An effective battery thermal management (BTM) system is required for lithium-ion batteries to ensure a desirable operating temperature range with minimal temperature gradient, and thus to guarantee ...their high efficiency, long lifetime and great safety. In this paper, a heat pipe and wet cooling combined BTM system is developed to handle the thermal surge of lithium-ion batteries during high rate operations. The proposed BTM system relies on ultra-thin heat pipes which can efficiently transfer the heat from the battery sides to the cooling ends where the water evaporation process can rapidly dissipate the heat. Two sized battery packs, 3 Ah and 8 Ah, with different lengths of cooling ends are used and tested through a series high-intensity discharges in this study to examine the cooling effects of the combined BTM system, and its performance is compared with other four types of heat pipe involved BTM systems and natural convection cooling method. A combination of natural convection, fan cooling and wet cooling methods is also introduced to the heat pipe BTM system, which is able to control the temperature of battery pack in an appropriate temperature range with the minimum cost of energy and water spray.
•We conduct a detailed review of the applications of recent deep learning models on machine health monitoring tasks and provide our own insights into these models.•Practical studies about ...conventional machine learning models and deep learning models on a challenging tool wear prediction have been given. Related data and code have also been open to public.•We present current deep learning works on machine health monitoring in a well-organized way to facilitate researchers to catch this topic and provide discussions about the future direction in this research topic.
Since 2006, deep learning (DL) has become a rapidly growing research direction, redefining state-of-the-art performances in a wide range of areas such as object recognition, image segmentation, speech recognition and machine translation. In modern manufacturing systems, data-driven machine health monitoring is gaining in popularity due to the widespread deployment of low-cost sensors and their connection to the Internet. Meanwhile, deep learning provides useful tools for processing and analyzing these big machinery data. The main purpose of this paper is to review and summarize the emerging research work of deep learning on machine health monitoring. After the brief introduction of deep learning techniques, the applications of deep learning in machine health monitoring systems are reviewed mainly from the following aspects: Auto-encoder (AE) and its variants, Restricted Boltzmann Machines and its variants including Deep Belief Network (DBN) and Deep Boltzmann Machines (DBM), Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN). In addition, an experimental study on the performances of these approaches has been conducted, in which the data and code have been online. Finally, some new trends of DL-based machine health monitoring methods are discussed.
Display omitted
•A total nitrogen removal rate of 0.18gNg−1VSSd−1 was obtained in the reactor.•Flexible flocs and compact granules were both found in the system.•Proteobacteria, Chloroflexi and ...Planctomycetes were selected over the startup.•Filamentous bacteria constitute a large portion (>25%) in the reactor.
The combined nitritation–anammox process has recently been studied extensively from an engineering perspective. However, the importance of microbial communities of this process was generally underestimated. In this study, a lab-scale nitritation–anammox sequencing batch reactor (SBR) was established and the microbial community structure was further characterized, in order to provide the comprehensive insight into the key microbial groups in one-stage nitritation–anammox system. In general, a total nitrogen removal rate of 0.18gNg−1VSSd−1 was obtained after 180days when the nitrogen loading rate was 0.5kgNm−3d−1 (hydraulic retention time of 1d). Flexible flocs and compact granules were both found in the system, and this morphological differences were basically caused by the different microbial compositions, that is, flocs mainly consisted of filamentous bacteria and granules dominated by anammox bacteria. Results from high-throughput sequencing analysis revealed that Proteobacteria, Chloroflexi and Planctomycetes were selected and established a stable foothold in the community over the startup period, probably driven by the availability of substrate in the influent. Apart from nitrifiers of the Proteobacteria and anammox bacteria of the Planctomycetes, members of Chloroflexi constitute a large portion (>25%), which indicate that heterotrophs (Chloroflexi) survived by soluble microbial products (SMP) of autotrophs should not be neglected in the autotrophic system. This study could be useful for better understanding of one-stage nitritation–anammox system, especially for the interaction between autotrophs and heterotrophs in the system.
Diabetes is a disorder of glucose metabolism, and over 90% are type 2 diabetes. Diabetic cardiomyopathy (DCM) is one of the type 2 diabetes complications, usually accompanied by changes in myocardial ...structure and function, together with cardiomyocyte apoptosis. Our study investigated the effect of curcumin on regulating oxidative stress (OS) and apoptosis in DCM. In vivo, diabetes was induced in an experimental rat model by streptozoticin (STZ) together with high‐glucose and high‐fat (HG/HF) diet feeding. In vitro, H9c2 cardiomyocytes were cultured with high‐glucose and saturated free fatty acid palmitate. Curcumin was orally or directly administered to rats or cells, respectively. Streptozoticin ‐induced diabetic rats showed metabolism abnormalities and elevated markers of OS (superoxide dismutase SOD, malondialdehyde MDA, gp91phox, Cyt‐Cyto C), enhanced cell apoptosis (Bax/Bcl‐2, Cleaved caspase‐3, TUNEL‐positive cells), together with reduced Akt phosphorylation and increased Foxo1 acetylation. Curcumin attenuated the myocardial dysfunction, OS and apoptosis in the heart of diabetic rats. Curcumin treatment also enhanced phosphorylation of Akt and inhibited acetylation of Foxo1. These results strongly suggest that apoptosis was increased in the heart of diabetic rats, and curcumin played a role in diabetic cardiomyopathy treatment by modulating the Sirt1‐Foxo1 and PI3K‐Akt pathways.
Low-level saliency cues or priors do not produce good enough saliency detection results especially when the salient object presents in a low-contrast background with confusing visual appearance. This ...issue raises a serious problem for conventional approaches. In this paper, we tackle this problem by proposing a multi-context deep learning framework for salient object detection. We employ deep Convolutional Neural Networks to model saliency of objects in images. Global context and local context are both taken into account, and are jointly modeled in a unified multi-context deep learning framework. To provide a better initialization for training the deep neural networks, we investigate different pre-training strategies, and a task-specific pre-training scheme is designed to make the multi-context modeling suited for saliency detection. Furthermore, recently proposed contemporary deep models in the ImageNet Image Classification Challenge are tested, and their effectiveness in saliency detection are investigated. Our approach is extensively evaluated on five public datasets, and experimental results show significant and consistent improvements over the state-of-the-art methods.
Person re-identification is to match pedestrian images from disjoint camera views detected by pedestrian detectors. Challenges are presented in the form of complex variations of lightings, poses, ...viewpoints, blurring effects, image resolutions, camera settings, occlusions and background clutter across camera views. In addition, misalignment introduced by the pedestrian detector will affect most existing person re-identification methods that use manually cropped pedestrian images and assume perfect detection. In this paper, we propose a novel filter pairing neural network (FPNN) to jointly handle misalignment, photometric and geometric transforms, occlusions and background clutter. All the key components are jointly optimized to maximize the strength of each component when cooperating with others. In contrast to existing works that use handcrafted features, our method automatically learns features optimal for the re-identification task from data. The learned filter pairs encode photometric transforms. Its deep architecture makes it possible to model a mixture of complex photometric and geometric transforms. We build the largest benchmark re-id dataset with 13, 164 images of 1, 360 pedestrians. Unlike existing datasets, which only provide manually cropped pedestrian images, our dataset provides automatically detected bounding boxes for evaluation close to practical applications. Our neural network significantly outperforms state-of-the-art methods on this dataset.
•A coupling model is developed to study the behaviors of Li-ion batteries.•Thick electrode battery (CEB) has high temperature response during discharge.•Thin electrode battery has a relative lower ...capacity fading rate.•Less heat is generated in thin electrode battery with even heat distribution.•CEBs underutilize active materials and stop discharge early at high rates.
Lithium ion (Li-ion) battery, consisting of multiple electrochemical cells, is a complex system whose high electrochemical and thermal stability is often critical to the well-being and functional capabilities of electric devices. Considering any change in the specifications may significantly affect the overall performance and life of a battery, an investigation on the impacts of electrode thickness on the electrochemical and thermal properties of lithium-ion battery cells based on experiments and a coupling model composed of a 1D electrochemical model and a 3D thermal model is conducted in this work. In-depth analyses on the basis of the experimental and simulated results are carried out for one cell of different depths of discharge as well as for a set of cells with different electrode thicknesses. Pertinent results have demonstrated that the electrode thickness can significantly influence the battery from many key aspects such as energy density, temperature response, capacity fading rate, overall heat generation, distribution and proportion of heat sources.
Abstract
Achieving CO
2
reduction with H
2
O on metal photocatalysts and understanding the corresponding mechanisms at the molecular level are challenging. Herein, we report that quantum-sized Au ...nanoparticles can photocatalytically reduce CO
2
to CO with the help of H
2
O by electron-hole pairs mainly originating from interband transitions. Notably, the Au photocatalyst shows a CO production rate of 4.73 mmol g
−1
h
−1
(~100% selectivity), ~2.5 times the rate during CO
2
reduction with H
2
under the same experimental conditions, under low-intensity irradiation at 420 nm. Theoretical and experimental studies reveal that the increased activity is induced by surface Au–O species formed from H
2
O decomposition, which synchronously optimizes the rate-determining steps in the CO
2
reduction and H
2
O oxidation reactions, lowers the energy barriers for the *CO desorption and *OOH formation, and facilitates CO and O
2
production. Our findings provide an in-depth mechanistic understanding for designing active metal photocatalysts for efficient CO
2
reduction with H
2
O.
A convolutional discriminative feature learning method is presented for induction motor fault diagnosis. The approach firstly utilizes back-propagation (BP)-based neural network to learn local ...filters capturing discriminative information. Then, a feed-forward convolutional pooling architecture is built to extract final features through these local filters. Due to the discriminative learning of BP-based neural network, the learned local filters can discover potential discriminative patterns. Also, the convolutional pooling architecture is able to derive invariant and robust features. Therefore, the proposed method can learn robust and discriminative representation from the raw sensory data of induction motors in an efficient and automatic way. Finally, the learned representations are fed into support vector machine classifier to identify six different fault conditions. Experiments performed on a machine fault simulator indicate that compared with the current state-of-the-art methods, the proposed method shows significant performance gains, and it is effective and efficient for induction motor fault diagnosis.