Jaime Vitola

A Sensor Data Fusion System Based on k-Nearest Neighbor Pattern Classification for Structural Health Monitoring Applications

Civil and military structures are susceptible and vulnerable to damage due to the environmental and operational conditions. Therefore, the implementation of technology to provide robust solutions in damage identification (by using signals acquired directly from the structure) is a requirement to reduce operational and maintenance costs. In this sense, the use of sensors permanently attached to the structures has demonstrated a great versatility and benefit since the inspection system can be automated. This automation is carried out with signal processing tasks with the aim of a pattern recognition analysis. This work presents the detailed description of a structural health monitoring (SHM) system based on the use of a piezoelectric (PZT) active system. The SHM system includes: (i) the use of a piezoelectric sensor network to excite the structure and collect the measured dynamic response, in several actuation phases; (ii) data organization; (iii) advanced signal processing techniques to define the feature vectors; and finally; (iv) the nearest neighbor algorithm as a machine learning approach to classify different kinds of damage. A description of the experimental setup, the experimental validation and a discussion of the results from two different structures are included and analyzed.

Distributed Piezoelectric Sensor System for Damage Identification in Structures Subjected to Temperature Changes

Structural health monitoring (SHM) is a very important area in a wide spectrum of fields and engineering applications. With an SHM system, it is possible to reduce the number of non-necessary inspection tasks, the associated risk and the maintenance cost in a wide range of structures during their lifetime. One of the problems in the detection and classification of damage are the constant changes in the operational and environmental conditions. Small changes of these conditions can be considered by the SHM system as damage even though the structure is healthy. Several applications for monitoring of structures have been developed and reported in the literature, and some of them include temperature compensation techniques. In real applications, however, digital processing technologies have proven their value by: (i) offering a very interesting way to acquire information from the structures under test; (ii) applying methodologies to provide a robust analysis; and (iii) performing a damage identification with a practical useful accuracy. This work shows the implementation of an SHM system based on the use of piezoelectric (PZT) sensors for inspecting a structure subjected to temperature changes. The methodology includes the use of multivariate analysis, sensor data fusion and machine learning approaches. The methodology is tested and evaluated with aluminum and composite structures that are subjected to temperature variations. Results show that damage can be detected and classified in all of the cases in spite of the temperature changes.

Fast content-based audio retrieval algorithm

Fingerprinting is one of the most used techniques for searching and identification audio with a wide spectrum of applications. Different algorithms defines different fingerprint extraction and the match techniques, with different efficiency, computational load, robustness, response time and location search. Nowadays music audio retrieval faces two main challenges in order to be efficient: robustness and speed. This article proposes a fast algorithm to the audio content-based retrieval with the fingerprint technique, based on the extraction of the frequency features of the audio and a hash function. Experiments determined a high success rate and a response time lower than other techniques, optimal to real time applications like monitoring radio stations or songs identifying.

Fast parallel audio fingerprinting implementation in reconfigurable hardware and GPUs

One of the main challenges that Music Information Retrieval (MIR) faces is performance. This paper presents an algorithm based on fingerprinting techniques implemented in a low-cost embedded reconfigurable platform. This fast algorithm is even faster when implemented in parallel for a GPU platform. The hit rate of the implementations is practically 100% and the response time is two times faster than the response time of a top class PC, which means MIR times of up to 65 audio tracks in real time.

Data-Driven Methodologies for Structural Damage Detection Based on Machine Learning Applications

Structural health monitoring (SHM) is an important research area, which interest is the damage identification process. Different information about the state of the structure can be obtained in the process, among them, detection, localization and classification of damages are mainly studied in order to avoid unnecessary maintenance procedures in civilian and military structures in several applications. To carry out SHM in practice, two different approaches are used, the first is based on modelling which requires to build a very detailed model of the structure, while the second is by means of data-driven approaches which use information collected from the structure under different structural states and perform an analysis by means of data analysis . For the latter, statistical analysis and pattern recognition have demonstrated its effectiveness in the damage identification process because real information is obtained from the structure through sensors installed permanently to the observed object allowing a real-time monitoring. This chapter describes a damage detection and classification methodology, which makes use of a piezoelectric active system which works in several actuation phases and that is attached to the structure under evaluation, principal component analysis, and machine learning algorithms working as a pattern recognition methodology. In the chapter, the description of the developed approach and the results when it is tested in one aluminum plate are also included.

Design and control of an exoskeleton in rehabilitation tasks for lower limb

According to statistical data provided by the National Administrative Department of Statistics (DANE), during 2005, 29.32% of Colombias disabled population had problems with their legs related to moving or walking. In order to contribute and help these people, medical science and engineering have been working together to provide solutions that can improve the quality of life of injured people. In this sense, it is possible to design and implement electromechanical devices to assist and facilitate movements in rehabilitation processes. Such devices can keep detailed records of movements performed, patient history, speed, force, muscle interaction, among others; also allows physiotherapists to have complete control of rehabilitation processes, improving it and offering a lot more possible treatments by analyzing the summarized collected data. Rehabilitation focused exoskeletons work as guide and support in physical therapy, those make sure that the treated patient performs correctly all its exercises, in case that the patient cannot perform the movement by himself such devices helps the patient to finish the exercise, improving the effectiveness of the therapy and reducing the time it takes to recover lost faculties. As a contribution to rehabilitation processes, this paper proposes the design of an exoskeleton by considering biome-chanical models involving the most possible characteristics of the human body to reduce the differences between the mathematical model and the real behavior of the body segments; That proposed model is used to constraint and design a controller in master-slave configuration to assist and ensure soft movements improving rehabilitation processes involving flection and extension movements in the sagittal plane of the lower limbs.

Cartesian genetic algorithm for boolean synthesis with power consumption restriction

The use of evolutionary algorithms in the boolean synthesis is an interesting technique to generate hardware structures with multiple restrictions. However, one characteristic of these algorithms is their high computational load. This paper presents the implementation of a parallel cartesian genetic programming (CGP) for boolean synthesis on a FPGA-CPU based platform. Power consumption and critical path restrictions were included into the algorithm in order to generate structures to solve any problem. As results a 2-bit comparator is presented, as well as response time and data transitions probability.