Research On Safety Diagnosis Technology Of Southto-north Water Diversion Project Based On Set Pair Analysis

This paper,taking the typical channel section of Huixian County,Xinxiang city of the South-to-North Water Diversion Project Middle Route as an example,through the analysis of the characteristics of the South-to-North Water Diversion Linear Project,carries out the study of seepage and settlement safety.It selects the typical channel sections such as high fill,deep excavation,and expansive soil to study the spatial and temporal characteristics of various detection items and monitoring data that reflect the observed behaviors of the South-to-North Water Diversion Linear Project,and analyzes the change law of monitoring data and its influencing factors(environment variables),establishing a single-point mathematical model based on the neural network forecasting analysis method,so as to realize the correlation between monitoring data and environment variables;it studies the characteristics of timevarying law of settlement values,osmotic pressure values as well as the change law of environment variables.Combining with the monitoring index value method and the cloud model theory method,it establishes the single-point multi-factor time-effect model of monitoring data.By integrating the model with the set pair analysis method,the determination of monitoring data evaluation level is realized.Besides,this paper collects the data of detection information of the channel section to realize the anomaly discrimination of the detection information;it focuses on the method of identifying single-project data hidden danger of the South-to-North Water Diversion Linear Project such as analytic hierarchy process and rough set theory to study the timeliness and accuracy in response to the abnormal conditions between different monitoring data and detection information,establishing a sensitivity index system for different data information of different projects,so as to provide a basis for the distribution of weights.Moreover,in this paper,a comprehensive analysis method is used to consider the type,area and location of the detection information of the South-to-North Water Diversion Linear Project,and to study the relationship among environment variables,monitoring data and detection information,building the adaptive evaluation method to reflect the working state and proposing a diagnostic technology for the safe operation of the South-to-North Water Diversion Linear Project.The main achievements are as follows:The single-point mathematical model of seepage and settlement as well as the multi-factor time-effect model of single project and single point are established.Because the long and short neural network is a feed-forward type,its internal hidden layer nodes can be connected with each other,and have a certain memory unit.Through setting the mechanisms of input gate,forgetting gate and output gate,etc,this network is used to control the transmission of timeeffect information of seepage or settlement at different times.Therefore,on the basis of this characteristic,a single-point mathematical model of channel seepage and settlement is established,and the correlation between environmental factors,detection information and seepage pressure value or settlement value is realized.What’s more,the establishment of a multi-factor time-effect model of single project and single point of channel seepage and settlement is based on the cloud model theory.The channel osmotic pressure value or settlement value both varies with time,so it is called timeliness.What’s more,timeliness is caused by the combined action of multiple factors.Multi-factor time-effect model of channel seepage and settlement is established by analyzing a large number of measured data that change with time under the guidance of cloud model theory,providing the basis for building evaluation criteria of monitoring data.(2)Based on the rough set theory and analytic hierarchy process,this paper establishes the sensitivity index system of channel safety evaluation and the evaluation level of monitoring data according to the set pair analysis method.According to the basic characteristics of each typical channel section,the sensitivity index system of channel seepage and settlement is established by using rough set theory and analytic hierarchy process.The sensitivity index system is divided into 3 levels.Through the analytic hierarchy process,the division of the 3-level index sensitivity and the weights distribution of the flat section monitoring points as well as the detection information method can be realized;by using the rough set theory,the decision table is built according to the relationship between osmotic pressure values and environment variables,settlement values and environment variables,so as to realize the importance division between condition attributes and decision attributes,thereby realizing the sensitivity division of environment variables.In line with set pair analysis,channel safety evaluation standard set and evaluation index set are established,and the establishment of evaluation level of monitoring data is achieved through the calculation of connection degree.The distribution of weights is mainly based on the sensitivity system.(3)Anomaly discrimination of detection information.By collecting data such as geological radar,high-density electrical method and surface wave method,and using tomography and other methods to achieve image output,qualitative analysis of channel segment safety and the data assignment are realized in accordance with experts’ experience and the binaryzation method respectively.(4)Establishment of diagnostic technology for safe operation of the South-to-North Water Diversion Linear Project.According to the monitoring data set,the analysis results are dimensionally processed.It means that the evaluation level of the monitoring data is normalized and is combined with the anomaly discrimination of detection information,and the weight is allocated according to the sensitivity.By doing so,the safety discrimination of the typical channel section of the South-to-North Water Division Project is achieved.The application example proves that the above research method can realize the safety evaluation of the channel section,and the model method realizes the integration of detection information and monitoring data,which provides a more comprehensive evaluation method for the safety of the channel section.