Advances in hardware development have created robotic hands able to mimic the appearance and functionality of the human counterpart (i.e., Anthropomorphic Robotic Hand (ARH)). Non-invasive signal identification and classification to control the robotic hands remains a challenge for real scenarios. Myoelectric sensors provide valuable information of muscle activity. However, the subject’s gesture identification through myoelectric signals is a complex task. In this work, we implemented a non-invasive EMG-Based control system for an ARH based on artificial neural nets. Electromyography (EMG) data were pre-processed using filtering and scaling. Then, for each subject, 20 relevant features were extracted and selected. Next, eight machine learning architectures were used for gesture identification. Finally, the control system was implemented using custom scripts in Python, communicating with a six DoF robotic hand using an ESP32 microcontrolled card. As a result of the evaluation with the classifiers, the Multi-layer Perceptron obtained the highest accuracy of 84.78%. The implemented control system was open-loop, based on multiple inputs with a single output for each servomotor, with two seconds of delay. The EMG data classification based on artificial neural networks could be an essential resource to increase the number of independent movements performed by robotic hands.
This document presents a Smith predictor based LAMDA (Learning Algorithm for Multivariable Data Analysis) Sliding-Mode Control "LSMC" applied to a Mixing Tank with Dead Time. The Smith Predictor structure is used in order to improve the performance of the controller due to the non-linearity generated by the dead time produced in the mixing tank. Smith Predictor requires an approximation of the plant model for good performance, however, in plants with variable dynamics, a robust controller capable of controlling the system while maintaining stability is required, for which the LSMC proposal is used. The proposed scheme is validated when disturbances are introduced and when reference changes are made, which causes the system to change its operating point. The results obtained by the controller are promising and show that it is a viable proposal to implement it due to the ease of design.
This work presents the Internal Model Control (IMC) of the tank water level in a small-scale industrial multipurpose water tank process. The variables involved in this process are the water flow and water level, internal tank temperature, and pressure. A conventional PID controller is also implemented for comparison purposes. Both control algorithms, IMC and PID, are implemented using a ControlLogix L73 PLC and the Studio 5000 software, where the computation of overshoot, steady-state error, and settling time are performed. Finally, both controllers are tested with disturbances added to the process variables.
High-speed applications represent a challenge in electrical machine design due to the thermal, mechanical and electromagnetic factors involved. Consequently, a good selection of the machine topology is crucial to satisfy the requirements of these applications. Nevertheless, devising appropriate selection criteria requires comprehensive comparative studies between prominent topologies, in order to provide relevant information of their suitability for different operating scenarios. This article presents a comparative study based on the finite element analysis (FEA) of two high-speed electrical machines sharing a common stator design: a synchronous reluctance machine and a solid rotor induction machine. Both with a mechanical speed of 20,000 r/min. The results show that, for the same mean torque objective, the SynRM provides lower core losses and higher overall efficiency, at the cost of developing higher torque ripple. Additionally, the SR-IM suffers from significantly lower mechanical stress, creating headroom to withstand higher operation speeds without compromising the rotor structure integrity.
Timely detection of potential engine failures is essential for preventive maintenance of equipment. This avoids damages to machines, potential security risks and reduce high costs involved in repairing or replacing them. The audio emitted by machines during operation is an indicator of a potential engine failure. Audio-based fault classification and detection is a research area dominated by solutions based on AutoEncoder Neural Networks. In this work, an end-to-end approach is proposed for classifying electric motor noise. The methodology includes a pre-processing phase to transform the audio time series into a matrix representation of spectrogram images, capturing time and frequency domain data, then classifying them using an Ensemble of Convolutional Neural Network models. To evaluate the performance a set of metrics have been measured and compared, reaching a performance accuracy above 97% and AUC above 99%. Therefore, acoustic registry of electric motors could also be a powerful source of useful data to include in predictive maintenance data analysis.
