Search

The crossbeam is frequently subjected to alternating loads during work as an essential load-bearing part of the crane. However, due to the large volume and the limitations of detection technology, it is impossible to realize online monitoring of the mechanical state. The ongoing advancement of ROMing and digital twin technology plays a pivotal role in facilitating the resolution of this particular issue. In this paper, we take the crane beam as the physical entity and combine the Twin Builder reduced-order technology and Deployer digital twin deployment technology to establish a digital twin of the beam. The load recognition model within the twin system exhibits a prediction error rate of ±5%. Furthermore, the accuracy of the ROM surpasses that of conventional machine learning models by a factor of 25. Upon deployment on the web platform, the results are delivered within 0.5 s, representing a substantial improvement as it is merely 1/15 of the time required for traditional 3D displays. The digital twin online monitoring system has the advantages of high accuracy and low requirements for monitoring equipment, which can be widely used in engineering practice to solve the problem that the mechanical state of large parts cannot be accurately monitored online.

To study the deformation law and forming defects of hemispherical molding of unidirectional carbon fiber reinforced polymer (UN-CFRP), hemispherical molding experiments were carried out, and a three-dimensional finite element model of single-ply UN-CFRP hemispherical molding was established. The results show that the established finite element model can accurately predict the molding deformation law and forming defects of UN-CFRP. The results indicate that the fiber at the top of the sample is uniformly distributed; the fiber on both sides of the sample top and parallel to the fiber direction is local dispersion; the fiber at both sides of the bottom is trapezoidal splitting. The results at the bending corner show that the bending quality is better when the angle between the fiber direction and the tangent of the circular bend is in the range of 70-90°; it is easy to produce extrusion wrinkling when the angle between the fiber direction and the tangent of the circular bend is 15-70°; it is easy to produce extrusion accumulation when the angle between the fiber direction and the tangent at the circular bending is in the range of 0-15°. The results of cross-ply UN-CFRP hemispherical molding show that the full spreading property is good, and there is no obvious fold in the bending part.

In this study, an application of a voltammetric electronic tongue for discrimination and prediction of different varieties of rice was investigated. Different pretreatment methods were selected, which were subsequently used for the discrimination of different varieties of rice and prediction of unknown rice samples. To this aim, a voltammetric array of sensors based on metallic electrodes was used as the sensing part. The different samples were analyzed by cyclic voltammetry with two sample-pretreatment methods. Discriminant Factorial Analysis was used to visualize the different categories of rice samples; however, radial basis function (RBF) artificial neural network with leave-one-out cross-validation method was employed for prediction modeling. The collected signal data were first compressed employing fast Fourier transform (FFT) and then significant features were extracted from the voltammetric signals. The experimental results indicated that the sample solutions obtained by the non-crushed pretreatment method could efficiently meet the effect of discrimination and recognition. The satisfactory prediction results of voltammetric electronic tongue based on RBF artificial neural network were obtained with less than five-fold dilution of the sample solution. The main objective of this study was to develop primary research on the application of an electronic tongue system for the discrimination and prediction of solid foods and provide an objective assessment tool for the food industry.

Peanut meal is the byproduct of high-temperature peanut oil extraction; it is mainly composed of proteins, which have complex tastes after enzymatic hydrolysis to free amino acids and small peptides. The enzymatic hydrolysis method was adopted by using two compound proteases of trypsin and flavorzyme to hydrolyze peanut meal aiming to provide a flavor base. Hence, it is necessary to assess the taste attributes and assign definite taste scores of peanut meal double enzymatic hydrolysis hydrolysates (DEH). Conventionally, sensory analysis is used to assess taste intensity in DEH. However, it has disadvantages because it is expensive and laborious. Hence, in this study, both taste attributes and taste scores of peanut meal DEH were evaluated using an electronic tongue. In this regard, the response characteristics of the electronic tongue to the DEH samples and standard five taste samples were researched to qualitatively assess the taste attributes using PCA and DFA. PLS and RBF neural network (RBFNN) quantitative prediction models were employed to compare predictive abilities and to correlate results obtained from the electronic tongue and sensory analysis, respectively. The results showed that all prediction models had good correlations between the predicted scores from electronic tongue and those obtained from sensory analysis. The PLS and RBFNN prediction models constructed using the voltage response values from the sensors exhibited higher correlation and prediction ability than that of principal components. As compared with the taste performance by PLS model, that of RBFNN models was better. This study exhibits potential advantages and a concise objective taste assessment tool using the electronic tongue in the assessment of DEH taste attributes in the food industry.