Wines in the northwest region of China often suffer from poor color stability, low wine acidity, and inelegant aroma description. The Lachancea thermotolerans yeasts with a high yield of lactic acid and pleasant aroma compounds can ameliorate such wines. In this study, the performances of native L. thermotolerans CVE-LT1 co-inoculated with Saccharomyces cerevisiae in different ratios and strategies (simultaneous and sequential mode) were evaluated in pilot scale fermenter (60 L) and industrial scale fermenter (520 hL) of Cabernet Sauvignon grape must with high sugar/low acidity. Results evidenced that the mixed culture of L. thermotolerans and S. cerevisiae can significantly increase lactic acid level (up to 6.98 g/L) and decrease wine pH (up to 3.57). Meanwhile, the color parameters of a* and C* in mixed culture wines were improved. The higher concentrations of volatile compounds were mainly observed in sequential inoculation treatments, including the enhancement of ethyl lactate and 2-phenylethanol. As for phenolic compounds, L. thermotolerans promoted the formation of anthocyanins derivatives and phenolic acids in pilot and industrial fermentation. In conclusion, these results indicated that L. thermotolerans LT1 can be used in hot winery regions to improve the acidity, color indexes, phenolic compounds, and aroma quality of wines.
To study the influence of external loading on the magnetic signal at the defect location of the component, Q345R steel was selected to make a groove defect specimen for axial tensile testing. The magnetic behavior of the specimens containing different initial magnetization strengths under different stress states (elastic deformation stage, yielding stage, strengthening stage, and fracture stage) was studied in comparison. The result showed that the value of the tangential component Hxave in the nondestructive region gradually decreased with the increase of the stress loading in the elastic loading stage, however, the tangential component Hxave at the defect location of the specimens had different variation patterns due to the different initial magnetization strengths. When the load was loaded to the yielding stage, the magnetic signals remained stable, and Hxave increased again in the strengthening stage and completed the excitation of the magnetic signals in the fracture stage. By analyzing the magnetic memory field parameters throughout the loading process, it can be found that the variation patterns were consistent despite the different initial magnetization intensities of the specimens. The above findings are of great significance for assessing the stress state of defects in ferromagnetic materials.
Recent trend on transferable adversarial attacks focuses on applying different transformations on input images to improve input diversity. DEM, the recently proposed state-of-the-art method, adopts a diversity-ensemble method that contains multiple resize-padding-resize transformation branches to improve transferability. Despite its impressive attacking performance, we observe that the input diversity of DEM is still limited. In light of this, we propose an Advanced Diversity-Ensemble Method (ADEM) to further improve the transferability of adversarial examples. Specifically, we enlarge the range of image proportion and eliminate the overlap between different diversity scales existing in DEM to further improve the input diversity. Extensive experiments demonstrate the effectiveness of the proposed method.
To study the influence law and mechanism of the geomagnetic field on magnetic signals in the process of magnetic memory testing, the linear scanning detection results of defective specimens with three different initial magnetization strengths (0 A/m, -70 A/m, 105 A/m) along the horizontal and vertical directions in the geomagnetic field ambient are analyzed and discussed. The result shows that the magnetic signals in the defect-free region only change in amplitude with the change of detection angle, and have no effect on the overall shape of the magnetic signal curve; the characteristic parameters of the peaks of the magnetic signals in the defective region show a corresponding relationship with the measurement angle in the horizontal detection, and the initial magnetization strength of the specimen itself affects the change rule of the defective characteristic peaks. The mechanism analysis reveals that the extracted magnetic signals are affected by the combination of the geomagnetic field, the induced magnetic field of the specimen itself, and the leakage magnetic field. Finally, aiming at the problem that the geomagnetic field will change the topography of the defect peaks along different detection orientations, a geomagnetic compensation model is proposed for the cylinder equipment when performing circumferential scanning detection, which can reduce the interference of the geomagnetic field aberration on the detection results, to prevent the misjudgment and omission of the specimen's defect location.