| 摘要: |
| 为研究区域小气候对小型水库大坝渗流压力值的影响，建立了区域小气候观测系统和大坝渗流压力观测系统，并在水库上进行测试，实现了区域土壤水分、气压、气温、相对湿度等气象数据和水库渗流压力数据的自动采集。根据月累计降雨量的大小，分别选取干旱和多雨月份的监测数据，建立多气象因子偏最小二乘回归（PLSR）模型，开展影响渗流压力值的变量投影重要性（VIP）分析，在干旱和多雨两种条件下，量化区域小气候各气象因子对渗流压力值的贡献率。结果表明，干旱条件下，气压的VIP值＞1，是影响渗流压力值变化的主要原因，多雨条件下，气压、土壤水分含量VIP值＞1，是影响渗流压力值变化的主要原因。研究成果可对科学修正小型水库大坝渗流压力值提供理论和数据支撑。 |
| 关键词: 小气候 渗流压力 偏最小二乘回归 变量投影重要性 大坝安全监测 |
| DOI： |
| 分类号:TP274.2 |
| 基金项目:国家重点研发计划项目(2022YFC3005502)；河南省重点研发专项(241111220900)。 |
|
| Research on the Influence of Regional Microclimate on Seepage Pressure Values of Small Reservoirs |
|
TIAN Dongzhe1, WU Su1,2,3,2, JI Enyue4, YU Guohe3, WANG Zhongjin1
|
|
1.Henan Zhongyuan Photoelectric Measurement and Control Technology Co,Ltd;2.China;3.Cetc New Defense Technology Co,LTD;4.Nanjing Hydraulic Research Institute
|
| Abstract: |
| With the widespread construction of safety monitoring facilities for small reservoir dams, how to utilize the data value of these facilities and scientifically assess the service performance of the dams has become an important research content. The seepage pressure of the dam, as an important physical quantity for evaluating the seepage characteristics of earth-rock dams, is a necessary monitoring element for the construction of new safety monitoring facilities for small reservoirs. Accurately analyzing the sources of seepage pressure errors and improving the accuracy of data observation are the basis for achieving the scientific application of monitoring data and supporting the safety evaluation of the dam. Previous analyses of the sources of seepage pressure errors for dams included precipitation, water level in front of the dam, air pressure, sensor calibration parameters, and engineering construction. However, as a small-scale climate carrier, the reservoir is also affected by various meteorological factors within the region. The factors such as temperature, relative humidity, wind direction, wind speed, radiation, rainfall, and soil moisture form regional microclimate, which is closely related to the changes in the seepage pressure of the dam. There are relatively few research materials on the changes in seepage pressure values caused by multiple meteorological factors. In order to study the influence of regional microclimate on the seepage pressure value of small reservoir dams, a regional microclimate observation system and a dam seepage pressure observation system were established, and tests were conducted on the reservoir, achieving the automatic collection of meteorological data such as soil moisture, air pressure, temperature, and relative humidity, and the data of reservoir seepage pressure. Based on the monthly cumulative rainfall, data with a monthly cumulative rainfall of less than 30mm were selected as samples under drought conditions, and data with a monthly cumulative rainfall of more than 100mm were selected as samples under heavy rain conditions. A multiple meteorological factor partial least squares regression (PLSR) model was established, and the optimal number of components for the regression model was determined through 5-fold cross-validation. The reliability of the model was evaluated using the correlation coefficient R2 and the root mean square error RMSE. Through variable projection importance (VIP) analysis, the score values of each meteorological factor were calculated, quantifying the contribution rate of each meteorological factor to the seepage pressure value. Based on the measured seepage monitoring data from two monitoring points UP5 and UP6 on the maximum cross-section of a certain reservoir dam, the applicability of this method was verified. The results show that this method is relatively reliable. Under drought conditions, the R2 and RMSE of UP5 and UP6 are 0.873 and 0.942, and 0.3536 and 0.2399, respectively. By calculating the VIP score, the VIP score of air pressure is greater than 1, which is the main reason affecting the change in seepage pressure value. Under heavy rain conditions, the R2 and RMSE of P5 and UP6 are 0.832 and 0.418, and 0.4044 and 0.7524, respectively. By calculating the VIP score, the VIP scores of air pressure and soil moisture are greater than 1, which are the main reasons affecting the change in seepage pressure value. By calculating the standardized regression coefficient of soil moisture under heavy rain conditions and combining the spatial and temporal distribution of the monitoring data, it was found that soil moisture has a lag effect on the change in seepage pressure value. Compared with the existing research results, this study supplements the influence of meteorological factors on the variation of seepage pressure values. The research results can provide theoretical and data support for scientifically correcting the seepage pressure values of small reservoir dams. |
| Key words: Microclimate,Seepage pressure,Partial least squares regression,Variable Importance for Projection,Dam safety monitoring. |
