TFM: Design strategies for detecting action potentials in actions based on movements
Posted on by anacp

This work is located in the studies of the brain and their signals. The puspose is to know when someone wants to make a movement. Thus, it might help to people that actually are not able to move a member of their body or more. Mainly, it is focused in the design of strategies for detection of action potentials or spikes when a movement wants to be made. This study is not looking for action potentials form, it is looking for patterns and characteristics that allow to recognize the movement. Although there are action potentials covered by the signals taken from the electrodes, but they are unavailable.

To accomplish the objective, it is used the EEG signals of a public data base. It is selected the ones related to the movement of the hands, concretely, the movement of open and close the fist. Signal sources of noise that dirty the signal are analyzed, they are called artifacts, and then, filtering stage comes, giving the signals of below for movement and no movement.

slotMov15slotRest15

Now, possible algorithms are checked. It is decided to use the Wavelet transform and the way in which it obtains the energy of the signal. Thanks to the calculation of Wavelet energy in 22 subjects, it is reached to the conclusion that Wavelet energy for movement is higher than for no movement. So, electrodes that comply with this condition at 100% are 4.

The final algorithm is implemented three features: correlation, a parameter that gives a relation between two signals, their energy range and their energy average. It could be said that algorithm has two parts: a training stage and a decision stage. Inside decision part, there are three algorithms: ProMove, ProMove + improve and Logic. The basic difference among ProMove and Logic is an or (||) and an and (&). The improve is based on empiric knowledge.

commonstages

 

systemcomplete

Final conclusions show that the signals between subjects are very changing. Therefore, same algorithm is not useful for everybody. To some subjects, the successful probability is very high (92,86% – 1 fail), while for others is more low than what is expected (50% – 7 fail). With these test, the importance in the length of the signals is reflected, because if signals for subjects with more than 3 fails are inversely processed, the fails are reduced. The most useful algorithm for a larger number of subjects is ProMove + improve.

TFG: Design and implementation of modules for a low cost radar system.
Posted on by b105

In the last years many low-cost radar modules have appeared on the market, allowing the implementation of this technology in a large number of applications, such as medical applications or people detection.

The B105 Electronic Systems Lab developed a prototype for the control, management and processing of the signals generated by these transceiver radars. The aim of the project is increasing the system versatility while correcting the problems it presented. So, the first step consisted on analysing the existing system and evaluating the aspects in which it could be improved.
radar

Then, the focus shifted to the design of several circuits that allowed to digitally change the amplification and filtering of the analog signals of the radar module. The circuit that modulates the radar transceiver was also modified to make it configurable. Once the designs were made a printed circuit board (PCB) was developed and manufactured.

An update of the existing software was needed since the hardware modules have been modified. Functions that handle the different amplification and filtering configurations of the system were developed. Also, a communication that would allow sending orders from a computer to the module was added. This communication allows the modification of the parameters during the operation of the system. The parameters include the amplification, the filtering characteristics, as well as the modulation parameters of the radar transceiver.

TFG: DESIGN AND IMPLEMENTATION OF AN ELECTRONIC MEASUREMENT SYSTEM FOR DEFORMATIONS ON MATERIALS
Posted on by b105

Currently there are various systems to measure deformations such as optical systems of video or laser as well as direct contact systems,  which can be classified in mechanical and electrical systems. The strain gauges belong to this last group. These gauges are devices that resemble a rectangular sheet whose dimensions typically span just a few centimeters length. In its interior there is a conductive or semiconducting wire with the form of a grid which has the ability to vary its electrical resistance when it is deformed. Compared to other technologies, strain gauges offer a much more affordable price and its use is very simple. Given their increasing perfection, they can offer benefits similar to other technologies and that is why the interest they receive has been increasing considerably, giving rise to a wide gauge market with a great variety of features and prices.

This work was born with the goal of developing a system that measures deformations based on its use for different materials with certain levels of precision, accuracy and reliability, as well as designing it as generic as possible to allow the use of any gauge that is offered in the market.

The design of the system consists of a Discovery board that tries to sample the signal coming from the gauges for its later transformation and processing. The data are sended and displayed on the computer screen through a program that reads the USB port.

The study covers different measurement techniques based on the use of different configurations that connect the gauges to the Discovery board for a comparison of results and effectiveness with each method. It also seeks to analyze the performance of different types of strain gauges with different characteristics.

TFG: DESIGN AND IMPLEMENTATION OF A DEMONSTRATOR SYSTEM FOR COGNITIVE WIRELESS SENSOR NETWORKS
Posted on by b105

 

A wireless sensor network (WSN) is a kind of network that contains nodes communicating wireless. It has sensors that allow to obtain information directly from the environment in order to learn or act on it.

Since the use of this wireless networks is growing, it appears the need of creating cognitive networks which are able to learn from the environment and adapt themselves efficiently.

The B105 Electronic Systems Lab research group developed a test-bench containing some nodes called ‘cognitive New Generation Device (cNGD)’. Currently, each of them is programmed by connecting it physically to a computer. However, this situation produces a lot of problems, like the required time to perform the node programming or the necessity of reprogramming a node that is out of reach. This is the main reason why a wireless programming method becomes very handy.

The aim of this project is to improve the already available Bootloader getting a better reception and to manage the available random access memory. For this purpose, a Wake On Radio (WOR) board was used to wake up a specific cNGD node and then work on this node independently. However, some modifications were required due to hardware and software limitations. Even though the node has three transceivers on ISM (Industrial, scientist and medical) free bands, it was used the 434 MHz band for the WOR and the 2.45GHz band for the Bootloader due to its speed.

Captura de pantalla 2018-01-23 a las 17.13.26

In addition, an graphical interface was implemented for the test-bench in order to see the status of the cNGD nodes, the code transmission and the connection processes. It also has another tab for the choice of the cNGD nodes to wake up and reprogram. This interface is a web application with the server side implemented with the Python programming language, so we can reach it only with an internet connection.

Captura de pantalla 2018-01-23 a las 17.15.06

Finally, some tests were run to verify the expected behavior of the test-bench. These test are documented at the end of the memoir.

Obtención de indicadores de fatiga mediante el electroencefalograma.
Posted on by Gis

Desde el proyecto Simbiosys, buscamos nuevas formas de detección de fatiga. Puesto que el sistema está destinado a ser usado en un simulador para conductores de vehículos, se busca que sea lo menos intrusivo posible, para facilitar el movimiento y comodidad del conductor.

Con este fin se está desarrollando un sistema multisensor con una parte importante de investigación como es la detección de distintos estados de fatiga mediante la actividad cerebral del conductor.

Para la obtención del electroencefalograma (EEG) se eligió un casco con un único electrodo, ya que los EEG convencionales presentan más de veinte electrodos, lo cual sería muy intrusivo para el conductor.

21_electrodes_of_International_10-20_system_for_EEG.svg

Tras la obtención de la señal en bruto del cerebro, el sistema se basa en la detección de la cantidad de energía que existe en las diferentes bandas del cerebro. En este caso las bandas de interés serán la banda alpha, betha y tetha, todas ellas relacionadas con estados de cansancio, fatiga o sueño.

5d3d36b587c91d50c495eeb3ed8f5ca2

El sistema consta de dos partes diferenciadas, basadas en machine learning. En la primera parte se obtiene las características – la energía de cada banda- del sujeto en estado de consciencia (no fatigado) para formar dos clústeres.
2clus_2

El objetivo es generar dos esferas que engloben todas las características en este estado, de tal forma que, si en la segunda parte del algoritmo se obtiene alguna característica que no pertenece a los clústeres, se considera una anomalía. Será la acumulación de anomalías durante un periodo de tiempo la que nos indique la presencia de fatiga en el sujeto.