No abstract

A new haptic sensor that is based on vibration produced by mechanical excitation from a clock coupled to a resonant cavity is presented. This sensor is intended to determine the chemical composition of liquid mixtures in a completely non-destructive method. In this case, a set of 23 samples of water, ethanol, and fructose mixtures has been used to simulate different kinds of alcoholic beverage. The spectral information from the vibrational absorption bands of liquid samples is analyzed by a Grouping Genetic Algorithm. An Extreme Learning Machine implements the fitness function that is able to classify the mixtures according to the concentration of ethanol and fructose. The 23 samples range from 0%–13% by volume of ethanol and from 0–3 g/L of fructose, all of them with different concentration. The new technique achieves an average classification accuracy of 96%.

This paper deals with the application of time-frequency analysis for transforming the received radar echoes in order to facilitate a neural network classification task. So as to compress the time-frequency representations maintaining most of the information, a feature extractor is designed. The proposed detector is compared with a single Multilayer Perceptron (MLP). The results show that time-frequency decompositions improve the performance of neural networks for slow fluctuating radar targets detection, specially for low values of Probability of False Alarm. The performance of the new detector is nearly independent on the Training-Signal-to-NoiseRatio (TSNR) and the training initial conditions.