RESUMO
Due to the complexity of electromechanical equipment and the difficulties in obtaining large-scale health monitoring datasets, as well as the long-tailed distribution of data, existing methods ignore certain characteristics of health monitoring data. In order to solve these problems, this paper proposes a method for the fault diagnosis of rolling bearings in electromechanical equipment based on an improved prototypical network-the weight prototypical networks (WPorNet). The main contributions of this paper are as follows: (1) the prototypical networks, which perform well on small-sample classification tasks, were improved by calculating the different levels of influence of support sample distributions in order to achieve the prototypical calculation. The change in sample influence was calculated using the Kullback-Leibler divergence of the sample distribution. The influence change in a specific sample can be measured by assessing how much the distribution changes in the absence of that sample; and (2) The Gramian Angular Field (GAF) algorithm was used to transform one-dimensional time series into two-dimensional vibration images, which greatly improved the application effect of the 2D convolutional neural network (CNN). Through experiments on MAFAULDA and CWRU bearing datasets, it was shown that this network effectively solves the shortcomings of a small number of valid samples and a long-tail distribution in health monitoring data, it enhances the dependency between the samples and the global data, it improves the model's feature extraction ability, and it enhances the accuracy of model classification. Compared with the prototypical network, the improved network model increased the performance of the 2-way 10-shot, 2-way 20-shot, and 2-way 50-shot classification tasks by 5.23%, 5.74%, and 4.37%, respectively, and it increased the performance of the 4-way 10-shot, 4-way 20-shot, and 4-way 50-shot classification tasks by 12.02%, 10.47%, and 4.66%, respectively. Experimental results show that the improved prototypical network model has higher sample classification accuracy and stronger anti-interference ability compared with traditional small-sample classification models.
RESUMO
Airborne infrared optical systems equipped with multiple cooled infrared cameras are commonly utilized for quantitative radiometry and thermometry measurements. Radiometric calibration is crucial for ensuring the accuracy and quantitative application of remote sensing camera data. Conventional radiometric calibration methods that consider internal stray radiation are usually based on the assumption that the entire system is in thermal equilibrium. However, this assumption leads to significant errors when applying the radiometric calibration results in actual mission scenarios. To address this issue, we analyzed the changes in optical temperature within the system and developed a simplified model to account for the internal stray radiation in the non-thermal equilibrium state. Building upon this model, we proposed an enhanced radiometric calibration method, which was applied to the absolute radiometric calibration procedure of the system. The radiometric calibration experiment, conducted on the medium-wave channel of the system within a temperature test chamber, demonstrated that the proposed method can achieve a calibration accuracy exceeding 3.78% within an ambient temperature range of -30 °C to 15 °C. Additionally, the maximum temperature measurement error was found to be less than ±1.01 °C.
RESUMO
The ball joint is a multi-degree-of-freedom transmission pair, if it can replace the inner frame in the aviation photoelectric pod to carry the optical load, which will greatly simplify the system structure of the photoelectric pod and reduce the space occupied by the inner frame. However, installation errors in ball joint siting introduce nonlinear errors that are difficult to correct and two degree of freedom angular displacement of the ball joint is difficult to detect, which limits application in the precision control of two degrees of freedom systems. Studies of spherical capacitive sensors to date have not tested sensors for use in an inner frame stabilisation mechanism nor have they analysed the influence of installation error on sensor output. A two-axis angular experimental device was designed to measure the performance of a ball joint capacitive sensor in a frame stabilisation mechanism in an aeronautical optoelectronic pod, and a mathematical model to compensate for ball joint capacitive sensor installation error was created and tested. The experimental results show that the resolution of the capacitive sensor was 0.02° in the operating range ±4°, the repeatability factor was 0.86%, and the pulse response time was 39 µs. The designed capacitive sensor has a simple structure, high measurement accuracy, and strong robustness, and it can be integrated into ball joint applications in the frames of aeronautical photoelectric pods.
RESUMO
The spherical pair has an important role in the inner frame of the stabilization mechanism of the aviation optoelectronic pod. However, its two-degrees-of-freedom (2-DOF) angular displacement signal is difficult to detect, seriously restricting its application in aviation optoelectronic pods. Therefore, this study proposes a new method to measure a spherical pair's 2-DOF angular displacement using a spherical capacitive sensor. The capacitive sensor presented by this method realizes the measurement of the 2-DOF angular displacement of the spherical pair by integrating the spherical electrode groups in the ball head and the ball socket of the spherical pair. First, based on the geometric structure of the spherical pair, the structure of the capacitive sensor is designed, and the mathematical model for the capacitive sensor is deduced. Then, the sensor's output capacitance, in different directions, is simulated by Ansoft Maxwell software. Finally, an experiment device is built for the measurement experiments. The simulation analysis and experimental results show that the spherical capacitive sensor has an approximately linear output in different directions, and the measured output capacitance is as high as 89.7% of the theoretical value. Compared with the existing sensors that measure the 2-DOF angular displacement signal of the ball pair, the sensor proposed in this study has an integrated structure, which can be integrated into the spherical pair. That makes it possible to apply the spherical pair to the inner frame of the aviation optoelectronic pod.
RESUMO
Green deep eutectic solvents (DESs) are widely used to extract bioactive components from plant biomass; however, hydrophilic DES and bioactive component isolation methods have not been developed. In this study, we synthesized hydrophilic DES (CL-CA-DES) using citric acid and choline chloride. We combined this with environmentally friendly anion- and cation-exchange resin column chromatographic isolation methods. This approach extracted and isolated four polyphenolic compounds (catechins, epicatechins, procyanidin B1, and procyanidin B2) from raspberry root and efficiently recovered the hydrophilic DES. CL-CA-DES extracted significantly higher contents of catechin and procyanidin B2 from raspberry root compared to other solvents. It also extracted substantially higher contents of epicatechin compared to butyl alcohol, 70% ethanol, and water, but there was no significant difference when compared with acetone and ethyl acetate. Additionally, CL-CA-DES extracted significantly higher contents of procyanidin B1 compared to butyl alcohol, water, and ethyl acetate, with no significant difference when compared with 70% ethanol and acetone. The isolation efficiency of the bioactive components in the raspberry root extract by anion- and cation-exchange resin column chromatography was higher than that of the organic solvent extraction and precipitation generation methods, and the method was effective in recovering CL-CA-DES with a recovery rate higher than 60%. In conclusion, this study developed a new method for the efficient recovery of hydrophilic CL-CA-DES, which can be used for isolating polyphenolic compounds from raspberry root.
RESUMO
Metabolic flux analysis (MFA) using stable isotope labeled tracers is a powerful tool to estimate fluxes through metabolic pathways. It finds applications in studying metabolic changes in diseases, regulation of cellular energetics, and novel strategies for metabolic engineering. Accurate and precise quantification of the concentration of metabolites and their labeling states is critical for correct MFA results. Utilizing an ultra-high performance liquid chromatography-high resolution mass spectrometry (UHPLC-HRMS) system, an analytical method for simultaneously quantifying the concentration of sugar metabolites and their mass isotopologue distribution (MID) was developed. The method performs with good linearity and coefficient of determination (R2) > 0.99, while the detection limit ranged from 0.1 to 50 mg L-1. Seven sugar metabolites were detected in a labeled Brevibacterium flavum sample using the method. The detected quantities ranged from 6.15 to 3704.21 mg L-1, and 13C abundance was between 12.77% and 66.67% in the fermentation fluid and 16.28% and 91.93% in the bacterial body. Overall, the method is efficient, accurate, and suitable for analysis of labeled sugar metabolites in 13C MFA studies.