Data-Driven Modeling: Concept, Techniques, Challenges and a Case Study

Habib, Maki K.; Ayankoso, Samuel A.; Nagata, Fusaomi

doi:10.1109/icma52036.2021.9512658

Cited by 18 publications

(7 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…where e is the error from (7), N is the number of samples, η is the number of nonzero elements in Ξ λ . 4) Model Selection : The gain coefficient Ξ λ with the best criteria has the lowest AIC score and can be obtained using min function in MATLAB.…”

Section: ) Model Selectionmentioning

confidence: 99%

“…By using the mathematical equations of the models enabled us to gain an in-depth understanding of the characteristics and dynamics of the robot manipulator, which in turn enabled the design of effective and optimized control systems [1], [2] . The dynamic modeling of a robot manipulator can be performed using three approaches: manual modeling [3], [4], computer-based modeling [5], [6], and hybrid modeling [7]. Robot manipulator dynamics modeling using manual or analytical approaches, such as the Lagrange-Euler [8] and Newton-Euler formulation [9], require a high level of knowledge of dynamics and physics laws.…”

Section: Introductionmentioning

confidence: 99%

“…The hybrid approach is a combination of analytical and computer-based modeling approaches, which is also known as parameter estimation, where the model equations are obtained through an analytical approach, and the parameters are estimated through the process of a computer-based modeling approach. However, owing to the presence of analytical modeling, parameter estimation is less effective when applied to robot manipulator dynamics modeling, because obtaining complex equations through an analytical approach is difficult [7], particularly if the degrees of freedom, which are determined by the number of joints, are increased. System identification refers to the use of a computer to build mathematical models of a system based on the measured input and output data [12].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Enhanced Equation Discovery of 3-DoF Robotic Manipulator Dynamics Using LASSO Model Selection Criteria With Variable Segregation Algorithm

Istiqphara,

Wahyunggoro,

Cahyadi

2024

IEEE Access

View full text Add to dashboard Cite

The challenge in controlling a manipulator robot is to model the system to obtain an efficient control system design. One approach that can be used to model the dynamics of a manipulator robot is data-driven modeling. However, in its implementation, data-driven modeling is highly sensitive to sensor noise, which significantly affects the accuracy of the system identification. In addition, the existing approach yields only a generalized form of the differential equation for each joint, which has not been divided into inertial, Coriolis, and gravitational variables that can be used for other purposes. In this study, a LASSO model selection criteria with a variable segregation algorithm (LMSCVS) is proposed to derive the dynamic equation of a 3-DoF manipulator robot, segregating the generalized form variables into Coriolis and centrifugal, inertia, and gravitational variables. Additionally, a Dynamic Expression Nonlinearization(DEx-N) algorithm is introduced to generate nonlinear candidates more efficiently to express the dynamics of the robot manipulator. The experimental results on the ROB3 hardware demonstrate that the proposed method successfully discovers mathematical equations, resulting in higher accuracy and sparsity compared to the previous method. The processing time of the proposed method is also significantly faster. Based on these results, the proposed method has a better performance in identifying real systems that usually have noise in the sensor data and in discovering the equation of robot manipulator dynamics for broader purposes.

show abstract

Section: ) Model Selectionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Enhanced Equation Discovery of 3-DoF Robotic Manipulator Dynamics Using LASSO Model Selection Criteria With Variable Segregation Algorithm

Istiqphara,

Wahyunggoro,

Cahyadi

2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Data-driven model-based methods utilize the historical or current data of an object under certain functional constraints to establish a type that can approximate the implicit mapping mechanism between object data and lifespan for prediction [ 14 ]. The main steps of data-driven methods are data preprocessing, feature extraction, feature selection, model selection, and model evaluation [ 15 – 17 ].…”

Section: Introductionmentioning

confidence: 99%

An integrated method for the leakage fault mode diagnosis and life prediction of the reactor coolant pump

Shao,

Kang,

et al. 2024

PLoS ONE

View full text Add to dashboard Cite

The reactor coolant pump is a key equipment in a nuclear power plant. If the leakage exceeds a certain threshold, it may cause reactor overheating and shutdown. The reactor coolant pump leakage fault usually has two problems: corrosion and scaling. Accurately and efficiently diagnosing the leakage fault mode as early as possible and predicting its remaining useful life (RUL) are important for taking timely maintenance measures. In this paper, an integrated method is proposed. First, the cross-sectional area of the first seal is extracted as a fault indicator. The motivation is that corrosion may enlarge the cross-sectional area, and scaling may reduce the cross-sectional area. Based on the fluid mechanics theory, an integrated model with several uncertain parameters is established among the cross-sectional area, temperature, and leakage at the inlet and outlet of the first seal. In the diagnosing process, a modified change-detection method is proposed to detect the starting point of degradation. Then, the unknown parameters in the previous relation are estimated, and the degrading data before the starting point of degradation are used to diagnose the leakage fault mode. Second, a time-series model of the autoregressive integrated moving average (ARIMA) is established to predict the remaining useful life based on the degrading data after the starting point of degradation. Finally, the leakage degrading data from six reactor coolant pumps of a nuclear power plant is used to perform the leakage fault mode diagnosis and life prediction with degradation point detection error rates not exceeding 4%, fault mode diagnosis correction rates 100% and practical RUL predicting results, which proves that the proposed integrated method is accurate and efficient. The proposed integrated method combines the advantages of both the physical model diagnosis and the data-driven model diagnosis and innovatively make use of the quantity of flow from the output side of the primary pump as the monitoring indicator and the cross-sectional area as the characteristic index together to diagnose the leakage fault mode happened to the seal and predict its RUL, which can meet the needs of actual operation and maintenance to ensure a healthy and stable operation of the pump and prevent unexpected shutdowns of nuclear power plants and serious accidents.

show abstract

“…Machine learning (ML) techniques [4], and its subset, that is deep learning (DL) methods [5] such as Neural networks (NNs), have received significant attention in the field of science and engineering due to their capability in modeling nonlinear and complex systems [6]. The modeling of a dynamic system using NNs or ML techniques can be implemented in two ways: purely data-driven or physics-informed data-driven implementation.…”

Section: Introductionmentioning

confidence: 99%

Time-Series Machine Learning Techniques for Modeling and Identification of Mechatronic Systems with Friction: A Review and Real Application

Ayankoso,

Olejnik

2023

Electronics

View full text Add to dashboard Cite

Developing accurate dynamic models for various systems is crucial for optimization, control, fault diagnosis, and prognosis. Recent advancements in information technologies and computing platforms enable the acquisition of input–output data from dynamical systems, resulting in a shift from physics-based methods to data-driven techniques in science and engineering. This review examines different data-driven modeling approaches applied to the identification of mechanical and electronic systems. The approaches encompass various neural networks (NNs), like the feedforward neural network (FNN), convolutional neural network (CNN), long short-term memory (LSTM), transformer, and emerging machine learning (ML) techniques, such as the physics-informed neural network (PINN) and sparse identification of nonlinear dynamics (SINDy). The main focus is placed on applying these techniques to real-world problems. A real application is presented to demonstrate the effectiveness of different machine learning techniques, namely, FNN, CNN, LSTM, transformer, SINDy, and PINN, in data-driven modeling and the identification of a geared DC motor. The results show that the considered ML techniques (traditional and state-of-the-art methods) perform well in predicting the behavior of such a classic dynamical system. Furthermore, SINDy and PINN models stand out for their interpretability compared to the other data-driven models examined. Our findings explicitly show the satisfactory predictive performance of six different ML models while also highlighting their pros and cons, such as interpretability and computational complexity, using a real-world case study. The developed models have various applications and potential research areas are discussed.

show abstract

Data-Driven Modeling: Concept, Techniques, Challenges and a Case Study

Cited by 18 publications

References 36 publications

Enhanced Equation Discovery of 3-DoF Robotic Manipulator Dynamics Using LASSO Model Selection Criteria With Variable Segregation Algorithm

Enhanced Equation Discovery of 3-DoF Robotic Manipulator Dynamics Using LASSO Model Selection Criteria With Variable Segregation Algorithm

An integrated method for the leakage fault mode diagnosis and life prediction of the reactor coolant pump

Time-Series Machine Learning Techniques for Modeling and Identification of Mechatronic Systems with Friction: A Review and Real Application

Contact Info

Product

Resources

About