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.     

Machine fault diagnosis with small sample based on variational information constrained generative adversarial network

In actual engineering scenarios, limited fault data leads to insufficient model training and over-fitting, which negatively affects the diagnostic performance of intelligent diagnostic models. To solve the problem, this paper proposes a variational information constrained generative adversarial network (VICGAN) for effective machine fault diagnosis. Firstly, by incorporating the encoder into the discriminator to map the deep features, an improved generative adversarial network with stronger data synthesis capability is established. Secondly, to promote the stable training of the model and guarantee better convergence, a variational information constraint technique is utilized, which constrains the input signals and deep features of the discriminator using the information bottleneck method. In addition, a representation matching module is added to impose restrictions on the generator, avoiding the mode collapse problem and boosting the sample diversity. Two rolling bearing datasets are utilized to verify the effectiveness and stability of the presented network, which demonstrates that the presented network has an admirable ability in processing fault diagnosis with few samples, and performs better than state-of-the-art approaches.     

Nacre-Inspired,Liquid Metal-Based Ultrasensitive Electronic Skin by Spatially Regulated Cracking Strategy

The realization of liquid metal-based wearable systems will be a milestone toward high-performance, integrated electronic skin. However, despite the revolutionary progress achieved in many other components of electronic skin, liquid metal-based flexible sensors still suffer from poor sensitivity due to the insufficient resistance change of liquid metal to deformation. Herein, a nacre-inspired architecture composed of a biphasic pattern (liquid metal with Cr/Cu underlayer) as “bricks” and strain-sensitive Ag film as “mortar” is developed, which breaks the long-standing sensitivity bottleneck of liquid metal-based electronic skin. With 2 orders of magnitude of sensitivity amplification while maintaining wide (>85%) working range, for the first time, liquid metal-based strain sensors rival the state-of-art counterparts. This liquid metal composite features spatially regulated cracking behavior. On the one hand, hard Cr cells locally modulate the strain distribution, which avoids premature cut-through cracks and prolongs the defect propagation in the adjacent Ag film. On the other hand, the separated liquid metal cells prevent unfavorable continuous liquid-metal paths and create crack-free regions during strain. Demonstrated in diverse scenarios, the proposed design concept may spark more applications of ultrasensitive liquid metal-based electronic skins, and reveals a pathway for sensor development via crack engineering.     

Decreased CD10-positive granulocytes for the differential diagnosis of myelodysplastic syndrome

Myelodysplastic syndromes (MDS) are highly heterogeneous myeloid neoplasms, and a large number of patients are difficult to diagnose and classify by blood and bone marrow examination. As a surface marker of granulocyte, studies have shown CD10 can be used to define the degree of granulocyte maturation in MDS patients. However, whether it can be used for differential diagnosis of MDS and other hematological diseases remains inconclusive. To explore the value of CD10 for differential diagnosis of MDS, 60 newly diagnosed MDS, 20 aplastic anemia (AA) patients, and 35 iron-deficient anemia (IDA) patients were selected for this study. Bone marrow (BM) specimens were processed for surface marker analysis and labeled with pre-conjugated monoclonal antibodies. Stained cells were detected by flow cytometry. Our results indicated that CD10-positive granulocytes were significantly decreased in BM of MDS patients than AA and IDA patients, and the level of CD10-positive mature granulocytes was not associated with the clinical stages of malignancy. Receiver operating characteristic (ROC) areas under the curve (AUC) of CD10-positive granulocytes was 0.86 and 0.85, respectively, in MDS patients than the IDA group and AA group with good specificity and sensitivity. Further, CD10-positive granulocytes were increased after effective treatment. In conclusion, we found the decrease in CD10-positive granulocytes has a differential diagnostic value of MDS.     

A Voltage Control Method for a Distributed Line Feeder by means of Multiagent Approach

This paper proposes a method for the coordinated control of power factor by means of a multiagent approach. The proposed multiagent system consists of two types of agent: single feeder agent (F_AG) and bus agent (B_AG). In the proposed system, an F_AG plays as an important role, which decides the power factors of all distributed generators by executing the load flow calculations repeatedly. The voltage control strategies are implemented as the class definition of Java into the system. In order to verify the performance of the proposed method, it has been applied to a typical distribution model system. The simulation results show that the system is able to control very violent fluctuation of the demands and the photovoltaic (PV) generations.     

基于残差决策的自动飞行控制系统远程实时故障诊断

针对自动飞行控制系统结构复杂、关联部件众多,发生故障时诊断时间长,从而影响飞机运行效率的问题,提出一种基于飞机通信寻址报告系统(ACARS)的远程实时故障诊断方案。首先,分析自动飞行控制系统的故障特点,设计搭建检测滤波器;然后,利用ACARS数据链实时发送的自动飞行控制系统的关键信息进行相关部件的残差计算,并根据残差决策算法进行故障诊断及定位;最后,针对不同故障部件残差间的差异大、决策门限无法统一的缺点,提出基于二次差值的残差决策改进算法,减缓了检测对象的整体变化趋势,降低了随机噪声和干扰的影响,避免了将瞬态故障诊断为系统故障的情况。实验仿真结果表明,基于二次差值的改进残差决策算法避免了多决策门限的复杂性,在采样时间为0.1 s的情况下,故障检测所需时间大约为2 s,故障检测时间大幅降低,有效故障检测率大于90%。     

基于深度信念网络的列车故障音频诊断方法

铁路在交通运输行业有着举足轻重的地位，一旦列车发生故障将会导致严重的生命财产损失。由于列车发生故障的概率相对较低，因此难以捕获列车的故障样本。针对上述问题，提出了一种无监督学习的列车故障识别方法，通过检测列车音频信号来识别列车故障。该方法基于深度信念网络(DBN)，利用小波包分解提取检测信号的特征向量并将其作为DBN的输入，待网络充分训练后，由训练好的DBN识别当前列车的运行状况。现场监测实验结果表明，该方法能够在无监督的条件下有效识别列车故障，保障了列车的运行安全。     

Magnetic Resonance Signal Amplification by Reversible Exchange of Selective PyFALGEA Oligopeptide Ligands Towards Epidermal Growth Factor Receptors

The biorelevant PyFALGEA oligopeptide ligand, which is selective towards the epidermal growth factor receptor (EGFR), has been successfully employed as a substrate in magnetic resonance signal amplification by reversible exchange (SABRE) experiments. It is demonstrated that PyFALGEA and the iridium catalyst IMes form a PyFALGEA:IMes molecular complex. The interaction between PyFALGEA:IMes and H₂ results in a ternary SABRE complex. Selective 1D EXSY experiments reveal that this complex is labile, which is an essential condition for successful hyperpolarization by SABRE. Polarization transfer from parahydrogen to PyFALGEA is observed leading to significant enhancement of the ¹H NMR signals of PyFALGEA. Different iridium catalysts and peptides are inspected to discuss the influence of their molecular structures on the efficiency of hyperpolarization. It is observed that PyFALGEA oligopeptide hyperpolarization is more efficient when an iridium catalyst with a sterically less demanding NHC ligand system such as IMesBn is employed. Experiments with shorter analogues of PyFALGEA, that is, PyLGEA and PyEA, show that the bulky phenylalanine from the PyFALGEA oligopeptide causes steric hindrance in the SABRE complex, which hampers hyperpolarization with IMes. Finally, a single-scan ¹H NMR SABRE experiment of PyFALGEA with IMesBn revealed a unique pattern of NMR lines in the hydride region, which can be treated as a fingerprint of this important oligopeptide.