Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Research and Application of an Intelligent Prediction of Rock Bursts Based on a Bayes-Optimized Convolutional Neural Network
https://orcid.org/0000-0002-6961-7258
https://orcid.org/0000-0002-7930-3789
Intelligent prediction of rock bursts has great significance in rock mechanics research and a high value in engineering applications. An intelligent rockburst prediction method based on a Bayes-optimized convolutional neural network (BOCNN) was proposed. First, an exploratory analysis of data was conducted using joint distribution diagrams and the heat map of the correlation matrix to establish a high-quality data set of rockburst engineering cases and a parameter system for rockburst prediction. Second, six rockburst prediction models were built by combining machine learning algorithms, such as random forest, k-nearest neighbor (KNN), and Bayes, deep learning (CNN1d and CNN2d), and BOCNN. In addition, we used accuracy, precision, recall, F1 score, receiver operating characteristic curve, Taylor diagram, and the probability indicator of prediction results as indicators to evaluate the accuracy of the models. A comparative analysis of the six rockburst prediction models was conducted to explore rockburst prediction models with good robustness, generalization performance, and high accuracy. Moreover, 11 comparative models were established for comparative analysis. Then, we used the MATLAB tool to build an intelligent rockburst prediction system and applied the findings to the Jiangbian Hydropower Station in Sichuan Province, China. The results of the study show that the intelligent rockburst prediction system can provide technical support for predicting rockburst hazards in mining, transportation, and water conservancy and hydropower projects and a scientific basis for later construction and the design of support structures.
Research and Application of an Intelligent Prediction of Rock Bursts Based on a Bayes-Optimized Convolutional Neural Network
https://orcid.org/0000-0002-6961-7258
https://orcid.org/0000-0002-7930-3789
Intelligent prediction of rock bursts has great significance in rock mechanics research and a high value in engineering applications. An intelligent rockburst prediction method based on a Bayes-optimized convolutional neural network (BOCNN) was proposed. First, an exploratory analysis of data was conducted using joint distribution diagrams and the heat map of the correlation matrix to establish a high-quality data set of rockburst engineering cases and a parameter system for rockburst prediction. Second, six rockburst prediction models were built by combining machine learning algorithms, such as random forest, k-nearest neighbor (KNN), and Bayes, deep learning (CNN1d and CNN2d), and BOCNN. In addition, we used accuracy, precision, recall, F1 score, receiver operating characteristic curve, Taylor diagram, and the probability indicator of prediction results as indicators to evaluate the accuracy of the models. A comparative analysis of the six rockburst prediction models was conducted to explore rockburst prediction models with good robustness, generalization performance, and high accuracy. Moreover, 11 comparative models were established for comparative analysis. Then, we used the MATLAB tool to build an intelligent rockburst prediction system and applied the findings to the Jiangbian Hydropower Station in Sichuan Province, China. The results of the study show that the intelligent rockburst prediction system can provide technical support for predicting rockburst hazards in mining, transportation, and water conservancy and hydropower projects and a scientific basis for later construction and the design of support structures.
Research and Application of an Intelligent Prediction of Rock Bursts Based on a Bayes-Optimized Convolutional Neural Network
https://orcid.org/0000-0002-6961-7258
https://orcid.org/0000-0002-7930-3789
Intelligent prediction of rock bursts has great significance in rock mechanics research and a high value in engineering applications. An intelligent rockburst prediction method based on a Bayes-optimized convolutional neural network (BOCNN) was proposed. First, an exploratory analysis of data was conducted using joint distribution diagrams and the heat map of the correlation matrix to establish a high-quality data set of rockburst engineering cases and a parameter system for rockburst prediction. Second, six rockburst prediction models were built by combining machine learning algorithms, such as random forest, k-nearest neighbor (KNN), and Bayes, deep learning (CNN1d and CNN2d), and BOCNN. In addition, we used accuracy, precision, recall, F1 score, receiver operating characteristic curve, Taylor diagram, and the probability indicator of prediction results as indicators to evaluate the accuracy of the models. A comparative analysis of the six rockburst prediction models was conducted to explore rockburst prediction models with good robustness, generalization performance, and high accuracy. Moreover, 11 comparative models were established for comparative analysis. Then, we used the MATLAB tool to build an intelligent rockburst prediction system and applied the findings to the Jiangbian Hydropower Station in Sichuan Province, China. The results of the study show that the intelligent rockburst prediction system can provide technical support for predicting rockburst hazards in mining, transportation, and water conservancy and hydropower projects and a scientific basis for later construction and the design of support structures.
Research and Application of an Intelligent Prediction of Rock Bursts Based on a Bayes-Optimized Convolutional Neural Network
Int. J. Geomech.
Li, Mingliang (Autor:in) / Li, Kegang (Autor:in) / Qin, Qingci (Autor:in) / Yue, Rui (Autor:in) / Shi, Ji (Autor:in)
01.05.2023
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Optimized Deep Convolutional Neural Network for the Prediction of Breast Cancer Recurrence
| DOAJ | 2023
| Engineering Index Backfile | 1928