In this study, an intelligent monitoring platform is established for continuous quantification of soil,vegetation, and atmosphere parameters (e.g. soil suction, rainfall, tree canopy, air temperature, and windspeed) t...In this study, an intelligent monitoring platform is established for continuous quantification of soil,vegetation, and atmosphere parameters (e.g. soil suction, rainfall, tree canopy, air temperature, and windspeed) to provide an efficient dataset for modeling suction response through machine learning. Twocharacteristic parameters representing suction response during wetting processes, i.e. response time andmean reduction rate of suction, are formulated through multi-gene genetic programming (MGGP) usingeight selected influential parameters including depth, initial soil suction, vegetation- and atmosphererelated parameters. An error standardebased performance evaluation indicated that MGGP has appreciable potential for model development when working with even fewer than 100 data. Global sensitivityanalysis revealed the importance of tree canopy and mean wind speed to estimation of response timeand indicated that initial soil suction and rainfall amount have an important effect on the estimatedsuction reduction rate during a wetting process. Uncertainty assessment indicated that the two MGGPmodels describing suction response after rainfall are reliable and robust under uncertain conditions. Indepth analysis of spatial variations in suction response validated the robustness of two obtained MGGPmodels in prediction of suction variation characteristics under natural conditions.展开更多
基金the financial support funded by the Science and Technology Development Fund of Macao SAR (Grant Nos. 0026/2020/AFJ and SKL-IOTSC(UM)-2021-2023)the Funds for University of Macao (Grant No. MYRG2018-00173-FST)
文摘In this study, an intelligent monitoring platform is established for continuous quantification of soil,vegetation, and atmosphere parameters (e.g. soil suction, rainfall, tree canopy, air temperature, and windspeed) to provide an efficient dataset for modeling suction response through machine learning. Twocharacteristic parameters representing suction response during wetting processes, i.e. response time andmean reduction rate of suction, are formulated through multi-gene genetic programming (MGGP) usingeight selected influential parameters including depth, initial soil suction, vegetation- and atmosphererelated parameters. An error standardebased performance evaluation indicated that MGGP has appreciable potential for model development when working with even fewer than 100 data. Global sensitivityanalysis revealed the importance of tree canopy and mean wind speed to estimation of response timeand indicated that initial soil suction and rainfall amount have an important effect on the estimatedsuction reduction rate during a wetting process. Uncertainty assessment indicated that the two MGGPmodels describing suction response after rainfall are reliable and robust under uncertain conditions. Indepth analysis of spatial variations in suction response validated the robustness of two obtained MGGPmodels in prediction of suction variation characteristics under natural conditions.
