Autoencoder (AE) is a deep neural network that has been widely utilized in process industry owing to its superior abilities of feature extraction and data reconstruction. Recently, assuming the latent variables be random variables, probabilistic variant it called variational autoencoder (VAE) achieved major success different applications. In this article, we develop two novel submodels based on VAEs (DVAE), which are further establish soft sensor framework. By use our first submodel known as...

Soft capacitive pressure sensors with high performance are becoming increasingly in demand the emerging flexible wearable field. While fiber have achieved sensitivity, their sensitivity range is limited to low-pressure levels. As typically require preloading and fixation, this narrow of poses a challenge for practical applications. To overcome limitation, study proposes resistive-capacitive hybrid response (HFPSs) three-layer core–sheath structures. The trigger enhancement mechanisms...

Data-driven soft sensors, estimating the pivotal quality variables, have been widely employed in industrial process. This paper proposes a novel sensor modeling approach based on two-stream λ gated recurrent unit (T S - GRU) network. First, factors <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sub> and xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> are implemented to alter linear constraint existing original GRU unit, enriching...

Bisphenol-A polycarbonate (BPA-PC) has been widely used, but it restricted from use in several food-contact products due to the estrogen-like and antiandrogen effects of BPA. With aim developing novel, biobased, ecologically safe high-performance polycarbonates with potential replace BPA-PC, this work, a series novel biobased copolycarbonates, viz., poly(isosorbide carbonate-co-isoidide-2,5-dimethylene carbonate) (PIsIeC), were designed successfully synthesized via melt polymerization. The...

Fiber masterbatch production suffers from inherent agglomeration effect of high-concentration color masterbatches, negatively impacting uniformity, particle dispersion, and thermal stability in fiber masterbatch, ultimately the quality products. However, accurately recognizing classifying scanning electron microscopy (SEM) images remain a challenge due to complexity status, difficulty distinguishing micro-size agglomeration, limited data availability. To address this challenge, paper...

Capacitive pressure sensors play an important role in the field of flexible electronics. Despite significant advances two-dimensional (2D) soft sensors, one-dimensional (1D) fiber electronics are still struggling. Due to differences structure, theoretical research 2D has difficulty guiding design 1D sensors. The multiple response factors and capacitive mechanism have not been explored. Fiber urgently need a tailor-made development path. In this regard, we established pressure-sensing...

Fine denier polyester fibers have attracted attention in electronic information industry owing to their unique structural properties and ultra fine diameters. Since there is a close connection between melt-spinning process parameters diameters, the influence of on fiber diameters an issue worth further exploration. This study aims construct prediction models based Gaussian regression (GPR), radial basis function neural network (RBFNN) extreme learning machine (ELM) algorithms predict...

The uneven distribution of process industrial data poses a significant challenge for soft sensor modeling. Hence, it is necessary to employ generative models generate some new used augmenting the fitting ability model by full utilization sparse regions. Existing are mostly applied in image and text generation fields, which more suitable discrete where each variable only consists integers. To enhance applicability modeling, novel vector-quantized weighted-Wasserstein variational autoencoder...

In recent years, more and attention has been attracted by fine denier polyester fibers in electronic information industry due to their unique structural excellent mechanical properties. As known, the properties of have a close relationship with melt-spinning process parameters, thus, it is worth further exploring influence parameters on fiber (namely breaking strength, elongation at break CV strength). This study aims develop novel prediction model based artificial neural network (ANN)...

In this paper, a prediction method of the melt spinning properties based on Gated Recurrence Units (GRU) neural network model is proposed considering time series property data. Spinning process plays key role in production process, product called primary fiber. Primary fiber has three major indicators: breaking strength, elongation at break and 1.5 times elongation, its quality directly affects subsequent processing The algorithm experimentally data break. This experiment result demonstrates...

Abstract This paper considers a model of melt spinning with stress-induced crystallization, where the uses Phan Thien-Tanner model, and solid adopts rubber elastic model. We design computationally efficient algorithm for solving The temperature, radius, birefringence profile in low-speed high-speed spun PET fibers were predicted compared published experimental data. simulation results are consistent data from literature. Then parametric analysis is conducted to investigate effects operating...

Abstract Fiber crossbars, an emerging electronic device, have become the most promising basic unit for advanced smart textiles. The demand highly sensitive fiber crossbar sensors (FCSs) in wearable electronics is increased. However, unique structure of FCSs presents challenges replicating existing sensitivity enhancement strategies. Aiming at sensors, a second‐order synergistic strategy proposed that combines air capacitance and equipotential bodies, resulting remarkable over 20 times FCSs....

Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI

In polyester fiber production process, oligomer density is the most important indicator which reflects quality of and stability polymerization process. Accurate prediction essential for monitoring subsequent series chemical reactions. However, affected by numerous process factors. The use raw high-dimensional data will leads to high model complexity training difficulties. Therefore, this paper proposes a novel adaptive long-short term memory (LSTM) based on attention mechanism solve above...

An energy storage thermo-battery from a thermocell was obtained by utilizing the thermo-responsive hydrophobic interaction between methyl cellulose and I 3 − (Δ C ), thermoelectric properties were further enhanced confinement of BC S ).

The demand for information exchange between humans and machines or virtual spaces is growing rapidly. However, traditional contact sensing makes the spread of bacteria viruses, single also limits development. Hence, as an innovative product, a non-contact/contact dual-mode sensor has to come into being, which consists stretchable bacterial cellulose (BC)-BC/graphene (Gr) helical fibers. fiber with sheath-core structure, degradable BC sheath BC/Gr core, exhibit tensile strength 113 MPa...