Application of Neural Network Model in Construction Quality Evaluation of Core Wall Rockfill Dam

Wang Qianwei,Zhong Denghua,Zhao Jianghao
DOI: https://doi.org/10.2991/iceep-16.2016.139
2016-01-01
Abstract:At present, high core rockfill dam mainly uses the real-time monitoring system to control the roller compacted construction quality, and the quality of roller compacted construction unit was evaluated by the test pit experiment. However, in the course of the experiment, the filling body of the experimental points need to be excavated, so that the filling body has a larger disturbance. This will affect the results of the experiment. At the same time, test pit experiments need to continue for a long time, however, only after the quality inspection results qualified, the follow-up construction can be carried out after, which affect the subsequent construction. In order to effectively solve the above problems, this paper established the quality evaluation index system and model of rockfill dam construction by using neural network method, and trained and tested the model with actual data. The test results show that the model has a good evaluation accuracy, and it is of great practical significance to improve the accuracy and timeliness of the core wall rockfill dam construction quality evaluation.
What problem does this paper attempt to address?