Lucas Hermann Negri

Benchmark for Peak Detection Algorithms in Fiber Bragg Grating Interrogation and a New Neural Network for its Performance Improvement

This paper presents a benchmark for peak detection algorithms employed in fiber Bragg grating spectrometric interrogation systems. The accuracy, precision, and computational performance of currently used algorithms and those of a new proposed artificial neural network algorithm are compared. Centroid and gaussian fitting algorithms are shown to have the highest precision but produce systematic errors that depend on the FBG refractive index modulation profile. The proposed neural network displays relatively good precision with reduced systematic errors and improved computational performance when compared to other networks. Additionally, suitable algorithms may be chosen with the general guidelines presented.

An approach to improve the spatial resolution of a force mapping sensing system

This paper proposes a smart sensor system capable of detecting sparse forces applied to different positions of a metal plate. The sensing is performed with strain transducers based on fiber Bragg gratings (FBG) distributed under the plate. Forces actuating in nine squared regions of the plate, resulting from up to three different loads applied simultaneously to the plate, were monitored with seven transducers. The system determines the magnitude of the force/pressure applied on each specific area, even in the absence of a dedicated transducer for that area. The set of strain transducers with coupled responses and a compressive sensing algorithm are employed to solve the underdetermined inverse problem which emerges from mapping the force. In this configuration, experimental results have shown that the system is capable of recovering the value of the load distributed on the plate with a signal-to-noise ratio better than 12 dB, when the plate is submitted to three simultaneous test loads. The proposed method is a practical illustration of compressive sensing algorithms for the reduction of the number of FBG-based transducers used in a quasi-distributed configuration.

Efficient Computational Techniques in Bioimpedance Spectroscopy

Electrical Bioimpedance Analysis (BIA) is an important tool in the characterization of organic and biological material. For instance, its use may be mainly observed in the characterization of biological tissues in medical diagnosis (Brown, 2003), in the evaluation of organic and biological material suspensions in biophysics (Cole, 1968; Grimnes & Martinsen, 2008), in the determination of fat-water content in the body (Kyle et al., 2004) and in in vivo identification of cancerous tissues (Aberg et al., 2004), to name a few important works. It is also natural to have different computational approaches to bioimpedance systems since more complex computational techniques are required to reconstruct images in electrical impedance tomography (Holder, 2004), and this would open a myriad of other computational and mathematical questions based on inverse reconstruction problems.

Sparse Force Mapping System Based on Compressive Sensing

This paper reports the development and application of a reconstruction method based on differential evolution (DE) to solve an underdetermined tactile sensing system with quasi-distributed fiber sensors. The reconstruction relies on the coupled responses from eight fiber Bragg grating-based transducers. The sensing system is capable of locating and quantifying up to three loads simultaneously applied to a metallic plate divided into 16 regions. The signal reconstruction is performed using compressive sensing methods to infer the spatial distribution of the applied forces. A comparison between the implemented method based on DE and traditional sparse signal recovery schemes (LASSO, OMP, Robust-SL0, and CoSaMP) showed the better performance of the proposed algorithm in the demonstrated application.

Hardware embedded fiber sensor interrogation system using intensive digital signal processing

The description of an interrogation system for fiber Bragg grating sensors is reported. The full implementation in hardware of the required signal processing is proposed and made publicly available. The hardware description is implemented in a field programmable gate array (FPGA) development kit and the processing units allow one to control an optoelectronic interrogation system that uses the tunable filter method. Since the signal that drives the used Fabry-Perot filter (FFP) using a digital-to-analog converter (DAC) requires the generation of a triangular/saw-tooth waveform, the non-linear behavior of the DAC is compensated with a new methodology in this application using FPGA. When it operates controlled by a personal computer, off-board additional adaptive signal processing is used to suppress optical interference in an innovative way while removing undesired distortions in the signals caused by reflections in the optical circuit.

FBG refractometry and electrical impedance analysis in fuel samples characterization

This work reports the simultaneous use of electrical impedance spectroscopy and fiber Bragg grating (FBG) refractive index sensing in the estimation of the main components of specific fuel mixtures. Fuel samples containing gasoline, dehydrated ethanol, diesel, and kerosene were analyzed. Electrical impedance spectra and FBG sensor signals were registered for each mixture. Artificial Neural Networks (ANN) were used to estimate the ethanol concentration using the information from both sensors separately and to illustrate the methodology of fusing data from sensors that measure electrical permittivity at different frequency ranges, namely, an electrical impedance sensor and the etched FBG refractometric sensor. The behavior of the ANN to fuse data and the individual analysis of the sensor signals indicated that the joint use of the proposed techniques enhance the fuel estimation quality when compared to the usage of a singleton sensor.

Smartphone-based portable intensity modulated force sensor

This work proposes a low-cost force sensor, based on intensity modulation in an optical fibre. The transducer element is composed of a knot in a single mode fibre embedded to a silicone adhesive cuboid, and can be easily fabricated. A simple sensing scheme is devised by using a visible light source and a CCD camera of a smartphone, allowing implementation costs to be reduced. Experimental results have shown that the sensor presents a linear response and a standard uncertainty of 1:07N within the dynamical range from 0 to 30 N.

FBG interrogation and the benchmark for algorithms in the processing of experimental data

Fiber Bragg grating (FBG) spectrometric interrogation is widely used in strain, temperature, and refractive index sensing, where the acquired FBG sensor reflected signal has the spectrum peak monitored to infer a physical quantity. This monitoring, or peak detection, is subject to noise and different types of distortion, and many applications require the preprocessing of the signals to achieve the required precision and accuracy. This work extends the benchmarks already proposed in the literature with tests using specific experimental data with reduced signal to noise ratio and with systematic distortions in the FBG sensor. The obtained results show that specific conditions such as the distortion in the spectrum of the light reflected by the sensor can lead to poor performance of algorithms that reportedly show acceptable results under more ideal conditions. A metric for the performance evaluation of algorithms using experimental data in FBG interrogation is also presented.

Benchmark for standard and computationally intelligent peak detection algorithms for fiber Bragg grating sensors

Implementation and comparison of peak detection algorithms for fiber Bragg gratings spectra have been implemented and made publicly available. Benchmark experiments were performed by measuring accuracy, precision and efficiency of currently used algorithms, namely the centroid, least squares gaussian and polynomial fitting, and computational intelligence techniques using particle swarm optimization and perceptron neural network. Considering noisy apodized and uniform FBG spectra in the detection, it is shown that there is no general optimal algorithm for fast peak determination with high accuracy and precision, but it would be easier to choose quasi-optimal algorithms with the more general guidelines presented.

Tactile Sensor Array with Fiber Bragg Gratings in Quasi-Distributed Sensing

This work describes the development of a quasi-distributed real-time tactile sensing system with a reduced number of fiber Bragg grating-based sensors and reports its use with a reconstruction method based on differential evolution. The sensing system is comprised of six fiber Bragg gratings encapsulated in silicone elastomer to form a tactile sensor array with total dimensions of 60 × 80 mm, divided into eight sensing cells with dimensions of 20 × 30 mm. Forces applied at the central position of the sensor array resulted in linear response curves for the gratings, highlighting their coupled responses and allowing the application of compressive sensing. The reduced number of sensors regarding the number of sensing cells results in an undetermined inverse problem, solved with a compressive sensing algorithm with the aid of differential evolution method. The system is capable of identifying and quantifying up to four different loads at four different cells with relative errors lower than 10.5% and signal-to-noise ratio better than 12 dB.