameter monitoring is proposed. The Sequential Quadratic Programming (SQP) algorithm is utilized for trajectory planning. Building on this, a new algorithm that combines the Extended Kalman Filter (EKF) and SQP algorithm (EKF-SQP) is introduced. Simulation results demonstrate that the proposed algorithm significantly improves the accuracy of the robot manipulator’s trajectory planning. Compared to existing methods, the integration of real-time parameter identification and compensation enhances precision by effectively reducing position errors of the end joint. By continuously updating the kinematic parameters in real-time, the algorithm ensures that the trajectory is dynamically re-planned, allowing the robot manipulator to reach the target position with higher accuracy.</div></div>","PeriodicalId":51124,"journal":{"name":"Mechanical Systems and Signal Processing","volume":"229 ","pages":"Article 112518"},"PeriodicalIF":7.9,"publicationDate":"2025-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143548661","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号