Machine learning algorithm selection for windage alteration fault diagnosis of mine ventilation system

Machine learning algorithms have been widely used in mine fault diagnosis. The correct selection of the suitable algorithms is the key factor that affects the fault diagnosis. However, the impact of machine learning algorithms on the prediction performance of mine fault diagnosis models has not been fully evaluated. In this study, the windage alteration faults (WAFs) diagnosis models, which are based on K-nearest neighbor algorithm (KNN), multi-layer perceptron (MLP), support vector machine (SVM), and decision tree (DT), are constructed. Furthermore, the applicability of these four algorithms in the WAFs diagnosis is explored by a T-type ventilation network simulation experiment and the field empirical application research of Jinchuan No. 2 mine. The accuracy of the fault location diagnosis for the four models in both networks was 100%. In the simulation experiment, the mean absolute percentage error (MAPE) between the predicted values and the real values of the fault volume of the four models was 0.59%, 97.26%, 123.61%, and 8.78%, respectively. The MAPE for the field empirical application was 3.94%, 52.40%, 25.25%, and 7.15%, respectively. The results of the comprehensive evaluation of the fault location and fault volume diagnosis tests showed that the KNN model is the most suitable algorithm for the WAFs diagnosis, whereas the prediction performance of the DT model was the second-best. This study realizes the intelligent diagnosis of WAFs, and provides technical support for the realization of intelligent ventilation.     

Experimental Study of Oilfield Water Plugging with Organic Chromium Gel

Chemistry and Technology of Fuels and Oils - Application of water flooding technologies in the oilfield causes intensive heterogeneity of the oil reservoir and improper circulation of the injected...     

结合优化支持向量机与K-means++的工控系统入侵检测方法

针对工业控制系统传统单一检测算法模型对不同攻击类型检测率和检测速度不佳的问题，提出一种优化支持向量机和K-means++算法结合的入侵检测模型。首先利用主成分分析法（PCA）对原始数据集进行预处理，消除其相关性；其次在粒子群优化（PSO）算法的基础上加入自适应变异过程避免在训练的过程中陷入局部最优解；然后利用自适应变异粒子群优化（AMPSO）算法优化支持向量机的核函数和惩罚参数；最后利用密度中心法改进K-means算法与优化后的支持向量机组合成入侵检测模型，从而实现工业控制系统的异常检测。实验结果表明，所提方法在检测速度和对各类攻击的检测率上得到明显提升。     

因素空间理论下基点分类算法研究

目前，基于因素空间理论的背景基提取算法计算过程复杂，初始化必须依赖各因素极值，基点数量提取冗余等原因，未能在应用中取得很好效果。为此，结合内点判别法和知识可继承、可扩展的思想，提出一种计算简单、初始化独立、基点数量小的改进的背景基提取算法。然后，利用改进的背景基提取算法构造出一种全新的数据分类算法-基点分类算法，基点分类算法以提取每一类样本的背景基为预测模型，再通过新定义的λ-背景基，优化预测模型。数值实验表明:基点分类算法原理简单、构造难度小、分类模型泛化能力强，预测能力准确率高，同时严格的模型限定区域又能为识别新类别提供新方法。     

正交特征群

本文在证明了有限Ａｂｅ１群Ｇ与它的特征群Ｇ√存在同构的一个必要充分条件之后，构造了一类正交特征群，从而概括了相当一大类正交变换（它包括了离散付里叶变换、阿达玛变换、数论变换、Ｃｈｒｅｓｔｅｎｓｏｎ变换）这对构成有限域与整环上的正交变换具有指导作用。     

改进异步电机直接转矩控制系统的仿真研究

本文介绍了直接转矩控制系统的控制原理,并针对基于直接转矩控制的异步电动机运行时存在较大的电流及转矩脉动问题,提出一种新的控制方案,并基于Matlab 6.5对这种新方法进行仿真,仿真结果表明该方案能有效的解决电流及转矩脉动的问题.     

由碳五馏分合成甲基叔戊基醚工艺的研究

以催化裂化C_5馏分与工业甲醇为原料,D005型大孔强酸型阳离子交换树脂为催化剂,采用筒式膨胀床双反应器,考察了甲基叔戊基醚的合成工艺条件。一阶反应器的操作条件为：循环比为2：1,醇烯比为1.2：1,入口温度为70℃,空速为6 h~(-1),平均转化率为53.63%；二阶反应器的操作条件为：入口温度为50℃,空速为1.2h~(-1),平均转化率为71.32%,全调合方案可以使催化裂化汽油的辛烷值平均提高0.55个单位。     

Disposal of Waste Plastics With Traditional Coking Process

A new technology for treating waste plastics （WP） by traditional coking process was introduced. With a thermo-balance and a 10 g atmospheric fixed bed reactor, the thermo-gravimetric behavior and product were studied during co-coking of WP with blended coal. And then, using a coke-oven with capacity of 200 kg, the characteristics of products were assessed. The results showed that there is an overlapping temperature range （200-550℃ ） of decomposition between WP and blended coal, and the pyrolysis synergism index η and synergism strength β proposed could evaluate the synergism between them. 1% of added WP results in the maximum synergism in all series experiments. The increase of added WP decreases the synergism. Tar yield in co-coking is increased with the decrease of water yield for synergism. Moreover, it was also found that the quality indexes of coke, such as M10, M40, CRI and CSR, are degraded with the increase of WP until 4%, though the quality of tar and gas is optimized for WP addition.     

基于一种混合的节点定位算法的研究

针对基本的DV-Hop节点定位算法会产生不良的节点和多级跳带来的累积误差以及未知节点的坐标没有进行修正等缺点,提出了一种改进的混合节点定位算法WHDV-Hop算法,改进后的算法定位误差率明显比改进前的定位误差率小,通过MATLAB软件对节点定位误差方面进行仿真验证,仿真实验结果证明该改进后的算法稳定、可靠,易于实现。与DV-Hop算法相比,定位精确度提高了4.65%,计算量减少,降低了能量消耗水平,实现预期的目标。     

基于多模态推理图神经网络的场景文本视觉问答模型

文本阅读能力差和视觉推理能力不足是现有视觉问答(visual question answering, VQA)模型效果不好的主要原因,针对以上问题,设计了一个基于图神经网络的多模态推理(multi-modal reasoning graph neural network, MRGNN)模型。利用图像中多种形式的信息帮助理解场景文本内容,将场景文本图片分别预处理成视觉对象图和文本图的形式,并且在问题自注意力模块下过滤多余的信息;使用加入注意力的聚合器完善子图之间相互的节点特征,从而融合不同模态之间的信息,更新后的节点利用不同模态的上下文信息为答疑模块提供了更好的功能。在ST-VQA和TextVQA数据集上验证了有效性,实验结果表明,相比较此任务的一些其他模型,MRGNN模型在此任务上有明显的提升。