Selection of stabilization methods for roads slopes in northern Taiwan

摘要

As part of the slope management system for the mountain roads in Taiwan, a database of road slopes protected with various slope stabilization methods has been established and the failure history of slopes is recorded as well. Since there are several methods can be chosen to protect a slope, a selection mechanism which can help the site engineers to choose the adequate stabilization methods is studied in this paper. Based on the failed and not failed cases collected in the database of the slope management system and ten influencing factors on the stability of road slopes, a selection method for the road slopes is proposed here using the back-propagation neural (BPN) network method. After trained with the slope cases collected, the BPN method can be used to select the most adequate slope stabilization method for a road slope and to indicate the method which is most likely to result slope failure also. For the selected slope stabilization method, the most sensitive factors which have significant influence on the slope stability can also be identified. By identifying the sensitive factors, it can provide valuable information for the routine maintenance of road slopes.