This paper presents method of vibration based continuous monitoring system and analysis using machine learning approach. The reliable and effective performance of a braking system is fundamental operation of most vehicles. This study provides insight of fault diagnosis of hydraulic braking system by vibration analysis. A hydraulic brake system test rig was fabricated. The vibration signals were acquired from a piezoelectric transducer for both good as well as faulty conditions of brakes. The statistical parameters are extracted and the good features that discriminate different faulty conditions were formed using decision tree. This study presents the data model (J48, C4.5 decision tree algorithm) for fault diagnosis through descriptive statistical features extracted from vibration signals of good and faulty conditions of hydraulic brakes. The classification results of decision tree algorithm for fault diagnosis of a hydraulic brake system are presented. The model built can be used for condition monitoring of hydraulic brake system. The classification accuracy for decision tree algorithm using statistical features is found to be 97.45%.
A number of capacitance readout systems presented in the literature and covered by recent patents are briefly described while a biosensor array readout system developed by the authors is presented in this paper. This biosensor array readout system is based on a charge sensitive amplifier that estimates accurately the difference between a selfconfigurable reference capacitance and a biosensor. In the specific implementation, the self-configurable reference capacitance is automatically configured to a value between 0 and 630pF in steps of 10pF. The configuration of the reference capacitance is controlled by a simple finite state machine that can be implemented using a simple CPLD (Complex Programmable Logic Device) such as Altera EPM7064SLC44-10. The proposed system can be used to measure an absolute biosensor capacitance of up to 640pF with a theoretical sensitivity of 2.5fF. The 2.5fF resolution can be achieved using a low cost/high accuracy 12-bit ADC (Analogue/Digital Converter) like a one that has been described by the authors in a recent patent instead of a 17-bit ADC that would be required to measure the biosensor capacitance value with the same accuracy if a constant 320pF reference capacitor had been used. For this reason, the proposed architecture leads to a lower area/power and higher accuracy implementation compared to a read out circuit that would use a fixed reference capacitance. Although the design of the proposed system was inspired by the development of a biosensor array readout system, a similar architecture can be used by any sensor application that should support a broad capacitive input range with high resolution.
In this paper, a new patent technique was developed to support the geographical positioning planning and RF power calculation of Radio Base Stations (RBS) for the Cellular Mobile Systems using geoprocessed images of a mountain urban environment. The region of study is a neighborhood of Belo Horizonte City, in Minas Gerais State, Brazil. The mountain urban environment was considered in order to map the electromagnetic waves propagation for better coverage of cellular mobile service having the frequency band range of 800-900 MHz. The contributions of this research are: the implementation of a Geographic Information System (GIS) in order to exploit and simulate possibilities of geopositioning for the RBS to guarantee efficient area coverage, along with a quality service for radiofrequency system; the evaluation of power prediction levels for radiofrequency signals using Okumura-Hata propagation model, in order to obtain an optimum algorithm to register the geopositioning of the RBS. As a result, the minimum distances among fifteen suggested RBS and their coverage mapping regions are shown.
Rapid development of broadband multimedia services in the Internet and the demand for higher quality video implies that Quality of Experience (QoE) measurement for transmitted video sequences over packet error prone channels is a necessity. Due to the intrinsic properties of packet error prone channels in imposing quality degradation on the transmitted media and also, costly and time-consuming nature of subjective quality measurement techniques, exact modeling of the impact of packet loss on the measured video quality based on the on-line objective measurement methods is an important task. In the current work, a low-complexity objective video quality measurement algorithm is being developed in which by considering various factors that may affect the video quality, can estimate the subjective video quality with acceptable accuracy. Then, the performance of the proposed objective algorithm is compared with the popular subjective techniques. The simulation results verify that the proposed algorithm has high level of accuracy. On the other hand, the proposed algorithm has low-complexity which makes it suitable for online implementations. Some important and related patents in the field of QoE measurement and monitoring in the networks are being investigated.
Significant scientific efforts are constantly invested in the attempts to develop photonic circuitry integrated in microelectronic chip. Such circuitry should be capable of realizing fundamental processing building blocks such as modulators (or transistors) and logic gates in order to improve the performance of an optics communication short and/or long range links. The improvement is to be obtained in the sense of reduced power dissipation, lower noise level and increased information bandwidth. This paper reviews some of the technologies involved in those efforts. Specifically we focus on silicon based integrated nanophotonic modulation devices (based upon plasma dispersion as well as fundamental non linear processes in silicon) as well as passive devices for directing and spectrally filtering light via silicon nanophotonic crystal structures. The patents mentioned in this review relate to the realization of nanophotonic logic gates for data processing, they are related to photonic nano communication link apparatus, apparatus and method for switching, modulation and dynamic control of light transmission using photonic crystals and to modulator-based photonic chip-tochip interconnections for dense three-dimensional multichip module integration.
A measurement system is presented for evaluating the characteristics of both sides of an LED wafer using electroluminescence (EL). Integrating spheres (IS), photodiodes, pico-ammeters, spectrometers were installed in the front and rear sides to measure optical characteristics with minimum EL loss. A probe was connected to a source meter with a thin wire and attached using a transparent fixture in the front IS. The EL characteristics in a measuring point were peak wavelength, FWHM (full width at half maximum), forward current, forward voltage and reverse current. EL images were created by combining the motion of a stage and the EL characteristics. The correlation between the EL and the chip-level tests and the repeatability test showed that the dual measurement is useful for LED manufacturing. Our patents were applied to install the probe and create the EL image.
The wavelet transform converts a signal in one domain into another for better understanding of the structure and features of any given signal. Recently, performance of various denoising methods has been analyzed and the result shows that Fourth order moment based denoising provides better performance, in the same study the proposed modified Fourth order moment denoising provides better Bit Error Rate (BER) performance than previous work. Wiener filtering is an excellent method to estimate a noisy signal. In this work, the modified Fourth order moment based denoising technique is used in Wavelet domain (WD) Wiener filtering algorithm to estimate less noisy Wavelet coefficient for denoising purpose. The modified Fourth order moment reduces noise by allowing wavelet coefficients having noise below a certain threshold level. Computer simulation is performed to investigate the performance of the proposed technique. The simulation result shows that the joint use of WD-Wiener filter in modified Fourth order moment based denoising provides better BER performance than that of modified Fourth order moment based denoising. Finally, Wiener filter was also applied to detect noise free image through the same noisy fading channel.