Sensorless Estimation

Sensorless estimation aims to infer system states (e.g., speed, position, contact force) without relying on dedicated sensors, leveraging readily available signals like current measurements or voltage waveforms. Recent research focuses on applying machine learning techniques, including support vector machines, artificial neural networks (ANNs), and deep learning models, to extract relevant information from these signals and accurately estimate the desired parameters. This approach offers significant advantages in cost reduction, robustness to sensor failure, and miniaturization, impacting diverse fields such as motor control, robotics, and traffic management by enabling more efficient and reliable systems.