Introducing a New Benchmarked Dataset for Mechanical Stop Detection of Stepper Motors

Goubeaud, Maxime; Grunert, Tim; Lutzenkirchen, Jan; JouBen, Philipp; Ghorban, Farzin; Kummert, Anton

doi:10.1109/icecs49266.2020.9294985

Cited by 3 publications

(3 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Extensive simulation tests were executed in MATLAB/Simulink under different operating conditions for the evaluation of the performance of the AC-TSKFC. The data sets used for conducting the experiment was taken from Goubeaud et al (2020). Simulation was performed using MATLAB/Simulink software for the following cases: (1) no-load condition, (2) abrupt load-torque changes, (3) system parameter uncertainty, (4) sudden phase interruption and (5) impact of noise.…”

Section: Simulation Resultsmentioning

confidence: 99%

Actor–critic learning-based adaptive fuzzy controller for stepper motor

Babu¹,

Chellaswamy²,

Geetha³

2023

Transactions of the Institute of Measurement and Control

View full text Add to dashboard Cite

This paper presents a combination of both the Takagi–Sugeno–Kang fuzzy controller and the advantage of reinforcement learning for the reduction in the effect of external disturbances and system uncertainties. A neural network is used for the implementation of a critic network; the parameters are updated using the Lyapunov criteria for avoiding local minima problems. The fewer learning parameters used helps online tuning of the controller by the critic network based on the reward function. MATLAB/Simulink is used for simulation of the proposed system under different conditions in the investigation of the performance of the proposed controller. The actor parameters are updated based on the change in the reward function error and a control function. Five different conditions, namely, no-load condition, sudden load-torque changes, system parameter uncertainty, sudden phase interruption, and the impact of noise have been studied. The proposed method was found to meet the high-speed response with the tracking error of ±2.23% for the tracking reference trajectory and tracking error of ±2.5% under constant load-torque disturbance. The test results were compared with two benchmark controllers for the verification of the effectiveness of the proposed controller. Simulation results showed the proposed method providing an adaptive and precise speed response. Therefore, it is suitable for non-linear and uncertain applications.

show abstract

Section: Simulation Resultsmentioning

confidence: 99%

Actor–critic learning-based adaptive fuzzy controller for stepper motor

Babu¹,

Chellaswamy²,

Geetha³

2023

Transactions of the Institute of Measurement and Control

View full text Add to dashboard Cite

show abstract

“…Stepper motors For stepper motor monitoring, there is a dataset published by Goubeaud et al [19]. This includes measurements of current, voltage, and vibration (translational acceleration).…”

Section: Data Setsmentioning

confidence: 99%

“…The target variable is the operating mode of the stepper motor, which differentiates between clockwise and counterclockwise operation, as well as operation in the normal range and beyond the mechanical stop. A detailed description of the experimental setup and the data acquisition is given in the original publication [19].…”

Section: Data Setsmentioning

confidence: 99%

Autoencoder-based Representation Learning from Heterogeneous Multivariate Time Series Data of Mechatronic Systems

Kortmann¹,

Fehsenfeld²,

Wielitzka³

2021

Preprint

View full text Add to dashboard Cite

Sensor and control data of modern mechatronic systems are often available as heterogeneous time series with different sampling rates and value ranges. Suitable classification and regression methods from the field of supervised machine learning already exist for predictive tasks, for example in the context of condition monitoring, but their performance scales strongly with the number of labeled training data. Their provision is often associated with high effort in the form of person-hours or additional sensors. In this paper, we present a method for unsupervised feature extraction using autoencoder networks that specifically addresses the heterogeneous nature of the database and reduces the amount of labeled training data required compared to existing methods. Three public datasets of mechatronic systems from different application domains are used to validate the results.

show abstract