Search

Electronic skin (E-skin) is increasingly utilized in modern society, yet current E-skin technology suffers from issues, such as opacity, hardness, and fragility. To address these challenges, a novel E-skin was developed using polyurethane (PU) as the matrix material and silver nanowires (AgNWs) as the sensing material. By leveraging the small degree of microphase separation and lack of crystallization in the PU, combined with the appropriate length–diameter ratio of the AgNWs, the resulting E-skin exhibited a visible light transmittance of 75%. The E-skin also showed excellent self-healing properties (83.63% efficiency in the third repair) and mechanical properties (with almost no degradation after 60 tensile cycles) due to the reversible dynamic cross-linking network within the PU. The synergistic effect of PU and AgNWs resulted in exceptional sensing performance for the E-skin, with a gauge factor of 46 (when ε = 10%). Moreover, the E-skin demonstrated signal stability during human joint motion monitoring and successfully identified different movement states, highlighting its potential for diverse applications. This research presents a simple yet effective approach for producing transparent, durable, and stable E-skin.

A low band gap acceptor F-ThBr is designed for the front cell and CH1007 is selected for the rear cell. Using the two acceptors, a tandem OSC is fabricated and offers an efficiency of 18.67%.

Suppressing intramolecular vibration of non-fullerene acceptors (NFAs) by molecular rigidification has been proven to be an effective way to reduce the non-radiative recombination loss and energetic disorder of organic solar cells (OSCs). Thus far, extensive attention has been drawn on rigidifying the fused-ring backbones of NFAs, whereas the highly flexible alkyl side chains are barely concerned. Herein, an effective strategy of side chain rigidification by introducing a spiro-ring is developed for the first time and applied to construct the NFA of Spiro-F. Compared to its counterpart F-2F, the rigid spirocyclic side chain can constrain the vibrational-rotational motion and control the orientation of two highly flexible

Good temporal representations are crucial for video understanding, and the state-of-the-art video recognition framework is based on two-stream networks. In such framework, besides the regular ConvNets responsible for RGB frame inputs, a second network is introduced to handle the temporal representation, usually the optical flow (OF). However, OF or other task-oriented flow is computationally costly, and is thus typically pre-computed. Critically, this prevents the two-stream approach from being applied to reinforcement learning (RL) applications such as video game playing, where the next state depends on current state and action choices. Inspired by the early vision systems of mammals and insects, we propose a fast event-driven representation (EDR) that models several major properties of early retinal circuits: (1) logarithmic input response, (2) multi-timescale temporal smoothing to filter noise, and (3) bipolar (ON/OFF) pathways for primitive event detection. Trading off the directional information for fast speed ( $>$ 9000 fps), EDR enables fast real-time inference/learning in video applications that require interaction between an agent and the world such as game-playing, virtual robotics, and domain adaptation. In this vein, we use EDR to demonstrate performance improvements over state-of-the-art reinforcement learning algorithms for Atari games, something that has not been possible with pre-computed OF. Moreover, with UCF-101 video action recognition experiments, we show that EDR performs near state-of-the-art in accuracy while achieving a 1,500x speedup in input representation processing, as compared to optical flow.

In order to solve the problems of high voltage switches’ operation characteristic tests, such as no double end grounding measurement, the inconvenient opening of grounding switch, easy injury, and equipment damage due to induced high voltage, a new measurement method is proposed in this paper, which adopts frequency conversion technology, time coding technology, and contactless flexible sensor technology, which can accurately measure the action characteristics of high voltage switch equipment under the condition of double terminal grounding. This method can effectively reduce the influence of field measurement factors and improve the accuracy, safety, and convenience of field measurement of high voltage switch action characteristics.

The combination of nanoparticles and polycarboxylate ether superplasticizers is expected to simultaneously improve the durability and mechanical properties of mortar and concrete. However, the uniform dispersion of nanoparticles in cement materials has always been a difficult challenge. In this paper, nanoparticles are modified with surfactants to provide a research direction for better dispersion of nanoparticles in cementitious materials. Modified allyl ether polycarboxylate superplasticizer with short-branched chains was synthesized from acrylic acid and allyl glycidyl ether. Nano-silica/polycarboxylate ether was obtained by grafting nano-silica onto polycarboxylate ether molecular chains through intermolecular dehydration, thus achieving a homogeneous dispersion of nano-silica in the cementitious material. The results show that nano-silica/polycarboxylate ether has better dispersibility in cement matrix and strong clay resistance, which compensates for the poor dispersion of polycarboxylate ether synthesized from conventional polyethylene glycol graft chains in the presence of clay impurities.

Graph representation learning embeds graph nodes in a low-dimensional latent space, which allows for mathematical operations on nodes using low-dimensional vectors for downstream tasks, such as link prediction, node classification, and recommendation. Traditional graph embedding methods rely on hyper-parameters to capture the rich variation hidden in the structure of real-world graphs. In many applications, it may not be computationally feasible to search for optimal hyper-parameters. In this work, built on WatchYourStep which a graph embedding method leveraging graph attention, we propose a method that utilizes node-personalized context attention to capture the local variation in a graph structure. Specifically, we replace the shared context distribution among nodes with learnable personalized context distribution for each node. We evaluate our model on seven real-world graphs and show that our method outperforms the state-of-the-art baselines on both link prediction and node classification tasks. We further analyze the learned node context distribution to provide insights into its connection to graph structural properties.

A novel text data dimension reduction technique, called the tree-structured multi-linear principal component anal- ysis (TMPCA), is proposed in this work. Being different from traditional text dimension reduction methods that deal with the word-level representation, the TMPCA technique reduces the dimension of input sequences and sentences to simplify the following text classification tasks. It is shown mathematically and experimentally that the TMPCA tool demands much lower complexity (and, hence, less computing power) than the ordinary principal component analysis (PCA). Furthermore, it is demon- strated by experimental results that the support vector machine (SVM) method applied to the TMPCA-processed data achieves commensurable or better performance than the state-of-the-art recurrent neural network (RNN) approach.

A better placement of proppants has been always the goal pursued in sand fracturing in order to get longer effective fractures and higher flow conductivity. However, it is always difficult to achieve satisfactory effects by conventional processes. On the basis of theoretical analysis and simulation with FracproPT software, basic experiments, and innovative physical modeling experiment, a new impulse-stage fracturing process has been developed by combining a special pumping process with fiber, liquid and other auxiliary engineering means. Compared with conventional fracturing, the open seepage channel created by the new fracturing process has an obvious edge in effective fracture length and flow conductivity. Moreover, the open seepage channel can also improve fracture cleanliness and reduce pressure loss in artificial fractures, thus reaching the goal of prolonging the single-well production time and maximizing productivity. After the research on principles and optimal design of this new process, on-site pilot test and detailed post-fracturing evaluation were conducted. The results indicated that (1) the new process is highly operable and feasible; (2) compared with the adjacent wells with similar geological conditions, the proppant' cost is reduced by 44%–47%, the ratio of effective fracture length to propped fracture length is increased by about 16%, the fracturing fluid recovery rate is up to 63% after 18 h in the test, and the normalized production is 1.9–2.3 times that of the adjacent wells; and (3) the new process can significantly lower the cost and enhance production. The process has a broad application prospect in shallow-middle sand gas reservoirs and shale gas reservoirs in western Sichuan Basin.