Search results

The current studies on remaining useful life (RUL) prediction mainly rely on convolutional neural networks (CNNs) and long short-term memories (LSTMs) and do not take full advantage of the attention mechanism, resulting in lack of prediction accuracy. To further improve the performance of the above models, this study aims to propose a novel end-to-end RUL prediction framework, called convolutional recurrent attention network (CRAN) to achieve high accuracy.

The proposed CRAN is a CNN-LSTM-based model that effectively combines the powerful feature extraction ability of CNN and sequential processing capability of LSTM. The channel attention mechanism, spatial attention mechanism and LSTM attention mechanism are incorporated in CRAN, assigning different attention coefficients to CNN and LSTM. First, features of the bearing vibration data are extracted from both time and frequency domain. Next, the training and testing set are constructed. Then, the CRAN is trained offline using the training set. Finally, online RUL estimation is performed by applying data from the testing set to the trained CRAN.

CNN-LSTM-based models have higher RUL prediction accuracy than CNN-based and LSTM-based models. Using a combination of max pooling and average pooling can reduce the loss of feature information, and in addition, the structure of the serial attention mechanism is superior to the parallel attention structure. Comparing the proposed CRAN with six different state-of-the-art methods, for the predicted results of two testing bearings, the proposed CRAN has an average reduction in the root mean square error of 57.07/80.25%, an average reduction in the mean absolute error of 62.27/85.87% and an average improvement in score of 12.65/6.57%.

This article provides a novel end-to-end rolling bearing RUL prediction framework, which can provide a reference for the formulation of bearing maintenance programs in the industry.

Grey modeling technique is an important element of grey system theory, and academic articles applied to agricultural science research have been published since 1985, proving the broad applicability and effectiveness of the technique from different aspects and providing a new means to solve agricultural science problems. The analysis of the connotation and trend of the application of grey modeling technique in agricultural science research contributes to the enrichment of grey technique and the development of agricultural science in multiple dimensions.

Based on the relevant literature selected from China National Knowledge Infrastructure, the Web of Science, SpiScholar and other databases in the past 37 years (1985–2021), this paper firstly applied the bibliometric method to quantitatively visualize and systematically analyze the trend of publication, productive author, productive institution, and highly cited literature. Then, the literature is combed by the application of different grey modeling techniques in agricultural science research, and the literature research progress is systematically analyzed.

The results show that grey model technology has broad prospects in the field of agricultural science research. Agricultural universities and research institutes are the main research forces in the application of grey model technology in agricultural science research, and have certain inheritance. The application of grey model technology in agricultural science research has wide applicability and precise practicability.

By analyzing and summarizing the application trend of grey model technology in agricultural science research, the research hotspot, research frontier and valuable research directions of grey model technology in agricultural science research can be more clearly grasped.

This paper aims to develop an efficient numerical method for mid-frequency analysis of built-up structures with large convex uncertainties.

Based on the Chebyshev polynomial approximation technique, a Chebyshev convex method (CCM) combined with the hybrid finite element/statistical energy analysis (FE-SEA) framework is proposed to fulfil the purpose. In CCM, the Chebyshev polynomials for approximating the response functions of built-up structures are constructed over the uncertain domain by using the marginal intervals of convex parameters; the bounds of the response functions are calculated by applying the convex Monte–Carlo simulation to the approximate functions. A relative improvement method is introduced to evaluate the truncated order of CCM.

CCM has an advantage in accuracy over CPM when the considered order is the same. Furthermore, it is readily to consider the CCM with the higher order terms of the Chebyshev polynomials for handling the larger convex parametric uncertainty, and the truncated order can be effectively evaluated by the relative improvement method.

The proposed CCM combined with FE-SEA is the first endeavor to efficiently handling large convex uncertainty in mid-frequency vibro-acoustic analysis of built-up structures. It also has the potential to serve as a powerful tool for other kinds of system analysis when large convex uncertainty is involved.

The purpose of this paper is to synthesise carboxymethylcellulose and methyl methacrylate graft copolymers (CMC‐g‐PMMA), which is used as an effective additive, for reinforcing the rice‐hull‐cement composite.

Various CMC‐g‐PMMA copolymers were synthesised at different reaction temperatures, pH values of reaction solution and the dosages of monomer and initiator (potassium persulphate). The copolymers were characterised by Fourier transforms infrared (FT‐IR) spectra, thermal analysis (thermogravimetric and differential scanning calorimeter), X‐ray diffractometry (XRD) and scanning electron microscopy.

An optimal CMC‐g‐PMMA copolymer is obtained. For synthesis of the CMC‐g‐PMMA, the optimal reaction temperature is 80°C and pH value is 9. FT‐IR test of CMC‐g‐PMMA confirmed the existence of a chemical link between carboxymethylcellulose (CMC) and methyl methacrylate (MMA). The content of initiator has little effect on the reaction for synthesising the graft copolymer. Thermal analysis indicates the occurrence of graft reaction in CMC and MMA. XRD test proved that the chains of the graft copolymer can enlarge the proportion of the amorphous regions of CMC. Adding MMA has damage effect on the crystallisation.

Since the results of this paper are obtained from the laboratory experiments, further research should be conducted for evaluating the performances of this copolymer in practical application.

The mechanical test of the rice‐hull‐cement composite proved that CMC‐g‐PMMA is an effective additive for reinforcing the rice‐hull‐cement composite. The synthesis of CMC‐g‐PMMA provides a new alternative for modifying cellulose derivatives.

The CMC‐g‐PMMA obtained in this paper is a new kind of effective agent. It can reinforce the rice‐hull‐cement composite and expands the application of the composite in building industries.

The dynamic characteristics prediction and frequency-modulation of pipeline was an important work for the design of aircraft hydraulic structure.

A complex pipeline was deemed as a combination of several segments of straight-pipe-element (SPE). The 3D vibration equations of each SPE were established in their local coordinate system based on Timoshenko-beam model and Euler-beam model, respectively. The dynamic-stiffness-matrixes were deduced from the dispersion relation of these equations. According to the complex pipeline layout in the global coordinate system, a multi dynamic stiffness matrixes assembling (MDSMA) algorithm was carried out to establish the characteristic equations of the whole complex pipeline. The MDSMA solutions were verified to be consistent with experimental results.

The MDSMA method based on Timoshenko-Beam model was more suitable for the short span aviation pipeline and the vibration at high frequency stage (>350 Hz). The layout affected the pipeline's in-plane stiffness and out-plane stiffness, for the Z-shaped pipe, each order natural mode took place on the ZP and NP alternately. Reasonable designs of bending position and bending radius were effective means for complex pipeline frequency-modulation.

A new dynamic modeling method of aircraft complex pipeline was proposed to obtain the influence of pipeline layout parameters on dynamic characteristics.

This paper aims to introduce a novel structure for the physical internet (PI)–enabled sustainable supplier selection and inventory management problem under uncertain environments.

To address hybrid uncertainty both in the objective function and constraints, a novel interactive hybrid multi-objective optimization solution approach combining Me-based fuzzy possibilistic programming and interval programming approaches is tailored.

Various numerical experiments are introduced to validate the feasibility of the established model and the proposed solution method.

Due to its interconnectedness, the PI has the opportunity to support firms in addressing sustainability challenges and reducing initial impact. The sustainable supplier selection and inventory management have become critical operational challenges in PI-enabled supply chain problems. This is the first attempt on this issue, which uses the presented novel interactive possibilistic programming method.

This study aims to investigate the influence of soy protein isolate hydrolysates (SPIH) obtained using 4 h hydrolysis under 200 MPa on proximate composition, cooking loss, textural properties, color, water distribution, microstructure, thiobarbituric acid reactive substance (TBARS) value and carbonyl and sulfhydryl contents of emulsion sausages.

Sausages with SPIHs at four concentrations (0, 1.0, 2.0 and 3.0%) were prepared, and the sausage with 0.01% butylated hydroxyanisole (BHA) was used as a positive control. Some sausages were selected for the analyses of quality characteristics and microcosmic properties. Other sausages were stored under 4 °C for 0, 7, 14, 21 and 28 days to investigate the oxidative stability.

The addition of SPIHs at various levels (0–3.0%) or 0.01% BHA did not affect the proximate composition (protein, fat and ash) of emulsion sausages. The addition of 2.0% SPIH decreased cooking loss and increased moisture content, hardness, springiness, chewiness, resilience and L* value, compared to the sausages without SPIH and with 0.01% BHA (p < 0.05). Furthermore, low-field nuclear magnetic resonance results suggested that sausages with 2.0% SPIH had the shortest T2 relaxation time. In addition, 2.0% SPIH and 0.01% BHA could inhibit the oxidation of emulsion sausages when compared with the sample without SPIH (p < 0.05). Moreover, there were no differences between sausages with 2.0% SPIH and 0.01% BHA (p > 0.05).

These findings confirmed that the 2.0% SPIH obtained under 200 MPa can be used as a natural additive to improve quality properties and antioxidant potential of emulsion sausages during storage.

This study aims to introduce an innovative approach that uses a decoder with multiple layers to accurately identify Chinese nested entities across various nesting depths. To address potential human intervention, an advanced optimization algorithm is used to fine-tune the decoder based on the depth of nested entities present in the data set. With this approach, this study achieves remarkable performance in recognizing Chinese nested entities.

This study provides a framework for Chinese nested named entity recognition (NER) based on sequence labeling methods. Similar to existing approaches, the framework uses an advanced pre-training model as the backbone to extract semantic features from the text. Then a decoder comprising multiple conditional random field (CRF) algorithms is used to learn the associations between granularity labels. To minimize the need for manual intervention, the Jaya algorithm is used to optimize the number of CRF layers. Experimental results validate the effectiveness of the proposed approach, demonstrating its superior performance on both Chinese nested NER and flat NER tasks.

The experimental findings illustrate that the proposed methodology can achieve a remarkable 4.32% advancement in nested NER performance on the People’s Daily corpus compared to existing models.

This study explores a Chinese NER methodology based on the sequence labeling ideology for recognizing sophisticated Chinese nested entities with remarkable accuracy.

The purpose of this paper is to design a new type of magnetic suction wall-climbing robot suitable for the wall inspection of wind turbine towers to solve the problems in manual maintenance tasks.

By analyzing the shortcomings of existing wall-climbing robots, a magnetic suction integrated wheel structure is designed to effectively combine the adsorption structure and transmission structure. To enable the robot to adapt to the curvature of the wall surface of a wind turbine tower, a passive adaptive curvature structure is designed. The effects of the air gap, the thickness of the wheel plates on both sides, the size of permanent magnets and the size of aluminum rings on the adsorption force are studied. Through mechanical model analysis under different instability conditions, the magnetic circuit of the magnetic wheel is optimized and designed.

Applying the wall-climbing robot to engineering practice, experiments have shown that the developed wall-climbing robot can move safely and stably on the wall of the wind turbine tower. The robot can also carry a load of 20 kg, and the designed adaptive structure can cause the magnetic wheel to deflect up to 20° relative to the vehicle body, fully meeting the curvature requirements of the minimum diameter end of the wind turbine tower.

This paper proposes a magnetic suction integrated wheel structure through analysis of the working environment. And the parameters affecting the magnetic wheel adsorption performance were optimized. Meanwhile, a passive adaptive wind turbine tower curvature structure was proposed.

The purpose of this paper is to describe how the hydraulic performance and pressure fluctuations in the entire flow passage of a Francis turbine were predicted numerically for the highest head. The calculations are used to partition the turbine operating regions and to clarify the unsteady flow behavior in the entire flow passage including the blade channel vortex in the runner and vortex rope in the draft tube.

Three‐dimensional unsteady numerical simulations were performed for a number of operating conditions at the highest head. The unsteady Reynolds‐averaged Navier‐Stokes equations with the k‐ω based SST turbulence model were solved to model the unsteady flow within the entire flow passage of a Francis turbine.

The predicted pressure fluctuations in the draft tube agree well with the experimental results at low heads. However the peak‐to‐peak amplitudes in the spiral case are not as well predicted so the calculation domain and the inlet boundary conditions need to be improved. The unsteady simulation results are better than the steady‐state results. At the most unstable operating condition of case a0.5h1.26, the pulse in the flow passage is due to the rotor‐stator interference between the runner and the guide vanes, the blade channel vortex in the runner blade passage and the vortex rope in the draft tube.

This study investigates the characteristics of the dominant unsteady flow frequencies in different parts of the turbine for various guide vane openings at the highest head. The unsteady flow patterns in the turbine, including the blade channel vortex in the runner and the helical vortex rope in the draft tube, are classified numerically, and the turbine operating regions are partitioned to identify safe operating regions.