scholarly journals Real-Time Evaluation of the Mechanical Performance and Residual Life of a Notching Mold using Embedded PVDF Sensors and SVM Criteria

Sensors ◽  
2019 ◽  
Vol 19 (23) ◽  
pp. 5123
Author(s):  
Chang ◽  
Huang ◽  
Chung
Keyword(s):  
Real Time ◽  
Polyvinylidene Fluoride ◽  
Mechanical Performance ◽  
Residual Life ◽  
Sheet Material ◽  
Sensor Data ◽  
Fault Classification ◽  
Support Vector ◽  
Extensive Experience ◽  
Related Data

The geometric tolerance of notching machines used in the fabrication of components for induction motor stators and rotators is less than 50 µm. The blunt edges of worn molds can cause the edge of the sheet metal to form a burr, which can seriously impede assembly and reduce the efficiency of the resulting motor. The overuse of molds without sufficient maintenance leads to wasted sheet material, whereas excessive maintenance shortens the life of the punch/die plate. Diagnosing the mechanical performance of die molds requires extensive experience and fine-grained sensor data. In this study, we embedded polyvinylidene fluoride (PVDF) films within the mechanical mold of a notching machine to obtain direct measurements of the reaction forces imposed by the punch. We also developed an automated diagnosis program based on a support vector machine (SVM) to characterize the performance of the mechanical mold. The proposed cyber-physical system (CPS) facilitated the real-time monitoring of machinery for preventative maintenance as well as the implementation of early warning alarms. The cloud server used to gather mold-related data also generated data logs for managers. The hyperplane of the CPS-PVDF was calibrated using a variety of parameters pertaining to the edge characteristics of punches. Stereo-microscopy analysis of the punched workpiece verified that the accuracy of the fault classification was 97.6%.

Online Real-Time Quality Monitoring in Additive Manufacturing Processes Using Heterogeneous Sensors

2015 ◽  
Vol 137 (6) ◽  
Author(s):  
Prahalad K. Rao ◽  
Jia (Peter) Liu ◽  
David Roberson ◽  
Zhenyu (James) Kong ◽  
Christopher Williams
Keyword(s):  
Additive Manufacturing ◽  
Real Time ◽  
Dirichlet Process ◽  
Failure Modes ◽  
Evidence Theory ◽  
Sensor Data ◽  
Support Vector ◽  
Fused Filament Fabrication ◽  
Detection Approach ◽  
Real Time Identification

The objective of this work is to identify failure modes and detect the onset of process anomalies in additive manufacturing (AM) processes, specifically focusing on fused filament fabrication (FFF). We accomplish this objective using advanced Bayesian nonparametric analysis of in situ heterogeneous sensor data. Experiments are conducted on a desktop FFF machine instrumented with a heterogeneous sensor array including thermocouples, accelerometers, an infrared (IR) temperature sensor, and a real-time miniature video borescope. FFF process failures are detected online using the nonparametric Bayesian Dirichlet process (DP) mixture model and evidence theory (ET) based on the experimentally acquired sensor data. This sensor data-driven defect detection approach facilitates real-time identification and correction of FFF process drifts with an accuracy and precision approaching 85% (average F-score). In comparison, the F-score from existing approaches, such as probabilistic neural networks (PNN), naïve Bayesian clustering, support vector machines (SVM), and quadratic discriminant analysis (QDA), was in the range of 55–75%.

Real-time optimization using gradient adaptive selection and classification from infrared sensors measurement for esterification oleic acid with glycerol

2017 ◽  
Vol 10 (2) ◽  
pp. 130-144 ◽  
Author(s):  
Iwan Aang Soenandi ◽  
Taufik Djatna ◽  
Ani Suryani ◽  
Irzaman Irzaman
Keyword(s):  
Real Time ◽  
Sensor Data ◽  
Necessary Condition ◽  
Support Vector ◽  
Adaptive Selection ◽  
Content Type ◽  
The Real ◽  
Time Optimization ◽  
Starting Point ◽  
Real Time Optimization

Purpose The production of glycerol derivatives by the esterification process is subject to many constraints related to the yield of the production target and the lack of process efficiency. An accurate monitoring and controlling of the process can improve production yield and efficiency. The purpose of this paper is to propose a real-time optimization (RTO) using gradient adaptive selection and classification from infrared sensor measurement to cover various disturbances and uncertainties in the reactor. Design/methodology/approach The integration of the esterification process optimization using self-optimization (SO) was developed with classification process was combined with necessary condition optimum (NCO) as gradient adaptive selection, supported with laboratory scaled medium wavelength infrared (mid-IR) sensors, and measured the proposed optimization system indicator in the batch process. Business Process Modeling and Notation (BPMN 2.0) was built to describe the tasks of SO workflow in collaboration with NCO as an abstraction for the conceptual phase. Next, Stateflow modeling was deployed to simulate the three states of gradient-based adaptive control combined with support vector machine (SVM) classification and Arduino microcontroller for implementation. Findings This new method shows that the real-time optimization responsiveness of control increased product yield up to 13 percent, lower error measurement with percentage error 1.11 percent, reduced the process duration up to 22 minutes, with an effective range of stirrer rotation set between 300 and 400 rpm and final temperature between 200 and 210°C which was more efficient, as it consumed less energy. Research limitations/implications In this research the authors just have an experiment for the esterification process using glycerol, but as a development concept of RTO, it would be possible to apply for another chemical reaction or system. Practical implications This research introduces new development of an RTO approach to optimal control and as such marks the starting point for more research of its properties. As the methodology is generic, it can be applied to different optimization problems for a batch system in chemical industries. Originality/value The paper presented is original as it presents the first application of adaptive selection based on the gradient value of mid-IR sensor data, applied to the real-time determining control state by classification with the SVM algorithm for esterification process control to increase the efficiency.

Machine Learning Based IoT Geriatric Fall Intelligent System in Pandemic (Preprint)

2021 ◽  
Author(s):  
Gowri R ◽  
Rathipriya R
Keyword(s):  
Machine Learning ◽  
Blood Pressure ◽  
Heart Rate ◽  
Real Time ◽  
Old Age ◽  
Sensor Data ◽  
Stochastic Gradient Descent ◽  
Support Vector ◽  
The Real ◽  
Early Fall

UNSTRUCTURED In the current pandemic, there is lack of medical care takers and physicians in hospitals and health centers. The patients other than COVID infected are also affected by this scenario. Besides, the hospitals are also not admitting the old age peoples, and they are scared to approach hospitals even for their basic health checkups. But, they have to be cared and monitored to avoid the risk factors like fall incidence which may cause fatal injury. In such a case, this paper focuses on the cloud based IoT gadget for early fall incidence prediction. It is machine learning based fall incidence prediction system for the old age patients. The approaches such as Logistic Regression, Naive Bayes, Stochastic Gradient Descent, Decision Tree, Random Forest, Support Vector Machines, K-Nearest Neighbor and ensemble learning boosting techniques, i.e., XGBoost are used for fall incidence prediction. The proposed approach is first tested on the benchmark activity sensor data with different features for training purpose. The real-time vital signs like heart rate, blood pressure are recorded and stored in cloud and the machine learning approaches are applied to it. Then tested on the real-time sensor data like heart rate and blood pressure data of geriatric patients to predict early fall.

Sensor-Based Online Process Fault Detection in Additive Manufacturing

2015 ◽  
Author(s):  
Prahalad K. Rao ◽  
Jia (Peter) Liu ◽  
David Roberson ◽  
Zhenyu (James) Kong
Keyword(s):  
Additive Manufacturing ◽  
Real Time ◽  
Dirichlet Process ◽  
Failure Modes ◽  
Sensor Data ◽  
Support Vector ◽  
Fused Filament Fabrication ◽  
Loop Control ◽  
Am Processes ◽  
Non Parametric

The objective of this work is to identify failure modes and detect the onset of process anomalies in Additive Manufacturing (AM) processes, specifically focusing on Fused Filament Fabrication (FFF). We accomplish this objective using advanced Bayesian non-parametric analysis of in situ heterogeneous sensor data. The proposed method can ultimately lead to intelligent decision making and closed loop control in AM processes. Experiments are conducted on a desktop FFF machine (MakerBot Replicator 2X) instrumented with a heterogeneous sensor array including thermocouples, accelerometers, an infrared temperature sensor, and a real-time miniature video borescope. FFF process failures are detected online using the non-parametric Bayesian Dirichlet Process (DP) mixture model and evidence theory based on the experimentally acquired sensor data. This sensor data-driven defect detection approach facilitates real-time identification and correction of FFF process drifts with an accuracy and precision approaching 85% (average F-score). In comparison, the F-score from existing approaches, such as Probabilistic Neural Networks, Naïve Bayesian Clustering, Support Vector Machines, and Quadratic Discriminant Analysis was in the range of 55% to 75%.

scholarly journals A Survey on Distributed Fibre Optic Sensor Data Modelling Techniques and Machine Learning Algorithms for Multiphase Fluid Flow Estimation

Sensors ◽  
2021 ◽  
Vol 21 (8) ◽  
pp. 2801
Author(s):  
Hasan Asy’ari Arief ◽  
Tomasz Wiktorski ◽  
Peter James Thomas
Keyword(s):  
Machine Learning ◽  
Fluid Flow ◽  
Data Analysis ◽  
Real Time ◽  
Machine Learning Algorithms ◽  
Sensor Data ◽  
Support Vector ◽  
Measurement Technology ◽  
Fibre Optic ◽  
Multiphase Fluid

Real-time monitoring of multiphase fluid flows with distributed fibre optic sensing has the potential to play a major role in industrial flow measurement applications. One such application is the optimization of hydrocarbon production to maximize short-term income, and prolong the operational lifetime of production wells and the reservoir. While the measurement technology itself is well understood and developed, a key remaining challenge is the establishment of robust data analysis tools that are capable of providing real-time conversion of enormous data quantities into actionable process indicators. This paper provides a comprehensive technical review of the data analysis techniques for distributed fibre optic technologies, with a particular focus on characterizing fluid flow in pipes. The review encompasses classical methods, such as the speed of sound estimation and Joule-Thomson coefficient, as well as their data-driven machine learning counterparts, such as Convolutional Neural Network (CNN), Support Vector Machine (SVM), and Ensemble Kalman Filter (EnKF) algorithms. The study aims to help end-users establish reliable, robust, and accurate solutions that can be deployed in a timely and effective way, and pave the wave for future developments in the field.

Binary Spectrum Feature for Improved Classifier Performance

2020 ◽  
Author(s):  
Nalika Ulapane ◽  
Karthick Thiyagarajan ◽  
sarath kodagoda
Keyword(s):  
Machine Learning ◽  
Classification Performance ◽  
Feature Reduction ◽  
Sensor Data ◽  
Machine Learning Techniques ◽  
Support Vector ◽  
Svm Classifier ◽  
Monitoring Task ◽  
Classifier Performance ◽  
Spectrum Feature

<div>Classification has become a vital task in modern machine learning and Artificial Intelligence applications, including smart sensing. Numerous machine learning techniques are available to perform classification. Similarly, numerous practices, such as feature selection (i.e., selection of a subset of descriptor variables that optimally describe the output), are available to improve classifier performance. In this paper, we consider the case of a given supervised learning classification task that has to be performed making use of continuous-valued features. It is assumed that an optimal subset of features has already been selected. Therefore, no further feature reduction, or feature addition, is to be carried out. Then, we attempt to improve the classification performance by passing the given feature set through a transformation that produces a new feature set which we have named the “Binary Spectrum”. Via a case study example done on some Pulsed Eddy Current sensor data captured from an infrastructure monitoring task, we demonstrate how the classification accuracy of a Support Vector Machine (SVM) classifier increases through the use of this Binary Spectrum feature, indicating the feature transformation’s potential for broader usage.</div><div><br></div>

scholarly journals Study on Quality Prediction of 2219 Aluminum Alloy Friction Stir Welding Based on Real-Time Temperature Signal

Materials ◽  
2021 ◽  
Vol 14 (13) ◽  
pp. 3496
Author(s):  
Haijun Wang ◽  
Diqiu He ◽  
Mingjian Liao ◽  
Peng Liu ◽  
Ruilin Lai
Keyword(s):  
Aluminum Alloy ◽  
Friction Stir Welding ◽  
Prediction Model ◽  
Real Time ◽  
Support Vector ◽  
Weld Quality ◽  
Prediction Ability ◽  
Friction Stir ◽  
Temperature Signal ◽  
2219 Aluminum Alloy

The online prediction of friction stir welding quality is an important part of intelligent welding. In this paper, a new method for the online evaluation of weld quality is proposed, which takes the real-time temperature signal as the main research variable. We conducted a welding experiment with 2219 aluminum alloy of 6 mm thickness. The temperature signal is decomposed into components of different frequency bands by wavelet packet method and the energy of component signals is used as the characteristic parameter to evaluate the weld quality. A prediction model of weld performance based on least squares support vector machine and genetic algorithm was established. The experimental results showed that, when welding defects are caused by a sudden perturbation during welding, the amplitude of the temperature signal near the tool rotation frequency will change significantly. When improper process parameters are used, the frequency band component of the temperature signal in the range of 0~11 Hz increases significantly, and the statistical mean value of the temperature signal will also be different. The accuracy of the prediction model reached 90.6%, and the AUC value was 0.939, which reflects the good prediction ability of the model.

Sign in / Sign up

Export Citation Format

Share Document