Explore the vast range of titles available.

A new LeNet-5 gas identification convolutional neural network structure for electronic noses is proposed and developed in this paper. Inspired by the tremendous achievements made by convolutional neural networks in the field of computer vision, the LeNet-5 was adopted and improved for a 12-sensor array based electronic nose system. Response data of the electronic nose to different concentrations of CO, CH4 and their mixtures were acquired by an automated gas distribution and test system. By adjusting the parameters of the CNN structure, the gas LeNet-5 was improved to recognize the three categories of CO, CH4 and their mixtures omitting the concentration influences. The final gas identification accuracy rate reached 98.67% with the unused data as test set by the improved gas LeNet-5. Comparison with results of Multiple Layer Perceptron neural networks and Probabilistic Neural Network verifies the improvement of recognition rate while with the same level of time cost, which proved the effectiveness of the proposed approach.

This paper proposes a new method of mixed gas identification based on a convolutional neural network for time series classification. In view of the superiority of convolutional neural networks in the field of computer vision, we applied the concept to the classification of five mixed gas time series data collected by an array of eight MOX gas sensors. Existing convolutional neural networks are mostly used for processing visual data, and are rarely used in gas data classification and have great limitations. Therefore, the idea of mapping time series data into an analogous-image matrix data is proposed. Then, five kinds of convolutional neural networks—VGG-16, VGG-19, ResNet18, ResNet34 and ResNet50—were used to classify and compare five kinds of mixed gases. By adjusting the parameters of the convolutional neural networks, the final gas recognition rate is 96.67%. The experimental results show that the method can classify the gas data quickly and effectively, and effectively combine the gas time series data with classical convolutional neural networks, which provides a new idea for the identification of mixed gases.

The inherent cross-sensitivity of semiconductor gas sensors makes them extremely challenging to accurately detect mixed gases. In order to solve this problem, this paper designed an electronic nose (E-nose) with seven gas sensors and proposed a rapid method for identifying CH4, CO, and their mixtures. Most reported methods for E-nose were based on analyzing the entire response process and employing complex algorithms, such as neural network, which result in long time-consuming processes for gas detection and identification. To overcome these shortcomings, this paper firstly proposes a way to shorten the gas detection time by analyzing only the start stage of the E-nose response instead of the entire response process. Subsequently, two polynomial fitting methods for extracting gas features are designed according to the characteristics of the E-nose response curves. Finally, in order to shorten the time consumption of calculation and reduce the complexity of the identification model, linear discriminant analysis (LDA) is introduced to reduce the dimensionality of the extracted feature datasets, and an XGBoost-based gas identification model is trained using the LDA optimized feature datasets. The experimental results show that the proposed method can shorten the gas detection time, obtain sufficient gas features, and achieve nearly 100% identification accuracy for CH4, CO, and their mixed gases.

Gas identification/classification through pattern recognition techniques based on gas sensor arrays often requires the equilibrium responses or the full traces of time-series data of the sensor array. Leveraging upon the diverse gas sensing kinetics behaviors measured via the sensor array, a computational intelligence- based meta-model is proposed to automatically conduct the feature extraction and subsequent gas identification using time-series data during the transitional phase before reaching equilibrium. The time-series data contains implicit temporal dependency/correlation that is worth being characterized to enhance the gas identification performance and reliability. In this context, a tailored approach so-called convolutional long short-term memory (CLSTM) neural network is developed to perform the identification task incorporating temporal characteristics within time-series data. This novel approach shows the enhanced accuracy and robustness as compared to the baseline models, i.e., multilayer perceptron (MLP) and support vector machine (SVM) through the comprehensive statistical examination. Specifically, the classification accuracy of CLSTM reaches as high as 96%, regardless of the operating condition specified. More importantly, the excellent gas identification performance of CLSTM at early stages of gas exposure indicates its practical significance in future real-time applications. The promise of the proposed method has been clearly illustrated through both the internal and external validations in the systematic case investigation.

The visualization of the spatial distributions of gases from various sources is essential to understanding the composition, localization, and behavior of these gases. In this study, an inkjet-printed localized surface plasmon resonance (LSPR) subpixel gas sensor array was developed to visualize the spatial distributions of gases and to differentiate between acetic acid, geraniol, pentadecane, and cis-jasmone. The sensor array, which integrates gold nanoparticles (AuNPs), silver nanoparticles (AgNPs), and fluorescent pigments, was positioned 3 cm above the gas source. Hyperspectral imaging was used to capture the LSPR spectra across the sensor array, and these spectra were then used to construct gas information matrices. Principal component analysis (PCA) enabled effective classification of the gases and localization of their sources based on observed spectral differences. Heat maps that visualized the gas concentrations were generated using the mean squared error (MSE) between the sensor responses and reference spectra. The array identified and visualized the four gas sources successfully, thus demonstrating its potential for gas localization and detection applications. The study highlights a straightforward, cost-effective approach to gas sensing and visualization, and in future work, we intend to refine the sensor fabrication process and enhance the detection of complex gas mixtures.

As a typical machine olfactory system index, the accuracy of hybrid gas identification and concentration detection is low. This paper proposes a novel hybrid gas identification and concentration detection method. In this method, Kernel Principal Component Analysis (KPCA) is employed to extract the nonlinear mixed gas characteristics of different components, and then K-nearest neighbour algorithm (KNN) classification modelling is utilized to realize the recognition of the target gas. In addition, this method adopts a multivariable relevance vector machine (MVRVM) to regress the multi-input nonlinear signal to realize the detection of the concentration of the hybrid gas. The proposed method is validated by using CO and CH4 as the experimental system samples. The experimental results illustrate that the accuracy of the proposed method reaches 98.33%, which is 5.83% and 14.16% higher than that of principal component analysis (PCA) and independent component analysis (ICA), respectively. For the hybrid gas concentration detection method, the CO and CH4 concentration detection average relative errors are reduced to 5.58% and 5.38%, respectively.

Enhancing sensor sensitivity and gas identification capabilities is essential for the broad application of gas sensors. Developing efficient transducing methods for sensors can be applied to a wide range of sensors. However, developing such methods for resistive sensors remains challenging. In this study, an operating method that enhances both sensitivity and gas identification capability in resistive gas sensors is presented. The sensor operation is divided into two phases: the reaction phase and the signal detection phase, and propose optimized operating methods for each. In the reaction phase, the chemisorption of oxidizing and reducing gases are maximized through appropriate operating methods for each. In the signal detection phase, a read‐bias technique is introduced, enhancing sensitivity across all gases, with a 23‐fold increase for 500 ppb NO2 and a sixfold increase for 50 ppm H2S. Additionally, the limit of detection (LOD) can be improved, with the NO2 LOD reduced from 11.8 to 1.4 ppb. Furthermore, a method for obtaining gas‐specific signal patterns is presented that reflect the unique diffusion properties of each gas by simply adjusting the signal readout conditions. This approach demonstrates the accurate identification of four different gases using only a single sensor. This study presents a novel method to improve sensitivity and gas identification in resistive gas sensors by dividing operation into reaction and signal detection phases. The reaction phase optimizes chemisorption, while the signal detection phase, using a read‐bias technique, enhances sensitivity. Adjustable readout conditions enable distinct gas identification, allowing precise detection of multiple gases with a single sensor.

Sufficient gas accumulation is an essential factor that controls the effective development of tight sandstone gas reservoirs that are generally characterized by low porosity and permeability. Seismic methods are important for predicting potential gas areas in tight sandstones. However, the complex relationships between rock physical properties and gas saturation make gas enrichment estimation with seismic methods challenging. Nonetheless, seismic velocity dispersion using a wave-induced fluid flow mechanism can enable gas identification by utilizing the associated dispersion attributes. This paper proposes a method for improved gas identification using a new fluid dispersion attribute obtained by incorporating the decoupled fluid-solid seismic amplitude variation with offset representation into the frequency-dependent inversion scheme. Numerical analyses and synthetic data tests confirmed the enhanced sensitivity of the fluid dispersion attribute to gas saturation compared to the conventionally used compressional wave velocity dispersion attribute. Field data applications further validated the ability of the proposed fluid dispersion attribute to improve gas prediction in tight sandstone reservoirs. The results of the measurements enable rational interpretation of the geological significance of assessments of reservoir properties from gas-producing wellbores. The proposed fluid dispersion attribute is a reliable indicator for gas prediction and represents a useful tool for characterizing tight sandstone reservoirs.

With the development of sustainable economy, natural gas as a clean energy has received great attention, but there are still many problems to be solved in the exploration and development of natural gas. Among them, gas-water identification method of natural gas reservoir is a hot research topic at present. Correctly identifying reservoir properties and improving the ability of gas-water identification are the key points of expanding natural gas reserves and enhancing the exploration value of discovered gas reservoirs. With the deepening and development of the research on gas-water identification of natural gas reservoir in recent years, the experimental methods and technical means have been constantly enriched and improved, and more research results have been obtained, but they are still in the exploratory stage. Therefore, it is necessary to continuously strengthen the research on gas-water identification methods of natural gas reservoirs, make full use of current advanced technologies, and improve the coincidence rate of computer interpretation of gaswater layers.

A novel sparse representation classification method (SRC), namly SRC based on Method of Optimal Directions (SRC_MOD), is proposed for electronic nose system in this paper. By finding both a synthesis dictionary and a corresponding coefficient vector, the i-th class training samples are approximated as a linear combination of a few of the dictionary atoms. The optimal solutions of the synthesis dictionary and coefficient vector are found by MOD. Finally, testing samples are identified by evaluating which class causes the least reconstruction error. The proposed algorithm is evaluated on the analysis of hydrogen, methane, carbon monoxide, and benzene at self-adapted modulated operating temperature. Experimental results show that the proposed method is quite efficient and computationally inexpensive to obtain excellent identification for the target gases.