Search Results (11)

During robot-assisted rehabilitation, failure to recognize lower limb movement may efficiently limit the development of exoskeleton robots, especially for individuals with knee pathology. A major challenge encountered with surface electromyography (sEMG) signals generated by lower limb movements is variability between subjects, such as motion patterns and muscle structure. To this end, this paper proposes an sEMG-based lower limb motion recognition using an improved support vector machine (SVM). Firstly, non-negative matrix factorization (NMF) is leveraged to analyze muscle synergy for multi-channel sEMG signals. Secondly, the multi-nonlinear sEMG features are extracted, which reflect the complexity of muscle status change during various lower limb movements. The Fisher discriminant function method is utilized to perform feature selection and reduce feature dimension. Then, a hybrid genetic algorithm-particle swarm optimization (GA-PSO) method is leveraged to determine the best parameters for SVM. Finally, the experiments are carried out to distinguish 11 healthy and 11 knee pathological subjects by performing three different lower limb movements. Results demonstrate the effectiveness and feasibility of the proposed approach in three different lower limb movements with an average accuracy of 96.03% in healthy subjects and 93.65% in knee pathological subjects, respectively. Full article

The science of the provenance, sedimentary system, and distribution of sand bodies is unclear, limiting oil and gas exploration. Here, we combined heavy mineral, rare earth element, petrographic, and outcrop data to shed new light on the provenance, depositional environment, and paleoclimatic conditions of the Permian Jiamuhe Formation. The provenance is characterized by “a main provenance system, and four provenance zones,” and this result could be interpreted from analyses of its seismic reflection, clastic composition, REE_S, and heavy minerals. A detailed sedimentological study performed in the excellent outcrops, a comprehensive analysis of logging, and the legalistic cores of this formation allowed for the identification of ten lithofacies and three lithofacies associations. Four distributary/underwater channels were observed. Furthermore, the redox and paleoclimatic conditions based on trace elements (Th/U, V/(V + Ni), V/Cr, Cu/Zn, Sr/Ba, and Sr/Cu) suggested a weak reduction in the environment, as well as semi-humid and semi-dry conditions of the Jiamuhe Formation. These conditions are also supported by the mudstone color and plant fossils. The tectonic setting belongs to the acid island arc area based on the trace element discrimination diagram of La-Th-SC and the values of the La, Ce, ΣREE, L/H, La/Yb, and (La/Yb)N criteria. The research results further confirm that there are differences in the mineral compositions in the same provenance area, and they provide a geological basis for the fine sedimentary facies characterization and a favorable zone prediction in this area. Full article

Marine sediment is an important channel for methane leakage from the earth interior to the atmosphere. The investigation of gas invasion in fine-grained marine sediments is of great theoretical and practical significance in marine science and engineering. To study the mechanical mechanisms of fine-grained marine sediments subjected to shallow gas invasion, a gas injection test with a self-developed experimental apparatus was performed, and the gas invasion behavior was investigated. The results showed that the behavior of gas invasion in fine-grained sediments can be divided into different phases; the fracturing direction β gradually changes from vertical to horizontal, and finally fractures along the roof. Based on the 2D undrained elliptical cavity model and the tensile strength of sediments, considering both tensile and shear failure modes, a discrimination criteria of gas invasion was proposed. It revealed that gas invasion gradually changes from shear failure to tensile failure, and the fracturing angle θ predicted by the criteria is consistent with the experimental phenomenon. Full article

This study aimed to develop functional fruit wines using whole fruit, pulp, and peels from green bananas. The boiled banana homogenates were mixed with cane sugar before wine fermentation. Quality parameters, phenolic compounds, flavor components, and antioxidative properties of the green banana peel wine (GBPW), green banana pulp wine (GBMW), and whole banana wine (GBW) were determined. High-performance liquid chromatography was used to determine the phytochemical compounds in three wines, and the flavor components were further analyzed using headspace solid-phase microextraction combined with gas chromatography–mass spectrometry. The flavor components and in vitro antioxidant activities were, respectively, determined using the relative odor activity value and the orthogonal projections on latent structure discrimination analysis (OPLS-DA). In vitro antioxidative capacities for these wines were evaluated using antioxidant chemical assays and cell culture methods. The total phenolic and total tannin content of the GBPW, GBMW, and GBW showed reducing trends with increasing fermentation days, whereas the total flavonoid content of the wine samples exhibited downward trends. The antioxidant capacities of the three wine samples were higher than those of the raw fruit samples, except for the metal chelation rate (%). Additionally, the main flavor component in the wine samples was 3-methyl-1-butanol. Its percentages in the GBPW, GBMW, and GBW were 72.02%, 54.04%, and 76.49%, respectively. The OPLS-DA results indicated that the three wines presented significantly different antioxidant activities. The cell-culture-based antioxidant analysis showed that these wine samples had protective effects against the oxidative stress of the 3T3-L1 preadipocytes induced by hydrogen peroxide. This study provided a theoretical basis for defining the antioxidant characteristics of banana wines and expanding novel channels for using banana peels to develop nutraceuticals. Full article

In a typical ultra-deep high-temperature and high-pressure heterogeneous reservoir in Xinjiang, gas channeling quickly occurs during gas injection because of the heterogeneity of the reservoir, the low viscosity of gas injection, and the high gas-oil fluidity ratio. The identification and prediction methods of gas channeling in gas injection development were studied. First, gas channeling discrimination parameters were determined by the numerical simulation method. According to the ratio of gas to oil produced and the composition of oil and gas produced, the flow stages of formation fluid were divided into five regions: gas phase zone, two-phase zone, miscible zone, dissolved gas and oil zone, and original oil zone. The basis for gas channeling identification (namely, the field characterization parameters for gas channeling discrimination) was discovered through analysis and the knowledge of the operability of field monitoring data as the following two parameters: (1) the C1 content rising again on the previous platform when the trailing edge of the two-phase zone is produced and (2) the continuous rise of the gas-oil ratio in production. Then, considering the original high-pressure characteristics of the reservoir, the field characterization parameters of gas channeling under different formation pressures in the exploitation process (namely, C1 content and gas-oil ratio) were simulated and determined. Thus, a gas channeling discrimination method was established for gas injection development in ultra-deep high-temperature and high-pressure heterogeneous reservoirs. According to this gas injection approach, a gas channeling discrimination method was developed, and the field gas channeling judgment was carried out for a gas injection effective D1 well. The results of gas tracer detection were compared to verify the accuracy of this method, leading to strong support for this method in slowing down the gas channeling. Full article

Concept, design and practical implementation of a miniaturized spectrophotometer, utilized as a mid-infrared-based multi gas sensor is described. The sensor covers an infrared absorption wavelength range of 2.9 to 4.8 um, providing detection capabilities for carbon dioxide, carbon monoxide, nitrous oxide, sulphur dioxide, ammonia and methane. A lead selenide photo-detector array and customized MEMS-based micro-hotplate are used as the detector and broadband infrared source, respectively. The spectrophotometer optics are based on an injection moulded Schwarzschild configuration incorporating optical pass band filters for the spectral discrimination. This work explores the effects of using both fixed-line pass band and linear variable optical filters. We report the effectiveness of this low-power-consumption miniaturized spectrophotometer as a stand-alone single and multi-gas sensor, usage of a distinct reference channel during gas measurements, development of ideal optical filters and spectral control of the source and detector. Results also demonstrate the use of short-time pulsed inputs as an effective and efficient way of operating the sensor in a low-power-consumption mode. We describe performance of the spectrometer as a multi-gas sensor, optimizing individual component performances, power consumption, temperature sensitivity and gas properties using modelling and customized experimental procedures. Full article

Electroencephalogram (EEG) signals, which originate from neurons in the brain, have drawn considerable interests in identity authentication. In this paper, a face image-based rapid serial visual presentation (RSVP) paradigm for identity authentication is proposed. This paradigm combines two kinds of biometric trait, face and EEG, together to evoke more specific and stable traits for authentication. The event-related potential (ERP) components induced by self-face and non-self-face (including familiar and not familiar) are investigated, and significant differences are found among different situations. On the basis of this, an authentication method based on Hierarchical Discriminant Component Analysis (HDCA) and Genetic Algorithm (GA) is proposed to build subject-specific model with optimized fewer channels. The accuracy and stability over time are evaluated to demonstrate the effectiveness and robustness of our method. The averaged authentication accuracy of 94.26% within 6 s can be achieved by our proposed method. For a 30-day averaged time interval, our method can still reach the averaged accuracy of 88.88%. Experimental results show that our proposed framework for EEG-based identity authentication is effective, robust, and stable over time. Full article

Pressure vessels (PV) are designed to hold liquids, gases, or vapors at high pressures in various industries, but a ruptured pressure vessel can be incredibly dangerous if cracks are not detected in the early stage. This paper proposes a robust crack identification technique for pressure vessels using genetic algorithm (GA)-based feature selection and a deep neural network (DNN) in an acoustic emission (AE) examination. First, hybrid features are extracted from multiple AE sensors that represent diverse symptoms of pressure vessel faults. These features stem from various signal processing domains, such as the time domain, frequency domain, and time-frequency domain. Heterogenous features from various channels ensure a robust feature extraction process but are high-dimensional, so may contain irrelevant and redundant features. This can cause a degraded classification performance. Therefore, we use GA with a new objective function to select the most discriminant features that are highly effective for the DNN classifier when identifying crack types. The potency of the proposed method (GA + DNN) is demonstrated using AE data obtained from a self-designed pressure vessel. The experimental results indicate that the proposed method is highly effective at selecting discriminant features. These features are used as the input of the DNN classifier, achieving a 94.67% classification accuracy. Full article

Modelling and optimization of chemical reactors require a good understanding of the reactions mechanism with the corresponding kinetic description. Therefore, high quality kinetic data are needed, which can be challenging to obtain, especially for fast and highly exothermic reactions such as the CO₂ methanation. Traditionally, kinetic studies rely on measuring the exit gas composition (1 data point per species and experiment) using differential reactors with diluted catalyst beds and reactants to avoid temperature change. Therefore, an optically accessible catalytic channel reactor was designed, which allowed for the chance to gather spatially-resolved information on axial gas composition and catalyst surface temperature, specifically by means of a movable sampling capillary and shortwave infrared-thermography (SWIR), respectively. A catalyst coated plate was placed at the bottom of the channel, while a set of two quartz glass plates covers the top. In the current study 35 data points per gas species were collect for 1 experiment conducted under laminar flow conditions at 425 °C. Catalyst surface temperature determined via a SWIR camera was not influenced by polyatomic molecules partaking in the reaction and thus did not falsify the kinetic data. The catalyst mass distribution along the reactor axis was determined, enabling the development of a correct reactor model for kinetic parameter estimation and model discrimination. Full article

Despite tremendous potential and urgent demand in high-response low-cost gas identification, the development of gas identification based on a metal oxide semiconductor nanowire/nanobelt remains limited by fabrication complexity and redundant signals. Researchers have shown a multisensor-array strategy with “one key to one lock” configuration. Here, we describe a new strategy to create high-response room-temperature gas identification by employing gas as dielectric. This enables gas discrimination down to the part per billion (ppb) level only based on one pristine single nanobelt transistor, with the excellent average Mahalanobis distance (MD) as high as 35 at the linear discriminant analysis (LDA) space. The single device realizes the selective recognition function of electronic nose. The effect of the gas dielectric on the response of the multiple field-effect parameters is discussed by the comparative investigation of gas and solid-dielectric devices and the studies on trap density changes in the conductive channel. The current work opens up exciting opportunities for room-temperature gas recognition based on the pristine single device. Full article

Based on a laser diode, a 12 × 6 photodiode array sensor, and machine learning techniques, a new void fraction measurement method for gas-liquid two-phase flow in small channels is proposed. To overcome the influence of flow pattern on the void fraction measurement, the flow pattern of the two-phase flow is firstly identified by Fisher Discriminant Analysis (FDA). Then, according to the identification result, a relevant void fraction measurement model which is developed by Support Vector Machine (SVM) is selected to implement the void fraction measurement. A void fraction measurement system for the two-phase flow is developed and experiments are carried out in four different small channels. Four typical flow patterns (including bubble flow, slug flow, stratified flow and annular flow) are investigated. The experimental results show that the development of the measurement system is successful. The proposed void fraction measurement method is effective and the void fraction measurement accuracy is satisfactory. Compared with the conventional laser measurement systems using standard laser sources, the developed measurement system has the advantages of low cost and simple structure. Compared with the conventional void fraction measurement methods, the proposed method overcomes the influence of flow pattern on the void fraction measurement. This work also provides a good example of using low-cost laser diode as a competent replacement of the expensive standard laser source and hence implementing the parameter measurement of gas-liquid two-phase flow. The research results can be a useful reference for other researchers’ works. Full article