葵花籽粕综合利用研究进展

葵花籽粕是葵花籽提油后的副产品。葵花籽粕蛋白品质优,各氨基酸组成平衡,此外葵花籽粕中还含有绿原酸、黄酮等多种活性物质,营养价值高,具有广阔的应用前景。从葵花籽粕作为原料提取有效成分及作为饲料用原料两方面对葵花籽粕资源的综合利用情况进行了阐述,以期为葵花籽粕的进一步开发利用提供参考。     

基于主动学习的离群点集成挖掘方法研究

离群点检测任务通常缺少可用的标注数据，且离群数据只占整个数据集的很小一部分，相较于其他的数据挖掘任务，离群点检测的难度较大，尚没有单一的算法适合于所有的场景。因此，结合多样性模型集成和主动学习思想，提出了一种基于主动学习的离群点集成检测方法OMAL（Outlier Mining based on Active Learning）。在主动学习框架指导下，根据各种基学习器的对比分析，选择了基于统计的、基于相似性的、基于子空间划分的三个无监督模型作为基学习器。将各基学习器评判的处于离群和正常边界的数据整合后呈现给人类专家进行标注，以最大化人类专家反馈的信息量；从标注的数据集和各基学习器投票产生的数据集中抽样，基于GBM（Gradient BoostingMachine）训练一个有监督二元分类模型，并将该模型应用于全数据集，得出最终的挖掘结果。实验表明，提出方法的AUC有了较为明显的提升，且具有良好的运行效率，具备较好的实用价值。     

综采工作面活性磁化水防尘技术研究

煤矿生产是环境较为封闭的井下作业，具有一定的特殊性，作业空间湿度高、生产环境复杂，其中粉尘是危害最大的因素之一，会对机器和人体造成不同程度的损害。粉尘的危害主要由两方面决定:煤尘能够燃烧，在浓度较高时还会爆炸；粉尘会危害人的身体健康。以平煤股份十矿-320 m水平己四采区东翼下段的己_15-16-24130工作面为研究对象，研究磁化水除尘的方法及原理，通过对比分析优选出活性磁化水添加剂，在该工作面试验后具有良好的除尘效果，大大增加了可见度、降低了空气中悬浮的煤尘。研究对该矿井其他工作面除尘工作具有指导意义。     

风洞模型加速度负反馈振动主动控制方法研究

针对尾撑式风洞模型强方向性振动问题,基于异位配置加速度负反馈控制器对风洞模型振动主动控制方法进行研究。首先,通过模态分析得到系统低阶模态振动的强方向性并基于系统特性设计了具有结构耦合性的内嵌压电陶瓷作动器的风洞模型振动主动控制系统。然后,基于异位配置NAF控制器分别设计了针对第二阶模态和前两阶模态的单模态NAF控制算法和双模态NAF控制算法。最后,进行了实验验证,结果表明:双模态NAF控制算法控制效果较好,前两阶模态阻尼比分别提高近13倍和近40倍,镇定时间分别缩短近11s和近26s。     

冲击地压高应力回采巷道支护改革技术的研究与应用

通过改变传统巷道支护理念,优化巷道支护方式和工艺,解决冲击区域高应力与强卸压大蠕变巷道的主动支护技术难题,规范巷道支护材料、施工机具和巷道支护效果监测,降低巷道翻修率,保障矿井采掘接替正常,为巷道支护提供参考和理论依据,同时进一步完善冲击区域煤层开采巷道支护技术,也为煤炭资源的开采最大化和经济效益最大化提供科学依据和技术保障。     

Stationary and Progressive Phenotypes Caused by the p.G90D Mutation in Rhodopsin Gene

Mutations in rhodopsin gene (RHO) are a frequent cause of retinitis pigmentosa (RP) and less often, congenital stationary night blindness (CSNB). Mutation p.G90D has previously been associated with CSNB based on the examination of one family. This study screened 60 patients. Out of these 60 patients, 32 were affected and a full characterization was conducted in 15 patients. We described the clinical characteristics of these 15 patients (12 male, median age 42 years, range 8–71) from three families including visual field (Campus Goldmann), fundus autofluorescence (FAF), optical coherence tomography (OCT) and electrophysiology. Phenotypes were classified into four categories: CSNB (N = 3, 20%) sector RP (N = 3, 20%), pericentral RP (N = 1, 6.7%) and classic RP (N = 8, 53.3% (8/15)). The phenotypes were not associated with family, sex or age (Kruskal–Wallis, p > 0.05), however, cystoid macular edema (CME) was observed only in one family. Among the subjects reporting nyctalopia, 69% (22/32) were male. The clinical characteristics of the largest p.G90D cohort so far showed a large frequency of progressive retinal degeneration with 53.3% developing RP, contrary to the previous report.     

Detailed Balance Broken by Catch Bond Kinetics Enables Mechanical-Adaptation in Active Materials

Unlike nearly all engineered materials which contain bonds that weaken under load, biological materials contain “catch” bonds which are reinforced under load. Consequently, materials, such as the cell cytoskeleton, can adapt their mechanical properties in response to their state of internal, non-equilibrium (active) stress. However, how large-scale material properties vary with the distance from equilibrium is unknown, as are the relative roles of active stress and binding kinetics in establishing this distance. Through course-grained molecular dynamics simulations, the effect of breaking of detailed balance by catch bonds on the accumulation and dissipation of energy within a model of the actomyosin cytoskeleton is explored. It is found that the extent to which detailed balance is broken uniquely determines a large-scale fluid-solid transition with characteristic time-reversal symmetries. The transition depends critically on the strength of the catch bond, suggesting that active stress is necessary but insufficient to mount an adaptive mechanical response.     

Re-Looking into the Active Moieties of Metal X-ides (X- = Phosph-, Sulf-, Nitr-, and Carb-) Toward Oxygen Evolution Reaction

Non-precious metal-based catalysts for oxygen evolution reaction (OER) have been extensively studied, among which the transition metal X-ides (including phosph-ides, sulf-ides, nitr-ides, and carb-ides) materials are emerging as promising candidates to replace the benchmark Ir/Ru-based materials in alkaline media. However, it is controversial whether the metal Xides host the real active sites since these metal Xides are thermodynamically unstable under a harsh OER environment—it has been reported that the initial metal Xides can be electrochemically oxidized and transformed into corresponding oxides and (oxy)hydroxides. Therefore, the metal Xides are argued as “pre-catalysts”; the electrochemically formed oxides and (oxy)hydroxides are believed as the real active moieties for OER. Herein, the recent advances in understanding the transformation behavior of metal Xides during OER are re-looked; importantly, hypotheses are provided to understand why the electrochemically formed oxides and (oxy)hydroxides catalysts derived from metal Xides are superior for OER to the as-prepared metal oxides and (oxy)hydroxides catalysts.     

Cu-assisted induced atomic-level bivalent Fe confined on N-doped carbon concave dodecahedrons for acid oxygen reduction electrocatalysis

Atomically dispersed transition metals anchored on N-doped carbon have been successfully developed as promising electrocatalysts for acidic oxygen reduction reaction (ORR). Nonetheless, how to introduce and construct single-atomic active sites is still a big challenge. Herein, a novel concave dodecahedron catalyst of N-doped carbon (FeCuNC) with well confined atomically dispersed bivalent Fe sites was facilely developed via a Cu-assisted induced strategy. The obtained catalyst delivered outstanding ORR performance in 0.5 M H₂SO₄ media with a half-wave potential (E_1/2) of 0.82 V (vs reversible hydrogen electrode, RHE), stemming from the highly active bivalent Fe-N_x sites with sufficient exposure and accessibility guaranteed by the high specific surface area and curved surface. This work provides a simple but efficient metal-assisted induced strategy to tune the configurations of atomically dispersed active sites as well as microscopy structures of carbon matrix to develop promising PGM-free catalysts for proton exchange membrane fuel cell (PEMFC) applications.     

基于决策树的羽流追踪机器人自主决策方法研究

针对无法获得可靠羽流流向信息不利于实现羽流追踪的问题，提出了一种基于决策树的羽流追踪移动机器人自主决策方法。该方法通过移动机器人两侧的浓度传感器采集到的浓度信息，利用追踪的行为规则建立决策树模型，获得行为决策信息，使机器人高效地追踪到羽流并精确地定位。由于浓度变化关系蕴含了羽流的流向及流速信息，从而取代了传统方法中流向及流速传感器。在扩散环境下，通过移动机器人羽流追踪实验，实现了良好的源定位效果。