Data Coverage Assessment on Machine-Learning based Digital Twin for Diagnosis in a Nearly Autonomous Management and Control System

Abstract This article describes the design and the development of a novel six-legged robotic walking machine named SphereWalker. The six legs are arranged into pairs, and each pair of legs is supported and actuated by a single spherical four-bar mechanism. Two of the four-bar mechanisms are operated in a synchronous fashion, while the middle one is operated at 180 deg out of phase with respect to the other two. A physical prototype has been built, a digital twin has been generated, an actuation and control system has been designed, and the technology has been patented.

Download Full-text

Development and Assessment of a Nearly Autonomous Management and Control System During a Single Loss of Flow Accident

Volume 1: Beyond Design Basis; Codes and Standards; Computational Fluid Dynamics (CFD); Decontamination and Decommissioning; Nuclear Fuel and Engineering; Nuclear Plant Engineering ◽

10.1115/icone2020-16908 ◽

2020 ◽

Author(s):

Linyu Lin ◽

Paridhi Athe ◽

Pascal Rouxelin ◽

Nam Dinh ◽

Jeffrey Lane

Keyword(s):

Knowledge Base ◽

Confusion Matrix ◽

Data Generation ◽

Digital Twin ◽

Control Functions ◽

Digital Twins ◽

Twin Models ◽

Autonomous Management ◽

And Performance ◽

And Control

Abstract In this work, a Nearly Autonomous Management and Control (NAMAC) system is designed to diagnose the reactor state and provide recommendations to the operator for maintaining the safety and performance of the reactor. A three layer-hierarchical workflow is suggested to guide the design and development of the NAMAC system. The three layers in this workflow corresponds to knowledge base, digital twin developmental layer (for different NAMAC functions), and NAMAC operational layer. Digital twin in NAMAC is described as knowledge acquisition system to support different autonomous control functions. Therefore, based on the knowledge base, a set of digital twin models is trained to determine the plant state, predict behavior of physical components or systems, and rank available control options. The trained digital twin models are assembled according to NAMAC operational workflow to support decision-making process in selecting the optimal control actions during an accident scenario. To demonstrate the capability of the NAMAC system, a case study is designed, where a baseline NAMAC is implemented for operating a simulator of the Experimental Breeder Reactor II (EBR-II) during a single loss of flow accident. Training database for development of digital twin models is obtained by sampling the control parameters in the GOTHIC data generation engine. After the training and testing, the digital twins are assembled into a NAMAC system according to the operational workflow. This NAMAC system is coupled with the GOTHIC plant simulator, and a confusion matrix is generated to illustrate the accuracy and robustness of implemented NAMAC system. It is found that within the training databases, NAMAC can make reasonable recommendations with zero confusion rate. However, when the scenario is beyond the training cases, the confusion rate increases, especially when the scenarios are more severe. Therefore, a discrepancy checker is added to detect unexpected reactor states and alert operators for safety-minded actions.

Download Full-text

Seismic RTDT: Real-Time Digital Twin for Boosting Performance of Seismic Operations

Volume 7A: Ocean Engineering ◽

10.1115/omae2019-95885 ◽

2019 ◽

Cited By ~ 1

Author(s):

Severin Sadjina ◽

Stian Skjong ◽

Armin Pobitzer ◽

Lars T. Kyllingstad ◽

Roy-Jostein Fiskerstrand ◽

...

Keyword(s):

Machine Learning ◽

Real Time ◽

State Of The Art ◽

Seismic Survey ◽

Digital Twin ◽

Detection Systems ◽

Early Results ◽

Actual Equipment ◽

Automation Systems ◽

And Control

Abstract Here, we present the R&D project Real-Time Digital Twin for Boosting Performance of Seismic Operations, which aims at increasing the overall operational efficiency of seismic vessels through digitisation and automation. The cornerstone in this project is the development of a real-time digital twin (RTDT) — a sophisticated mathematical model and state estimator of all the in-sea seismic equipment, augmented with real-time measurements from the actual equipment. This provides users and systems on-board the vessel with a live digital representation of the state of the equipment during operations. By combining the RTDT with state-of-the-art methods in machine learning and control theory, the project will develop new advisory and automation systems that improve the efficiency of seismic survey operations, reduce the risk of equipment damage, improve health monitoring and fault detection systems, and improve the quality of the seismic data. This will lead to less unproductive time, reduced costs, reduced fuel consumption and reduced emissions for a given operational scope. The main focus in this paper is the presentation of today’s challenges in offshore seismic surveys, and how state-of-the-art technology can be adopted to improve various operations. We discuss how simulation technology, machine learning and live sensor measurements can be integrated in on-board decision support and automation systems, and highlight the importance of such systems for designing the complex, autonomous offshore vessels of the future. Finally, we present some early results from the project in the form of two brief case studies.

Download Full-text

Development and Assessment of Data-Driven Digital Twins in a Nearly Autonomous Management and Control System for Advanced Reactors

10.1115/1.0000482v ◽

2021 ◽

Author(s):

Linyu Lin ◽

Pascal Rouxelin ◽

Paridhi Athe ◽

Jeff Lane ◽

Nam Dinh

Keyword(s):

Control System ◽

Data Driven ◽

Digital Twins ◽

Autonomous Management ◽

And Control

Download Full-text

Development and Assessment of a Nearly Autonomous Management and Control System During a Single Loss of Flow Accident

10.1115/1.0000462v ◽

2021 ◽

Author(s):

Linyu Lin ◽

Paridhi Athe ◽

Pascal Rouxelin ◽

Jeff Lane ◽

Nam Dinh

Keyword(s):

Control System ◽

Autonomous Management ◽

And Control

Download Full-text

Exploration of Flight State and Control System Parameters for Prediction of Helicopter Loads via Gamma Test and Machine Learning Techniques

Annals of Information Systems - Real World Data Mining Applications ◽

10.1007/978-3-319-07812-0_18 ◽

2014 ◽

pp. 359-385 ◽

Cited By ~ 2

Author(s):

Catherine Cheung ◽

Julio J. Valdés ◽

Matthew Li

Keyword(s):

Machine Learning ◽

Control System ◽

Machine Learning Techniques ◽

Gamma Test ◽

System Parameters ◽

Learning Techniques ◽

And Control

Download Full-text

IMPROVING STUDENTS PERFORMANCE PREDICTION USING MACHINE LEARNING AND SYNTHETIC MINORITY OVERSAMPLING TECHNIQUE

Journal of Engineering and Sustainable Development ◽

10.31272/jeasd.25.6.6 ◽

2021 ◽

Vol 25 (6) ◽

pp. 56-64

Author(s):

Nibras Z. Salih ◽

◽

Walaa Khalaf ◽

Keyword(s):

Machine Learning ◽

Control System ◽

Performance Prediction ◽

Supervised Classification ◽

Learning Algorithm ◽

University Education ◽

Machine Learning Algorithm ◽

And Control ◽

The Relationship ◽

Academic Failures

Classification under supervision is the most common job that performed by machine learning. However, most Educators were worried about the rising evidence of student academic failures in university education. So, this study presents a supervised classification strategy of machine learning algorithm using an actual dataset contains 44 students, fourteen attributes for three previous academic years. We have proposed features that show the relationship among three main subjects which are, calculus, mathematical analysis, and control system in the education course. The objective of this study is to identify the student’s failure in the control system subject and to enhance his performance by Multilayer Perceptron (MLP) algorithm. The dataset is unbalanced, which causes overfitting of the results. Synthetic Minority Oversampling Technique has applied to a dataset for obtaining balance dataset using Weka tool. Several standard metrics used to evaluate the classifier results. Therefore, the suitable results occurred after applying SMOTE with an accuracy of 76.9%.

Download Full-text

IOT-HML: A Hybrid Machine Learning Technique for IoT Enabled Industrial Monitoring and Control System

10.21203/rs.3.rs-365980/v1 ◽

2021 ◽

Author(s):

Kota P.N. ◽

Chandak A.S. ◽

Patil B.P.

Keyword(s):

Machine Learning ◽

Control System ◽

Information Security ◽

Monitoring And Control ◽

Machine Learning Technique ◽

Monitoring And Control System ◽

Learning Technique ◽

Industrial Monitoring ◽

Hybrid Machine ◽

And Control

Abstract Industry 4.0 makes manufacturers more vulnerable to current challenges and makes it easier to adapt to market changes. This will increase the speed of innovation, make it more customer-oriented and lead to faster design processes. It is essential to focus on monitoring and controlling the production system before complex accidents occur. Moreover, an industrial control system facing information security problems in recent times because of the nature of IoT which affects the evaluation of abnormal predication. To overcome above research gaps, we shift to industrial 4.0 which combine IoT and mechanism learning for industrial monitor and manage. We propose a hybrid machine learning technique for IoT enabled industrial monitoring and control system (IoT-HML). Here, we concentrate both information security issues with accurate monitoring and control system. The first section of proposed IoT-HML system is to introduce the cat induced wheel optimization (IWO) algorithm for cluster formation. The process consists of clustering and cluster head (CH) selection. The source node forward information to destination through CH only which avoids the unwanted data loss and improve the security, because the information travel through trusted path. For route selection process, we utilize the cuckoo search algorithm to compute the optimal best path among multiples. In second section, we illustrate a coach and player learned neural network (CP-LNN) for monitoring the industrial and prevent from accidents by basic control strategies. Finally, the proposed IoT-HML system can evaluate with different set of data’s to prove the effectiveness.

Download Full-text