It is of great significance to explore the evolution law of shrinkage cracks in soil sample structure for the rational protection of earthen ruins. In order to study the dry shrinkage deformation characteristics of Zhouqiao earthen soil, an indoor dry shrinkage test was carried out. In order to further reveal the internal mechanism of soil shrinkage cracking and surface tension change, the surface tension gradient parameter is introduced. Based on SEM images, the microscopic model of Zhouqiao earthen soil was established in ABAQUS software, and the fracture evolution was simulated. The results show that the fracture evolution process of soil sample can be summarized as: tensile stress field, single main fracture, secondary horizontal and tertiary deep fractures, and the formation of fracture network. The duration of tensile stress field and single main fracture is long, while the duration of secondary horizontal and tertiary deep fractures is short. The mechanism of crack types can be summarized as follows: local shear, local tension and local mixed tension shear. With the development of crack, it gradually develops from tensile force to mixed tensile shear force, and the main crack extends outward perpendicular to the stress surface. Due to different length of water molecule migration path, the upper soil sample is more likely to produce large shrinkage deformation than the deeper soil sample, and causes the development of deep cracks. With the stability of water migration path, horizontal cracks become the main factor for the development of soil cracks. By analyzing the rose chart of fracture rate, it is found that the fracture rate of the soil sample in the upper part is generally lower than that in the lower part, and the fractures are mostly concentrated in the direction of 90° and 270°, indicating that the micro capillary pores and excessive internal shrinkage of the soil sample are the reasons for the cracks. The normalization results show that the model can better reflect the generation, evolution and development of soil micro damage in Zhouqiao site. The research results have reference value for the repair and protection of soil fissures in Zhouqiao site.

Recommender systems can effectively alleviate the problem of information overload caused by the rapid development of the Internet. However, the occurrence of shilling attacks restricts the healthy development of recommender systems. Therefore, how to detect shilling attacks accurately and efficiently is an important problem in the field of recommender systems security. The existing detection methods usually design hand-crafted detection features based on expert knowledge or automatically learn features from a single perspective using deep learning, then the attack users are identified according to the extracted features by hard classification, resulting in the poor detection performance. By automatically learning features from multiple perspectives and introducing a hesitant fuzzy decision, a novel detection method based on CNN and hesitant fuzzy set was proposed and named CNN-HFS. Firstly, for each user, three behavior matrices were extracted from the perspectives of rating, preference and rating time, respectively. To reduce the influence of data sparse, these matrices were scaled by bicubic interpolation to correspondingly obtain a dense rating matrix, a dense preference matrix and a dense time matrix. Next, each scaling matrix of users was regarded as an image, and three different CNN classifiers were trained based on these scaling matrices in three different views respectively. For each user, three membership degrees to the classifier of attack users were calculated. Finally, a fuzzy hesitant set was introduced to make a comprehensive decision, and the attack users were identified according to the decision results. To validate the effectiveness of the proposed CNN-HFS, the extensive experiments were conducted on the MovieLens 1M and Amazon datasets. The evaluation metrics of precision, recall and F1-measure were used to compare the proposed method with SVM-TIA, CoDetector, CNN-SAD, SDAEs-PCA, CNN-R, CNN-P and CNN-T. The experimental results showed that the proposed method is superior to seven baseline methods in terms of three detection metrics and achieves an excellent detection performance under various attacks.

For the landslide susceptibility prediction (LSP) based on machine learning (ML) models, the reasonable selection of negative samples has an important influence on the LSP performance. Generally, the main selection methods include randomly selecting from the whole study area or from the specific attribute areas such as low slopes. The negative samples selected by the above methods are often inaccurate or biased, resulting in low accuracy and low reliability of LSP. To solve this problem, the coupling model of ML and information value (IV) method was proposed for LSP. Taking Ruijin City as the study area, the attribute values of the environmental factors were transformed into the IV values of the contribution to the landslide to obtain the very low and low susceptibility areas. The negative samples were randomly selected in the above areas for the training and validation of machine learning models. The new coupling models of IV–SVM and IV–RF were constructed for the LSP of Ruijin. Further, IV–SVM and IV–RF models were compared with the single SVM and RF model with negative samples randomly selected from the whole study area, as well as the low-slope SVM and RF model with negative samples randomly selected from specific attribute areas with a slope less than 2°. Finally, Kappa coefficient (KC) and receiver operating characteristic (ROC) curve were used to verify and compare the modeling results. The AUC values of the ROC curve and KC of IV–SVM and IV–RF models were 0.828, 0.920 and 0.876, 0.988, which were higher than those of single SVM, RF model and low-slope SVM, RF model, respectively. Meanwhile, IV–SVM and IV–RF models have a smaller mean value and larger standard deviation of a susceptibility probability distribution. Results showed that: 1) IV–SVM and IV–RF models had the higher LSP accuracies than those of the single SVM, RF model and low-slope SVM, RF model, respectively; 2) RF model had higher LSP accuracy compared to the SVM model; 3) The coupling model such as IV–RF could address the inaccuracy of negative sample sampling existing in the single model and the shortcomings of the low slope model in the selection of slope interval, thus improving the LSP accuracy. In conclusion, this study provided a new idea for the negative sample sampling method for LSP using ML models

To verify the applicability of wave theory in studying the free vibration characteristics and impact response of the cable–beam structure, and to preliminarily explore the behaviors of elastic wave propagation through the structure under moving load, the dynamic response function was derived from the transverse vibration differential equation of Timoshenko beam and the longitudinal wave equation of cables. The reverberation-ray matrix was used to obtain the waveform solution of the structural response. Based on the idea of discrete Fourier transform (DFT), the series solution of structural transient response was derived to solve the inverse problem of the traditional reverberation-ray matrix (MRRM). The improved MRRM was verified by experiment and finite element method (FEM). The results showed that at the speed of 30 km/h, the deviation between the theoretical maximum strain and FEM results was 5% and that between theoretical and experimental results was 8%. When the vehicle speed was 40 km/h, the deviation between the theoretical maximum strain and FEM results was 4.0%, which was 9.8% between theoretical and experimental results. Taking the beam-cable system as the research object, the maximum deviation between the first five natural frequencies calculated by improved MRRM and FEM was 0.29%, and the deviation between the first two natural frequencies was 0. The wave response characteristics of the cable-stayed beam under moving load were analyzed and the theoretical results were in good agreement with the FEM results. It could be inferred that the improved MRRM had high reliability in calculating the transient wave response of bridge structure under moving load. Based on the analyses of the frequency domain response, it was found that the flexural waves in the Timoshenko beam under moving load were mainly low-frequency responses whose frequencies were lower than 2 times the fundamental frequency of the structure. Furthermore, the reasonable selection criteria of frequency range were explored in the process of finding the wave response, to further improve the calculation efficiency of MRRM

To scientifically and reasonably analyze the bending vibration frequency of the new type composite box girder bridge with corrugated steel webs, taking into account the shear deformation effect of corrugated steel webs and the shear lag effect of box girder, the Galerkin method and the Hamilton principle were used to deduce the free vibration control differential equation and natural boundary conditions of the bridge type. According to the natural boundary conditions, the calculation formula of the bending vibration frequency of the new type composite box girder bridge with corrugated steel webs under the influence of shear deformation effect and shear lag effect was solved. The results were compared with the measured results and ANSYS finite element results, and the influencing factors of the bending vibration frequency were analyzed. The results showed that the calculation results of the bending vibration frequency were in good agreement with the measured results and the ANSYS finite element results, which verified the correctness of the derived frequency calculation formula. The bending vibration frequency increased with the increase of the high-span ratio. When it was less than 0.05, the increase of the bending vibration frequency was relatively gentle. When the height-span ratio was greater than 0.05, the increase of the bending vibration frequency was more significant. The bending vibration frequency increased with the increase of the width-span ratio, but the overall increase was not obvious. The frequency increased with the increase of the thickness of the corrugated steel web, and the higher the order, the more significant the increase. The shear lag effect of the box girder had little effect on the bending vibration frequency, and the maximum error of the first 5 orders was only 6.73%. The shear deformation of the corrugated steel webs had a great influence on the bending vibration frequency, and the maximum error of the first five orders was as high as 51.18%