分类： 机械工程 >> 机械设计 提交时间： 2024-04-01
摘要： Overstaffing production in underground coal mining is not convenient for daily management, and incomplete information of coal miners hinders the rescue process of firefighters during mine accidents. To address this safety sustainability issue, a novel face recognition method based on an improved multiscale neural network is proposed in this paper. A new depthwise seperable (DS)-inception block is designed and a joint supervised loss function based on center loss theory is developed to constructe a new multiscale model. The miniers can be recognized in the harsh underground environment during the life rescue. Experimental results show that the accuracy, recall and F1-score indexes of the proposed method for the miner face recognition in the underground mining environment are 97.26%, 94.17% and 95.42%, respectively. Transfer model with joint supervised loss can effectively improve the recognition accuracy by about 0.5~1.5%. In addition, the average recognition accuracy of the proposed face recognition method achieves to 91.34% and the miss detection rate is less than 5% in the dugout tunnel of coal mine.
分类： 机械工程 >> 机械制造工艺与设备 提交时间： 2024-03-28
摘要： The heat transfer mechanism andtemperature distributioninlaser weldingapplications have a great impact on the quality of the weld bead geometry, mechanical properties and the resultant microstructure characterizations of the welding process. In this study, the effects of pulsedlaser weldingparameters including the frequency and pulse width on the melt velocity field andtemperature distributionin dissimilarlaser weldingof stainless steel 420 (S.S 420) and stainless steel 304 (S.S 304) was investigated. A comprehensive comparison was conducted through the numerical simulation and artificial neural network (ANN). The results of numerical simulation indicated thatbuoyancy forceandMarangonistress are the most important factors in the formation of the flow of liquid metal. Also, increasing the pulse width from 8 to 12ms due to increasing the pulse energy, the temperature in the center of the melt pool increased about 250°C. This leads to increasing the convective heat transfer in the molten pool and heat affected zone (HAZ). The temperature difference at a distance of 1mm from the beam center at both metals at a frequency of 15 and 20Hz is bout 58 and 75°C, respectively. Furthermore, reducing the frequency to 5Hz, due to diminishment of thermal energy absorption time, has clearly decreased the weld penetration depth in the workpiece. According to the ANN results, increasing both pulse duration and frequency has the significant effect on increasing melting ratio from 0.4 to 0.8 compared to the other input parameters. The ANN results confirmed that under the same input conditions, because of the differences in thermal conductivity coefficient, absorption coefficient and melting point of the two pieces, S.S 304 has experienced higher temperatures about 10% more than S.S 420. Also, among the 13 back propagation learning algorithms, the Bayesian regularization algorithm had the best performance. Among the number of different neurons in the hidden layer, comparison was performed to prevent network overfitting. The maximum relative error of network output data and target data for S.S 304 and S.S 420 temperatures and melting ratio were 7.297, 10.16 and 11.33%, respectively.
分类： 地球科学 >> 地理学 提交时间： 2019-03-28 合作期刊: 《干旱区科学》
摘要： Soil water is an important composition of water recycle in the soil-plant-atmosphere continuum. However, intense water exchange between soil-plant and soil-atmosphere interfaces only occurs in a certain layer of the soil profile. For deep insight into water active layer (WAL, defined as the soil layer with a coefficient of variation in soil water content >10% in a given time domain) in the Loess Plateau of China, we measured soil water content (SWC) in the 0.0–5.0 m soil profile from 86 sampling sites along an approximately 860-km long south-north transect during the period 2013–2016. Moreover, a dataset contained four climatic factors (mean annual precipitation, mean annual evaporation, annual mean temperature and mean annual dryness index) and five local factors (altitude, slope gradient, land use, clay content and soil organic carbon) of each sampling site was obtained. In this study, three WAL indices (WAL-T (the thickness of WAL), WAL-CV (the mean coefficient of variation in SWC within WAL) and WAL-SWC (the mean SWC within WAL)) were used to evaluate the characteristics of WAL. The results showed that with increasing latitude, WAL-T and WAL-CV increased firstly and then decreased. WAL-SWC showed an opposite distribution pattern along the south-north transect compared with WAL-T and WAL-CV. Average WAL-T of the transect was 2.0 m, suggesting intense soil water exchange in the 0.0–2.0 m soil layer in the study area. Soil water exchange was deeper and more intense in the middle region than in the southern and northern regions, with the values of WAL-CV and WAL-T being 27.3% and 4.3 m in the middle region, respectively. Both climatic (10.1%) and local (4.9%) factors influenced the indices of WAL, with climatic factors having a more dominant effect. Compared with multiple linear regressions, pedotransfer functions (PTFs) from artificial neural network can better estimate the WAL indices. PTFs developed by artificial neural network respectively explained 86%, 81% and 64% of the total variations in WAL-T, WAL-SWC and WAL-CV. Knowledge of WAL is crucial for understanding the regional water budget and evaluating the stable soil water reserve, regional water characteristics and eco-hydrological processes in the Loess Plateau of China.
分类： 物理学 >> 核物理学 分类： 核科学技术 >> 核科学与技术 分类： 核科学技术 >> 核探测技术与核电子学 提交时间： 2024-01-08
摘要： To detect radioactive substances with low activity levels, an anticoincidence detector and a high-purity germanium (HPGe) detector are typically used simultaneously to suppress Compton scattering background, thereby resulting in an extremely low detection limit and improving the measurement accuracy. However, the complex and expensive hardware required does not facilitate the application or promotion of this method. Thus, a method is proposed in this study to discriminate the digital waveform of pulse signals output using an HPGe detector, whereby Compton scattering background is suppressed and a low minimum detectable activity (MDA) is achieved without using an expensive and complex anticoincidence detector and device. The electric-field-strength and energy-deposition distributions of the detector are simulated to determine the relationship between pulse shape and energy-deposition location, as well as the characteristics of energy-deposition distributions for full- and partial-energy deposition events. This relationship is used to develop a pulse-shape-discrimination algorithm based on an artificial neural network for pulse-feature identification. To accurately determine the relationship between the deposited energy of gamma rays in the detector and the deposition location, we extract four shape parameters from the pulse signals output by the detector. Machine learning is used to input the four shape parameters into the detector. Subsequently, the pulse signals are identified and classified to discriminate between partial- and full-energy deposition events. Some partial-energy deposition events are removed to suppress Compton scattering. The proposed method effectively decreases the MDA of an HPGe -energy dispersive spectrometer. Test results show that the Compton suppression factors for energy spectra obtained from measurements on 152Eu, 137Cs, and 60Co radioactive sources are 1.13 (344 keV), 1.11 (662 keV), and 1.08 (1332 keV), respectively, and that the corresponding MDAs are 1.4%, 5.3%, and 21.6% lower, respectively
分类： 物理学 >> 核物理学 提交时间： 2023-12-07
摘要： To detect radioactive substances with a low activity level, an anti-coincidence detector and high-purity germanium detector (HPGe) are often used in combination to suppress the Compton scattering background, thereby obtaining an extremely low detection limit and improving the measurement accuracy. However, the complex and expensive hardware system required does not facilitate application and promotion of this method. Thus, a method is proposed to discriminate the digital waveform of pulse signals output by a HPGe detector, whereby the Compton scattering background is suppressed and a low minimum detectable activity (MDA) is obtained without using an expensive and complex anti-coincidence detector and device. The electric field strength distribution and the energy deposition distribution in the detector are simulated to determine the relationship between the pulse shape and location of energy deposition, as well as the characteristics of the energy deposition distribution for full- and partial-energy deposition events. This relationship is used to develop a pulse shape discrimination (PSD) algorithm based on employing an artificial neural network (ANN) for pulse feature identification. To accurately determine the relationship between the deposited energy of gamma (g)-rays in the detector and deposition location, we extract four shape parameters from the pulse signals output by the detector. Machine learning is used to input the four shape parameters to the detector. Then, the pulse signals are identified and classified to discriminate between partial- and full-energy deposition events, and some partial-energy deposition events are removed to suppress Compton scattering. The proposed method effectively lowers the MDA of a HPGe -energy dispersive spectrometer. Test results show that the Compton suppression factors for energy spectra obtained from measurements on 152Eu, 137Cs and 60Co radioactive sources are 1.13 (344 keV), 1.11 (662 keV) and 1.08 (1332 keV), respectively, and the corresponding MDAs are lowered by 1.4%, 5.3% and 21.6%.
分类： 地球科学 >> 地理学 提交时间： 2021-10-11 合作期刊: 《干旱区科学》
摘要： Zarrineh River is located in the northwest of Iran, providing more than 40% of the total inflow into the Lake Urmia that is one of the largest saltwater lakes on the earth. Lake Urmia is a highly endangered ecosystem on the brink of desiccation. This paper studied the impacts of climate change on the streamflow of Zarrineh River. The streamflow was simulated and projected for the period 1992–2050 through seven CMIP5 (coupled model intercomparison project phase 5) data series (namely, BCC-CSM1-1, BNU-ESM, CSIRO-Mk3-6-0, GFDL-ESM2G, IPSL-CM5A-LR, MIROC-ESM and MIROC-ESM-CHEM) under RCP2.6 (RCP, representative concentration pathways) and RCP8.5. The model data series were statistically downscaled and bias corrected using an artificial neural network (ANN) technique and a Gamma based quantile mapping bias correction method. The best model (CSIRO-Mk3-6-0) was chosen by the TOPSIS (technique for order of preference by similarity to ideal solution) method from seven CMIP5 models based on statistical indices. For simulation of streamflow, a rainfall-runoff model, the hydrologiska byrans vattenavdelning (HBV-Light) model, was utilized. Results on hydro-climatological changes in Zarrineh River basin showed that the mean daily precipitation is expected to decrease from 0.94 and 0.96 mm in 2015 to 0.65 and 0.68 mm in 2050 under RCP2.6 and RCP8.5, respectively. In the case of temperature, the numbers change from 12.33°C and 12.37°C in 2015 to 14.28°C and 14.32°C in 2050. Corresponding to these climate scenarios, this study projected a decrease of the annual streamflow of Zarrineh River by half from 2015 to 2050 as the results of climatic changes will lead to a decrease in the annual streamflow of Zarrineh River from 59.49 m3/s in 2015 to 22.61 and 23.19 m3/s in 2050. The finding is of important meaning for water resources planning purposes, management programs and strategies of the Lake's endangered ecosystem.
点击量
通过
点击量
下载量
评论
