Search

Inverse problems are typically tackled using deterministic optimization methods that may become trapped in a local minimum or probabilistic methods that can be computationally demanding. In this study, we explore the potential of the back propagation neural network (BPNN) optimized by the genetic algorithm (GA) for onshore transient electromagnetic (TEM) inversion. The GA is employed to optimize the initial parameters of the BPNN, enhancing its global optimization ability. Once the BPNN optimized by GA (GA-BPNN) is properly trained, it can provide the distribution of subsurface electrical conductivity (σ) in 0.1 s. We train the GA-BPNN using synthetic datasets generated by TEM forward modeling and assess its reliability using both synthetic and field data. Theoretical simulations demonstrate that compared with BPNN, the error of GA-BPNN on the inversion results of six samples is reduced by 23.2%. Furthermore, this method can provide reliable results even in the presence of noise in the TEM response. Finally, applying this inversion method to karst exploration with measured data proves the reliability and robustness of the proposed method. The proposed method can support quasi-real-time imaging of subsurface structures and provides a powerful tool for the interpretation of field TEM data. [ABSTRACT FROM AUTHOR]

The river levee settlement derived from shield tunnel excavation face instability is an anticipated safety issue. Serious engineering safety hazards may occur if it is not restricted. Thus, this study selected six influence indicators to quantify the stability level of the excavation face using the AHP-Fuzzy method. A database of the levee settlement prediction indicators containing tunneling speed, supporting pressure, grouting pressure, horizontal distance, lagging distance, and stability coefficient of excavation face was built by collecting 60 samples from a river-crossing shield tunnel project in China. Five models were carried out for predicting the levee settlement, i.e., modified Peck formula, RF, SVM, BPNN, and PSO-BPNN. Their prediction performance was verified by comparing the prediction results with the measured ones in a shield tunnel engineering application. The analytical results indicated that the PSO-BPNN model has the highest correlation coefficient, i.e., 0.8831(in training) and 0.8657(in testing) and the lowest error, i.e., RMSE = 1.901, and MAE = 0.8412, respectively. Thus the PSO-BPNN model has the best ability to capture the nonlinear relationship between the independent and dependent variables compared with the abilities of the RF, SVM, and BPNN model in this case study. The modified Peck formula performed unsatisfactory results (RMSE = 3.2588, MAE = 3.1172). The correlation analysis results of the input characteristic parameters of the model also indicate that the stability of the levee settlement is significantly related to the excavation face stability. The results indicate that the hybrid model is a very useful and powerful means to investigate the levee settlement caused by shield tunneling and can be improved by gathering more data from tunneling operations over time. [ABSTRACT FROM AUTHOR]

In geotechnical engineering practice, studying the characteristics of rock electrical resistivity under different physical and mechanical environments is important for improving the accuracy of unfavorable geological condition detection. In this study, Dali marble and granite were subjected to cyclic loading and unloading at different peak pressures, and the electrical resistivity of the core samples was continuously monitored. The aim was to derive the electrical resistivity variation patterns of two rocks under cyclic loading and unloading. Then, the theoretical models of electrical resistivity for the two rocks were established through theoretical derivation, and the model equations were consistent with the experimental findings. Finally, a MATLAB program was used for nonlinear fitting of the unloading phase electrical resistivity variation patterns of the rocks to reveal the quantified pressure-electrical resistivity patterns of the two rocks with different lithologies. Based on this, well-performing rock pressure-electrical resistivity relationship models were developed with three parameters: pressure, water content and electrical resistivity. The experimental results can significantly improve the accuracy of stratigraphic state inversion and unfavorable body boundary identification by providing realistic electrical resistivity constraints for electrical resistivity tomography. [ABSTRACT FROM AUTHOR]

Rockburst is one of the major engineering geological disasters of underground engineering. Accurate rockburst intensity level prediction is vital for disaster control during underground tunnel construction. In this work, a hybrid model integrating the back propagation neural network (BPNN) with beetle antennae search algorithm (BAS) has been developed for rockburst prediction. Before model building, 173 groups of rockburst dataset were collected. Six geological parameters are selected as predictors for rockburst, including the maximum tangential stress of the surrounding rock σθ, the uniaxial compressive strength of rock σc, the tensile strength of rock σt, the stress ratio σθ/σc, the rock brittleness ratio σc/σt, and the elastic energy index Wet. After preprocessed by outlier detection and synthetic minority oversampling technique (SMOTE), the new dataset was divided into training and test parts. BAS could optimize the weights and biases of BPNN from the training process. Then the established hybrid model was applied to the test samples with predicted accuracy of 94.3%, proving that the hybrid model has practical value in researching rockburst prediction. [ABSTRACT FROM AUTHOR]

The stability classification of loess deposits around tunnels is a vital prerequisite for safe construction in underground environment. Due to the fuzziness and randomness of loess physical and mechanical parameters, the stability prediction of loess deposits shows uncertainty. Existing loess deposit stability classification models rarely consider the uncertainty of influencing factors. A novel classification probability model of loess deposits is proposed for the above problems based on Monte Carlo simulation and multi-dimensional normal cloud (MCS-Cloud). Specifically, five loess parameters, including water content, cohesion, internal friction angle, elastic modulus, and Poisson ratio, were selected as predictors for the stability level of loess deposits. The weights of the predictors were obtained through 50 test samples. After acquiring the numerical characteristics of the normal cloud, the stability level can be comprehensively evaluated with the weighted multi-dimensional normal cloud model. The classification model was applied to the loess tunnel in Yan'an, China. The prediction results are in good agreement with practical engineering, denoting the rationality of the weighted multi-dimensional normal cloud. Finally, the stability classification of loess deposits was discussed from the perspective of uncertainty analysis with the application of MCS. Results proved that the MCS-Cloud model is feasible for classifying the stability of loess deposits surrounding tunnels. The obtained classification probability can be used for quantitative risk assessment of loess tunnels. [ABSTRACT FROM AUTHOR]

Engineering site selection is an essential and systematic work in the early engineering construction stage. At present, the subsea tunnel site selection mainly depends on manual experience. There is still a lack of subsea tunnel site selection systems based on environmental impact. This study develops a comprehensive site selection evaluation system based on the analytic hierarchy process (AHP) and fuzzy evaluation method for the subsea tunnel site selection. It is a multi-indicator mathematical model evaluation system. On this basis, the ecological site selection method of the subsea tunnel is further studied, an indicator system for evaluating the environmental carrying capacity of the island is established, and the site selection results of the subsea tunnel based on the environmental indicators are obtained. We compared the site selection results of the two methods. The results show that the conventional method and the ecological site selection method based on environmental indicators can well carry out the site selection of subsea tunnels. The two methods take into account both the overall and local optimum of the subsea tunnel route and organically combine the overall and local objectives. This way provides a reference for the design and construction of the subsea tunnel in the future and points out the direction for the site selection of other large-scale projects with significant environmental impact. [ABSTRACT FROM AUTHOR]

To achieve comprehensive analyses, the presentation of comprehensive geophysical results usually involves the use of separate imaging and the combination of various results. At present, few studies have considered the correlation degree and unified imaging of different types of geophysical data. We establish a set of data fusion imaging methods for multiple geophysical data based on their reflection coefficients. As geophysical exploration results are primarily provided through waveform and resistivity sections, waveform and resistivity data were selected for fusion and were converted into reflection coefficients, and ground-penetrating radar (GPR) and surface electrical resistivity tomography (ERT) were taken as examples. Re-sampling and feature reconstruction were performed to unify the data in space and resolution. Finally, principal component analysis was used to calculate the correlation of the reconstructed reflection coefficient and to perform data fusion; this led to unified imaging based on the reflection coefficient of the considered geophysical data. Numerical simulation analyses and field experiments proved the efficacy of this method for producing unified imaging of multiple geophysical data. In summary, we provide a novel method for the unified interpretation of multiple geophysical data and enhance the identification ability of geological interfaces and anomaly distribution. [ABSTRACT FROM AUTHOR]

Electrical resistivity tomography (ERT) is a widely used method in geophysical prospecting. In practical applications, the smoothness-constrained least squares method is usually used to invert the apparent data of ERT. Although the resistivity distribution in the detection area can be obtained according to the inversion result, it is difficult to resolve the specific boundary of anomalies directly, which will affect the accuracy of the interpretation results of the detection data. This paper proposes a high-resolution imaging and interpretation method for ERT based on reflection coefficient analysis. This method can amplify the data features at the boundary of anomalies by performing the secondary processing and analysis on inversion results, thereby identifying the specific boundary of anomalies. This research improves the accuracy of interpretation results and realizes high-resolution imaging. Compared with the multi-scale inversion and other methods that enhance the resolution by optimizing the inversion algorithm, this method can more intuitively determine the specific boundary of anomalies in the inversion cross-section. [ABSTRACT FROM AUTHOR]

An unbalanced rockburst dataset will restrict the accuracy and reliability of rockburst prediction based on machine learning. Therefore, a new oversampling algorithm was proposed based on Copula theory and Monte Carlo simulation to balance the dataset. This paper collected 243 rockburst cases worldwide. The predictors of rockburst used in this paper include the maximum tangential stress of the surrounding rock, the uniaxial compressive strength of rock, the tensile strength of rock, and the elastic energy index. During oversampling, Copula theory determined the predictors' joint distribution function by considering the correlation among predictors. Then Monte Carlo simulation was performed to generate enough data for oversampling according to rockburst classification standard. Six common machine learning methods with tenfold cross-validation were adopted to establish nonlinear models between predictors and rockburst grades. The accuracy rate of rockburst prediction has increased by 9.3–15.5% after oversampling. The prediction performance is better than the commonly used synthetic minority oversampling technique. Finally, rockburst predictors' importance was evaluated with the initial dataset, and the elastic energy index got the maximum value of 0.41. The proposed oversampling algorithm in this paper can reasonably overcome class imbalance and improve the prediction performance for rockburst. [ABSTRACT FROM AUTHOR]

As an eco-friendly management method, biological control of postharvest diseases, utilizing antagonistic yeasts, is a research topic receiving considerable attention. Detailed knowledge on the biology of yeast antagonists is crucial when considering their potential application and development as biocontrol products. Changes in the growth form, such as single-cell to pseudohyphae, have been associated with the mode of action in postharvest biocontrol yeasts. In this study, the antagonistic yeast, Candida diversa, reversibly shifted from a single-cell morphology on yeast peptone dextrose (YPD) medium with 2 % agar to a pseudohyphal morphology on YPD with 0.3 % agar. The tolerance of the pseudohyphal form to heat and oxidative stresses, as well as the biocontrol efficacy against Botrytis cinerea on apple and kiwifruit stored at 25 and 4 °C, was significantly higher as compared to the single-cell form. This study provides new information on the ability of C. diversa to change its morphology and the impact of the morphology shift on stress tolerance and biocontrol performance. [ABSTRACT FROM AUTHOR]