Multi-mode non-Gaussian variational autoencoder network with missing sources for anomaly detection of complex electromechanical equipment.

Anomaly detection is crucial to the safety of complex electromechanical equipment. With the rapid accumulation of industrial data, intelligent methods without human intervention have become the mainstream of anomaly detection. Among them, variational autoencoder (VAE) performs well in anomaly detection with missing fault samples due to the self-supervised learning paradigm. However, the data from electromechanical equipment is usually non-Gaussian, making it difficult for the standard VAE based on Gaussian distribution to recognize the abnormal states. To solve the above problems, we proposed multi-mode non-Gaussian VAE (MNVAE) to detect anomalies from unknown distribution vibration signals without fault samples or prior knowledge. Firstly, the encoder maps the input to a Gaussian mixture distribution in latent space and samples a latent variable from it, after which the Householder Flow is applied to the latent variable to capture more abundant features. Finally, to describe the non-Gaussianity of the signal, Weibull distribution serves as the likelihood function of the reconstructed signal output from the decoder and as the basis for anomaly discrimination. In comparison to 6 related methods, our method yields the best results across various datasets. Through further experiments, the robustness of our method is proved and the proposed improvements are effective.

Intelligent fault diagnosis scheme via multi-module supervised-learning network with essential features capture-regulation strategy.

The performance of data driven-based intelligent diagnosis method greatly depends on the quantity and quality of data. Nevertheless, due to realistic limitations, failure data is hard to acquire, which makes the training process of numerous intelligent models unsatisfactory and leads to performance degradation Aiming at this problem, considering the local impulse characteristics as minimum diagnosable units, this paper proposes a signal adaptive augmentation network (SAAN) to effectively construct artificial samples for amplifying fault data volume. The SAAN consists of three sub-structures: impulse extractor, regulator, and classifier. The impulse extractor combines inner product matching principle to extract the local impulse features from insufficient samples to construct massive initial artificial samples. The regulator adopts convolution and deconvolution frameworks to regulate and reconstruct the initial artificial samples by specially designed synthetic loss function, which makes artificial samples have same characteristic distribution as real samples. The augmented method is used for validation on three bearing data with some advanced algorithms. Besides, a focal normalized network is designed for classification under small samples. Relevant experiments indicate that the SAAN shows a competitive performance with existing state-of-art diagnostic methods, which can helpfully improve recognition accuracies of various diagnostic models by 5%-35% under small sample prerequisite.

Heterogeneous bi-directional recurrent neural network combining fusion health indicator for predictive analytics of rotating machinery.

Data-driven intelligent methods arise the increasing demand for predictive analytics to evaluate the operational reliability and natural degradation of rotating machinery. Nevertheless, accurate and timely predictive analytics is still regarded as an extremely challenging mission, because the quality of predictive maintenance depends not only on the capability of intelligent model, but also on the construction of effective health indicators To overcome this issue, a novel heterogeneous bi-directional gated recurrent unit (GRU) model combining with fusion health indicator (Fusion-HI) is proposed for predictive analytics in this paper. First, the support evidence space is constructed to reflect the operating state of mechanical equipment. Then the evidence features from multiple domains are integrated to obtain the optimal Fusion-HI by the modified de-noising auto-encoder (MDAE). Finally, a hybrid prediction network is designed combining with the gate attention algorithm, which consists of multi-scale convolution layers, bi-directional GRU layers, smoothed and de-noised layers, and regression layers. Three experimental whole lifetime data and one industrial entire life cycle data are analyzed to validate the feasibility of the proposed approach in two case studies respectively. Relevant experimental results indicate that the Fusion-HI is capable to sensitively characterize the degradation state of equipment, while the prediction accuracy of presented heterogeneous model is superior to that of conventional prediction approaches.

Intelligent Impulse Finder: A boosting multi-kernel learning network using raw data for mechanical fault identification in big data era.

Nowadays, most intelligent diagnosis methods focus on fault classification and the discriminative knowledge is unknown due to the 'black box' characteristic. However, impulse responses in vibration signals, which is important sign to determine whether mechanical equipment is faulty, are rarely studied under intelligent methods since their recognition is both difficult and time-consuming, especially mixed with noise. Aiming at these problems, a novel impulse recognition method was proposed to capture them from raw mechanical data. Firstly, a single-kernel convolutional neural network is proposed as weak classifier to learn discriminative information from raw data. Then, a coarse-to-fine search is proposed to locate position of impulse response. Finally, the boosting algorithm is used to ensemble several proposed weak classifiers for final output. Vibration signals of bearings with two different faults are utilized to validate the proposed model. The results prove that the proposed approach obtain higher accuracy compared with traditional Laplace wavelet method. Moreover, the extracted kernel functions reveal new knowledge about characteristics of impulse responses, which significantly differs from traditional hypothesis and sheds a light on improvement of relevant approaches.

Intelligent fault identification for industrial automation system via multi-scale convolutional generative adversarial network with partially labeled samples.

Rolling bearings are the widely used parts in most of the industrial automation systems. As a result, intelligent fault identification of rolling bearing is important to ensure the stable operation of the industrial automation systems. However, a major problem in intelligent fault identification is that it needs a large number of labeled samples to obtain a well-trained model. Aiming at this problem, the paper proposes a semi-supervised multi-scale convolutional generative adversarial network for bearing fault identification which uses partially labeled samples and sufficient unlabeled samples for training. The network adopts a one-dimensional multi-scale convolutional neural network as the discriminator and a multi-scale deconvolutional neural network as the generator and the model is trained through an adversarial process. Because of the full use of unlabeled samples, the proposed semi-supervised model can detect the faults in bearings with limited labeled samples. The proposed method is tested on three datasets and the average classification accuracy arrived at of 100%, 99.28% and 96.58% respectively Results indicate that the proposed semi-supervised convolutional generative adversarial network achieves satisfactory performance in bearing fault identification when the labeled data are insufficient.

Upregulation of microRNA­300 induces the proliferation of liver cancer by downregulating transcription factor FOXO1.

In the present study, we investigated whether miRNA­300 (miR­300) is an oncogene in human liver cancer and sought to determine the mechanism underlying its activity. We also investigated the effect of miRNA­300 on the growth in liver cancer. To identify its target molecule, we performed luciferase assays. The downstream signaling pathway was detected by immunohistochemical (IHC) analysis in human HCC tissues, and the protein levels of AKT, 4E­BP1, S6K1, SNAIL and MMP2 were determined using western blotting. miR­300 levels were higher in patients with high­stage HCC, and miR­300 promoted cell growth both in vitro and in vivo. miRNA­300 inhibited the luciferase activity of FOXO1 by targeting its 3'­untranslated region (UTR), and overexpression of miR­300 upregulated the protein levels of phospho­AKT, phospho­4E­BP1, phospho­S6K1, SNAIL and MMP2. These data revealed that miRNA­300 functions as an oncogene in liver cancer by inhibiting FOXO1 and interacting with the AKT/mTOR signaling pathway.

Long non-coding RNA FOXD2-AS1 plays an oncogenic role in hepatocellular carcinoma by targeting miR­206.

Recently, long non-coding RNA (lncRNA) FOXD2 adjacent opposite strand RNA 1 (FOXD2-AS1) has been recognized to function as an oncogene in several human tumors, and FOXD2­AS1 dysregulation has been closely associated with carcinogenesis and tumor progression. Nevertheless, the correlation between the aberrant expression of FOXD2­AS1 and the prognosis of hepatocellular carcinoma (HCC) has not yet been elucidated. In the present study, FOXD2­AS1 was found to be overexpressed in HCC tissues, and FOXD2­AS1 overexpression resulted in significantly shortened patient survival. FOXD2­AS1 overexpression enhanced the viability and metastasis of HCC cells in vitro and in vivo, as revealed by MTT, wound healing and cell migration assays. In addition, mechanistic studies revealed that FOXD2­AS1 upregulated the expression of the miR­206 target gene annexin A2 (ANXA2) by acting as a miR­206 sponge. In summary, FOXD2­AS1 was concluded to function as an oncogene in HCC and to upregulate ANXA2 expression in part by 'sponging' miR­206.

Curcumin inhibits hypoxia inducible factor­1α­induced epithelial­mesenchymal transition in HepG2 hepatocellular carcinoma cells.

Hepatocellular carcinoma (HCC) accounts for the majority of liver cancers. A hypoxic microenvironment is a common feature of HCC, and is associated with malignant invasion, metastasis and epithelial-mesenchymal transition (EMT) changes. Curcumin is a botanical agent derived from the dried rhizome of Curcuma longa. Although a number of preclinical studies have shown that curcumin has anticancer properties when administered in a normoxic microenvironment, no studies have directly examined the effect of curcumin on preventing HCC invasion and metastasis under hypoxic conditions. This study aimed to determine whether curcumin has effects on the hypoxia-induced malignant biological behavior of HCC. CoCl2 was used to establish a hypoxia model in vitro. The results showed that curcumin significantly decreased hypoxia-induced hypoxia inducible factor-1α (HIF-1α) protein level in HepG2 cells. Furthermore, cell proliferation, migration and invasiveness, as well as EMT changes associated with HIF-1α accumulation generated by a hypoxic microenvironment, were eliminated by curcumin. In conclusion, these data indicate that curcumin may be a viable anticancer agent in the treatment of HCC.

The SDF-1/CXCR4 axis induces epithelial­mesenchymal transition in hepatocellular carcinoma.

Hepatic stellate cells play a role in the migration process of hepatocellular carcinoma cells. Here, we address the role of the stromal-derived factor-1/CXC chemokine receptor 4 (SDF-1/CXCR4) axis on hepatocellular carcinoma progression. The expression of the SDF-1 and the CXCR4 was determined through western blotting and real-time PCR analysis using hepatic stellate (LX02) and hepatocellular carcinoma (MHCC97, SMMC7721, Hep3B, and HepG2) cell lines depleted of CXCR4 using shRNA. The migration of hepatocellular carcinoma cells following exogenous treatment with SDF-1 or in co-culture cell systems was measured using the Transwell assay. In parallel, the expression of epithelial­mesenchymal transition (EMT) markers was also determined. We found that SDF-1 is highly expressed in the hepatic stellate cell line LX02 and that the hepatocellular carcinoma cells express high levels of CXCR4. Co-culturing hepatocellular carcinoma cells with LX02 or exogenous treatment with SDF-1 induced an EMT as shown by increased migration. In contrast, ablation of CXCR4 expression in HepG2 cells attenuated the migration of HepG2 cells and suppressed the EMT. Thus, hepatic stellate cells can promote hepatocellular carcinoma cell invasion through the SDF-1/CXCR4 axis.

Notch1 silencing inhibits proliferation and invasion in SGC­7901 gastric cancer cells.

Downregulation of Notch1 has been shown to exert antineoplastic effects in vivo and in vitro. However, the role of the Notch1 gene in the proliferative and invasive ability of gastric cancer cells is not clear. In this study, we investigated the effect of Notch1 gene silencing on the proliferation and invasion of gastric cancer SGC­7901 cells. Small interfering RNA (siRNA) targeting Notch1 was transfected into SGC­7901 cells using Lipofectamine 2000. Proliferation of SGC­7901 cells was then determined by the MTT assay. Notch1 mRNA expression was determined by reverse transcription­polymerase chain reaction (RT­PCR). Invasion of the SGC­7901 cells was detected by the Transwell assay. The protein levels of cyclin D1, cyclin A1 and cyclin-dependent kinase 2 (CDK2) were determined by western blotting. The mRNA levels of matrix metalloproteinase­2 (MMP­2) and cyclooxygenase­2 (COX­2) were determined by RT­PCR. Compared to the control group, the Notch1 mRNA level was significantly decreased following transfection. The growth and invasion rates of SGC­7901 cells were significantly reduced after Notch1 silencing. Additionally, the expression of cyclin D1 and cyclin A1 proteins and of the MMP­2 and COX­2 mRNAs was markedly attenuated. From these results, it was concluded that Notch1 gene silencing inhibits the proliferation of gastric SGC­7901 cells by decreasing the expression of cyclins D1 and A1, and reduces the invasive ability of SGC­7901 cells through the downregulation of MMP­2 and COX­2 genes. Thus, silencing of the Notch1 pathway may be a novel approach in the treatment of gastrointestinal cancer.

[Curcumin inhibited hypoxia induced epithelial-mesenchymal transition in hepatic carcinoma cell line HepG2 in vitro].

OBJECTIVE: To explore effects and possible mechanisms of curcumin on hypoxia induced epithelial-mesenchymal transition (EMT) in hepatocellular carcinoma cell line HepG2. METHODS: HepG2 cells were divided to 3 groups, i.e., the normal control group, the CoCl2 group, and the CoCl2 plus 10 micromol/L curcumin group. The proliferation of HepG2 was determined using MTT assay. The migration of HepG2 was detected by wound healing assay.The mRNA expression of hypoxia-inducible factor-1 (HIF-1alpha) was evaluated with real-time RT-PCR. The protein expressions of HIF-1alpha, epithelial-cadherin (E-cadherin), and vimentin were determined using Western blot. RESULTS: Compared with the normal control group, the proliferation and migration of HepG2 cells under CoCl2-induced hypoxia significantly increased, the expression of HIF-1alpha was up-regulated, and the expression of E-cadherin protein was obviously down-regulated, and the expression of vimentin significantly increased (all P < 0.05). Intervention by curcumin significantly inhibited the proliferation and migration of hypoxic HepG2 cells, and expressions of HIF-1alpha and vimentin decreased, and the expression of E-cadherin was up-regulated, showing statistical difference when compared with those of the CoCl2 group (P < 0.05). There was no statistical difference in HIF-1alpha mRNA expression among the 3 groups (P > 0.05). CONCLUSION: Curcumin could reverse the proliferation and migration of HepG2 cells under CoCl2-induced hypoxia condition, which might be associated with inhibiting up-regulated expressions of HIF-1alpha protein and EMT.