Search

Increasing the number of sensors in a gas identification system generally improves its performance as this will add extra features for analysis. However, this affects the computational complexity, especially if the identification algorithm is to be implemented on a hardware platform. Therefore, feature reduction is required to extract the most important information from the sensors for processing. In this paper, linear discriminant analysis (LDA) and principal component analysis (PCA)-based feature reduction algorithms have been analyzed using the data obtained from two different types of gas sensors, i.e., seven commercial Figaro sensors and in-house fabricated $4 \times 4$ tin-oxide gas array sensor. A decision tree-based classifier is used to examine the performance of both the PCA and LDA approaches. The software implementation is carried out in MATLAB and the hardware implementation is performed using the Zynq system-on-chip (SoC) platform. It has been found that with the $4 \times 4$ array sensor, two discriminant functions (DF) of LDA provide 3.3% better classification than five PCA components, while for the seven Figaro sensors, two principal components and one DF show the same performances. The hardware implementation results on the programmable logic of the Zynq SoC shows that LDA outperforms PCA by using 50% less resources as well as by being 11% faster with a maximum running frequency of 122 MHz. © 2001-2012 IEEE.

This paper proposes a hardware/software co-design approach using the Zynq platform for the implementation of an electronic nose (EN) system based on principal component analysis (PCA) as a dimensionality reduction technique and decision tree (DT) as a classification algorithm using a 4x4 in-house fabricated sensor. The system was successfully trained and simulated in MATLAB environment prior to the implementation on the Zynq platform. High level synthesis was carried out on the proposed designs using different optimization directives including loop unrolling, array partitioning and pipelining. © 2016 IEEE.

The concept of passive Radio Frequency Identification (RFID) sensor tag is introduced to remove the dependency of current RFID platforms on battery life. In this paper, a gas identification system is presented using passive RFID sensor tag along with the processing unit. The RFID system is compliant to Electronics Product Code Generation 2 (EPC-Gen2) protocol in 902-928 MHz ISM band. Whereas the processing unit is implemented and analyzed in software and hardware platforms. The software platform uses MATLAB, whereas a High Level Synthesis (HLS) tool is used to implement the processing unit on a Zynq platform. Moreover, two sets of different gases are used along with Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) based feature reduction approaches to analyze in detail the best feature reduction approach for efficient classification of gas data. It is found that for the first set of gases, 90 % gases are identified using first three principal components, which is 7 % more efficient than LDA. However in terms of hardware overhead, LDA requires 50 % less hardware resources than PCA. The classification results for the second set of gases reveal that 91 % of gas classification is obtained using LDA and first four PCA, while LDA requires 52 % less hardware resources than PCA. The RFID tag used for transmission is implemented in 0.13 μm CMOS process, with simulated average power consumption of 2.6 μW from 1.2 V supply. ThingMagic M6e embedded reader is used for RFID platform implementation. It shows an output power of 31.5 dBm which allows a read range up to 9 meters.

This paper presents a gas identification system which is part of an ongoing research project aiming to develop a low power reconfigurable self-calibrated multi-sensing platform for gas applications where real time parameters such as temperatures and gas concentrations are monitored. The gas identification system is based on decision tree classifier; training is performed in MATLAB environment. It is first carried out using the steady states extracted from the raw data obtained from the sensors, and then using transformed data by applying principal component analysis. The data used for training is collected from a 16-Array SnO2 gas sensor. The sensor array is exposed to three types of gases (CO, C2H6O and H2) at ten different concentrations (20, 40, 60, 80, 100, 120, 140, 160, 180 and 200ppm). The resulting models are implemented in C and synthesized using Vivado High Level Synthesis tool for rapid prototyping on the heterogeneous Zynq platform.

Tin-oxide based gas sensor requires an operating temperature typically in the range of 200 °C to 400 °C and its performance dependents on this temperature. In this paper a deep examination has been made to analyze the best operating temperature suitable for gas identification application in which an array of sensors is used along with an appropriate feature reduction algorithm. The two most common feature reduction algorithms for gas classification are principal component analysis (PCA) and linear discriminant analysis (LDA); both of them have been used in this analytical work. The feature reduction is followed by a binary decision tree (BDT) or K-nearest neighbor (KNN) based classifier. Results obtained with data from an array of sensors used for detecting C6H6, CH2O, CO, NO2 and SO2 indicates that at 400 °C the BDT can classify 100% of gases after LDA based feature reduction, whereas KNN can classify 100% of gases at 200 °C and 300 °C using data before and after feature reduction. Furthermore, experimental results from the given sensor data suggest that with and without considering the operating temperature the BDT can classify 96% of gases using first four LDA components. While KNN can classify 98% to 99% of gases using first four LDA or first five PCA components of resulting data obtained after feature reduction. Thus, after LDA-based feature reduction both classifiers provide superior identification with minimum number of components. © 2006-2015 Asian Research Publishing Network (ARPN).

The Zynq-7000 based platforms are increasingly being used in different applications including image and signal processing. The Zynq system on chip (SoC) architecture combines a processing system based on a dual core ARM Cortex processor with a programmable logic (PL) based on a Xilinx 7 series field programmable gate arrays (FPGAs). Using the Zynq platform, real-time hardware acceleration can be performed on the programmable logic and controlled by a software running on the ARM-based processing system (PS). In this paper, a design and implementation of a gas identification system on the Zynq platform which shows promising results is presented. The gas identification system is based on a 16- Array SnO2 gas sensor, principal component analysis (PCA) for dimensionality reduction and decision tree (DT) for classification. The main part of the system that will be executed on the PL for hardware acceleration takes the form of an IP developed in C and synthesized using Vivado High Level Synthesis for the conversion from C to register transfer level, a hardware design for the entire system that allows the execution of the IP on the PL and the remaining parts of the identification system on the PS is developed in Vivado using IP Integrator. The communication between the processing system and programmable logic is performed using advanced extensible interface protocol (AXI). A software application is developed and executed on the ARM processor to control the hardware acceleration on the programmable logic of the previously designed IP core and the board is programmed using Software Development Kit. The maximum accuracy achieved by the system to classify three types of gases CO, C2H6O and H2 is 96.66%.