Real-time data-driven fault diagnosis of proton exchange membrane fuel cell system based on binary encoding convolutional neural network

The performance of proton exchange Membrane fuel cell (PEMFC) fault diagnosis system plays an important role in normal operation of PEMFC. Therefore, a new fault diagnosis algorithm based on binary matrix encoding neural network called BinE-CNN is proposed. In BinE-CNN, high-dimensional features are extracted through binary encoding, and the feature maps are transferred to a convolutional neural network (CNN) to realize seven-category fault classification. For development of BinE-CNN, a PEMFC model is modeled to generate simulative datasets. Simulative test precision and Frames per second (FPS) of BinE-CNN have reached respectively 0.973 and 999.8 (better than support vector machines (SVM), long short-term memory neural network (LSTM), etc.). In experimental verification section, fault datasets are collected during bench test. After that, BinE-CNN is deployed on vehicle control unit (VCU) to verify its engineering value (real-time and precision). The result meet both requirements, with time cost of 96.15 ms and precision of 0.931.     

FHUDS-2催化剂在天津石化的工业应用

中国石油化工股份有限公司天津分公司为生产硫含量符合欧Ⅲ排放标准S≯350（μg／g）柴油及满足未来更为严格的欧Ⅳ排放标准S≯50（μg/g）柴油,在320万t/a柴油加氢装置上选用了中国石油化工股份有限公司抚顺石油化工研究院（FRIPP）开发的柴油超深度加氢脱硫催化剂FHUDS-2。工业应用结果表明,FHUDS-2催化剂对各种油品的适应性强,反应温度低,具有优异的加氢脱硫性能,是生产清洁柴油的理想催化剂。     

Predicting reservoir property trends under heat exploitation: interaction between flow, heat transfer, transport, and chemical reactions in a deep aquifer at Stralsund, Germany

The subsurface flow and hydrogeothermal simulation system SHEMAT (Bartels, J., Kuhn, M., Pape, H., Clauser, C., 2000. A new aquifer simulation tool for coupled flow, heat transfer, multi-species transport and chemical water-rock interactions. In: Proceedings World Geothermal Congress 2000, Kyushu – Tohuku, Japan, May 28 – June 10, pp. 3997–4002) is used to investigate a typical hydrothermal sandstone reservoir situated in the North German Basin. This study focuses on the prediction of long-term behavior of reservoir properties for the entire operation time with reinjection during heat exploitation for district heating. The Stralsund location in NE Germany and the Detfurth sandstone horizon (Buntsandstein) are chosen due to the combination of its already confirmed geothermal potential and the availability of a complete data set. An installation of two production wells and one well for reinjection implements heat exploitation. Reinjection is required due to high salinity of the water. In order to quantify injectivity changes and allow the separation of thermal from chemical effects, changes in the hydraulic parameters of the reservoir are at first studied without chemical reactions. Reinjection of cooled water of higher viscosity than the natural reservoir fluid leads to a continuous reduction of the injectivity. This effect is partially balanced by thermally induced mineral reactions. Dissolution of anhydrite in the vicinity of the injection well dominates the effect of anhydrite precipitation at the propagating thermal front leading to a net increase of injectivity. Observed calcite precipitation around the injection well and dissolution at the thermal front are too small to alter reservoir properties significantly. Coupled numerical simulation indicates that the injectivity of the reservoir is influenced primarily by the viscosity effect, but that mineral reactions weaken this negative trend. Operation of a geothermal heating plant at the Stralsund location would not be restricted by a long-term reduction in the injectivity of the reinjection well.     

综合回收大厂七号坝老尾矿铅锑锌的浮选试验研究

针对大厂七号坝老尾矿铅锑锌品位低、嵌布粒度细、矿物组份复杂且深度氧化的特点,采用硫化钠作活化剂,丁基黄药与PC-1作组合捕收剂,新型有机抑制剂GZT抑制脉石及碳质矿物,成功地从大厂七号坝老尾矿中回收了铅锑锌矿物,铅锑锌精矿品位分别为21.36%、14.75%、45.63%,回收率分别达到70.12%、67.63%和83.29%。     

Moisture loss and oil uptake during deep fat frying of “Kroštula” dough

The mechanism of moisture transfer by diffusion and uptake of frying oil was studied during deep fat frying of Krotula dough with an initial moisture content of 0.4358 kg/kg (db). The experimental data were found to fit well to a first-order exponential model for moisture transfer (with a regression coefficient of 0.99). The effective moisture diffusivity of Krotula dough was determined for two periods of moisture loss (intensive in first 60 s and constant after 60 s of frying). In the frying range 0–60 s at temperatures of 160, 170, 180 and 190±1 °C, the effective moisture diffusivity values were 5.837, 6.607, 8.472 and 9.728×10^–9 m²/s, from which the activation energy (30 kJ/mol) was calculated by using an Arrhenius-type equation. The effective oil diffusivity values of Krotula dough in the frying time range 0–210 s at temperatures of 160, 170, 180 and 190±1 °C were 0.932, 1.135, 1.094 and 1.054×10^–9 m²/s, from which the activation energy (5.5 kJ/mol) was calculated by using an Arrhenius-type equation.     

有限信息下基于深度学习模型的小型分布式光伏电站功率预测

目的是解决小型分布式光伏电站在无气象站配备、无法测量气象变量(即太阳辐照度、温度、相对湿度等)的情况下,通过区域内光伏电站历史出力数据预测光伏发电的问题。基于有限信息,提出了两层的LSTM深度学习模型,对小型分布式光伏电站功率进行了预测,并对其超参数对其预测效果的影响进行了分析。此外,利用澳大利亚爱丽丝泉地区的分布式光伏电站数据来验证该模型的准确性,并与使用气象数据进行预测模型的效果进行了对比。结果表明,借助区域内光伏电站历史功率数据进行预测的效果良好,适用于无气象站情景下的光伏功率预测。     

An assessment of the antioxidant potential of coriander extracts in ghee when stored at high temperature and during deep fat frying

A coriander extract was evaluated for its potential to increase oxidative stability of ghee. Antioxidant activity using β‐carotene–linoleic acid model system and radical scavenging activity by 2‐diphenyl‐1‐picrylhydrazyl hydrate free (DPPH) was higher in steam distilled extract compared to its oleoresin counterpart. Steam distilled extract and oleoresin were significantly effective in retarding the deterioration of ghee relative to control, as observed in peroxide value, conjugated dienes, thiobarbituric acid (TBA) value and oxidative stability index at the end of 21 days of storage. However, these were less effective than butylated hydroxyanisole (BHA). During deep fat frying, steam distilled extracts showed higher antioxidant activity compared to oleoresin and BHA.     

崩落法深部开采岩移及地表塌陷规律分析研究

大红山铁矿为国内规模最大的地下开采矿山之一,其主采区无底柱分段崩落法结构参数为国内领先,大规模深部开采导致各采区之间应力较为复杂。针对大红山铁矿硐室爆破冒落上覆岩层后,崩落法深部开采引发的岩层移动及地表塌陷规律展开研究,总结地表塌陷区沉降及其裂缝的发展过程,系统分析了深部开采对地表沉降及开裂的影响,得出了岩移导致的陷落角。通过地表沉降及相关井下地压监测数据分析,得到了硐室爆破后,塌陷区地表及井下岩层趋于稳定的结论,表明崩落法上覆岩层冒落对于缓解深部开采高应力、保证足够覆盖岩层厚度、控制岩层移动具有重要作用。     

滨海矿山深部开采水力调查及稳定性分析

占地球表面71%的海洋蕴藏着丰富的矿产资源,随着陆地矿产资源日益减少,海底资源开采成为未来的发展方向,其中海底基岩矿产资源深井开采普遍面临富水和高地应力的问题,水和应力的分布及其耦合作用是影响海底深井开采安全性的重要因素。以滨海矿山三山岛金矿深井开采为背景开展水力研究,调查矿区深部含水带分布,开展现场地应力测量,统计深部岩体节理裂隙分布规律,研究深部岩体的渗透性,建立水力耦合数值模型,对深部矿体进行模拟开采。通过上述研究工作,获得了矿区地应力分布规律,深部岩体渗特性以及资源开采过程中围岩位移、应力的时空演化规律,并指出了潜在危险源的时空分布。     

冬瓜山铜矿微震监测系统及其应用研究

冬瓜山铜矿是千米深井开采有岩爆倾向的金属矿山,建立了微震监测系统以实现岩层活动的连续监测和岩爆的实时预报。综合实际工程条件和拟定传感器空间网络方案的地震事件定位精度和系统灵敏度计算结果,实现了微震监测网络的优化。建立的数字化多通道微震监测系统具有先进和丰富的地震学定量分析和可视化解释功能,可以完成多重聚类和时间序列分析。在冬瓜山铜矿已记录的监测数据的基础上,开展了矿山地震信号识别、矿山地震活动分析方法和冬瓜山地震活动规律等研究,取得到了初步成果。