Neural Networks Approach for Predicting Slope Failure Areas Based on the Onsite Slope Survey Parameters at a Specified Mountain Road

Abstract

This study proposed a new method of using a neural network model for predicting slope failure areas by training different slope parameters obtained from 49 onsite slope failure surveys in a specified mountain road section. The developed model has three input parameters including slope height, slope horizontal distance, and rainfall factor. With three neurons in the hidden layer, a relatively better performance for predicting a slope failure area can be obtained in the output layer, based on the evaluation indices of a correlation coefficient and a mean square error. For all the data sets in the neural network calculation process, the square value of a correlation coefficient between the prediction result and the survey data reached up to 0.8020, which may imply that the developed model has an acceptable reliability. The current study has presented a simple way for dealing with this type of slope failure problem, and the obtained results may provide useful information for the relevant engineering agency to improve road safety in the mountain road region investigated.

Authors and Affiliations

Tienfuan Kerh, et al.

Keywords

Related Articles

Control of Microgrid: Literature Review

Microgrid (MG) is a one of the novel concept in power generation. The Microgrid concept assumes a cluster of loads and micro-sources operating as a single controllable system that provides both power and heat to its loca...

A Bayesian Approach to Process Model Evaluation in Short Run SPC

With the availability of Big Data in manufacturing, historical data to initially characterize a process is available in abundance. In fact, evaluating and selecting the best-fitted data set replaces data availability as...

Output and Power Characteristics of LLC Resonant DC-DC Converters for Photovoltaic Applications

The output and power characteristics of different LLC resonant DC-DC converter configurations are obtained, which will inform the designer of the resonant converters, what is the type of LLC configuration, respectively t...

Designing of Efficient High Voltage Three Phase Bipolar SPWM Inverter and Analysis

The SPWM inverters are widely used for energy conversion in renewable sources viz. solar power, Wind and bio¬¬ plants. The paper has evaluated the FFT performance of three different inverting architectures triggered usin...

Cyclic Bending Failure of Local Sharp-Notched Circular Tubes with Different Diameter-to-Thickness Ratios

In this study, the failure of local sharp-notched 6061-T6 aluminum alloy tubes with different diameter-to-thickness ratios submitted to cyclic bending are studied. Different diameter-to-thickness ratios of 16.5, 31.0 and...

Download PDF file
  • EP ID EP497958
  • DOI -
  • Views 88
  • Downloads 0

How To Cite

Tienfuan Kerh, et al. (2018). Neural Networks Approach for Predicting Slope Failure Areas Based on the Onsite Slope Survey Parameters at a Specified Mountain Road. International Journal of Engineering Innovations and Research, 7(3), 192-199. https://europub.co.uk/articles/-A-497958