Water quality risk analysis method for inter-basin water transfer projects associated with flood events
By constructing a risk analysis method for inter-basin water transfer projects, and combining a hydrodynamic-water quality model and the Copula function, the problem that existing technologies cannot accurately reflect the associated water quality risks of flood events has been solved, thereby improving the accuracy of risk analysis and enabling scientific risk management.
Patent Information
- Application Number
- CN202411175778.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-26
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-08-26
AI Technical Summary
Existing risk analysis methods for inter-basin water transfer projects are insufficient to accurately reflect the water quality risks associated with flood events, especially during flood season water transfers, and cannot effectively predict risk changes after future project scheduling adjustments.
A method for analyzing water quality risks associated with flood events during inter-basin water transfer projects is developed. By collecting spatial geographic and water transfer project data, a three-level river and lake system is formed, risk factor sequences and risk value sequences are constructed, key risk factors are screened, and the risk contribution rate is analyzed using a hydrodynamic-water quality model and Copula function. Combined with data-driven and physical mechanism models, the method reflects the physical attributes, multi-factor effects, and randomness of the risks.
It improves the accuracy of quantitative risk analysis, accurately reflects the randomness and multi-factor interaction of risk events, and provides scientific risk management decision support.
Smart Images

Figure CN118691097B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to risk analysis associated with water diversion projects, and in particular to a method and system for analyzing water quality risks associated with flood events encountered by water diversion projects across river basins. Background Art
[0002] The construction of inter-basin water diversion projects is an effective way to solve the problem of uneven temporal and spatial distribution of water resources. While it brings social and economic benefits, it will inevitably bring certain negative effects. Among them, the water quality risk associated with water diversion projects has become one of the important factors restricting the sustainable development of projects. With the increase in global drought events caused by global warming, more and more water diversion projects are carrying out emergency water diversion during the flood season. During the flood season, the pollution sources and hydrological conditions of the reservoirs along the project are quite different from those in the non-flood season. When conducting the associated water quality risk analysis, it is necessary to take into account the randomness of the hydrological conditions and the physical mechanism of the migration and diffusion of pollutants in the water body, improve the accuracy of risk quantitative analysis, and provide a quantitative decision-making basis for the risk management of water diversion projects to achieve the operation goals of water diversion projects.
[0003] Existing risk analysis methods can be roughly divided into two categories, one based on data-driven methods and the other based on physical mechanism methods. Data-driven risk analysis methods can better reflect the random characteristics of risks and can also consider the impact of multiple risk factors, but their analysis process mainly relies on historical long-term data and lacks consideration of the physical mechanism of risk occurrence. It is especially difficult to predict the changes in risks after future engineering scheduling adjustments. On the other hand, although the risk analysis method based on the physical mechanism model can better reflect the physical mechanism of risk occurrence, it is unable to accurately reflect the randomness of risk events and the interaction properties of multiple risk factors due to the limited artificially set risk scenarios and the fact that each risk factor is often analyzed separately. Summary of the invention
[0004] The purpose of the invention is to provide a method for analyzing water quality risks associated with flood events in inter-basin water transfer projects to solve the above-mentioned problems in the prior art. On the other hand, a system for implementing the above-mentioned water quality risk analysis for inter-basin water transfer projects when water transfer encounters flood events is provided.
[0005] Technical solution: water quality risk analysis method for inter-basin water transfer projects encountering flood events, including the following steps:
[0006] Step S1, collecting spatial geography and water diversion project data, obtaining and analyzing the topological relationship between river and lake systems and water diversion projects, clustering rivers, and forming at least three-level river and lake systems;
[0007] Step S2, constructing a risk factor sequence and a risk value sequence of water quality risk for the river and lake system;
[0008] Step S3, selecting key risk factors from the risk factor sequence, and constructing a water quality risk analysis model based on the key risk factors and risk values;
[0009] Step S4, setting a risk scenario set and analyzing the associated water quality risk of the water diversion project encountering a flood event during water diversion through the water quality risk analysis model, and calculating the risk contribution rate of each secondary river zone;
[0010] Step S5, select a risk event from the risk event set, and select a time period during the occurrence of the risk event, analyze the associated water quality risks of the water diversion project in the selected time period when encountering a flood event during water diversion through the water quality risk analysis model, and calculate the change in the risk contribution rate of each secondary river division in the time period.
[0011] According to one aspect of the present application, step S1 further comprises:
[0012] Step S11, analyzing the topological relationship of the geographical location of rivers and lakes and the correlation of runoff;
[0013] Step S12: clustering rivers to construct a virtual river-lake system;
[0014] Step S13: Collect basin hydrological data, calculate the flow time series of the virtual river, and extract n flood processes.
[0015] According to one aspect of the present application, step S11 further comprises:
[0016] Step S11a, analyzing the topological relationship of the geographical location of rivers and lakes and the correlation of runoff;
[0017] Step S11b, construct the topological relationship between lakes, first-, second- and third-level rivers, and calculate the comprehensive correlation coefficient of the direct runoff correlation index of the second-level rivers.
[0018] According to one aspect of the present application, step S12 further comprises:
[0019] Step S12a, based on the water flow propagation time, the flow of all third-level rivers is accumulated to the second-level rivers that have a direct subordinate topological relationship with them;
[0020] Step S12b: for each first-level river, read the second-level rivers that have a direct topological relationship with it, and cluster several second-level rivers whose comprehensive correlation coefficient is greater than a threshold into a group.
[0021] Step S12c, repeat steps S12a-S12b to complete the clustering of the third-level and second-level rivers, and construct a virtual river-lake system with R1 first-level rivers, R2 second-level river divisions, and 1 lake, where R1 and R2 are positive integers.
[0022] According to one aspect of the present application, step S13 is further:
[0023] Step S13a, collecting basin hydrological data, calling the measured flow of the clustered rivers, and calculating the flow time series of the virtual river;
[0024] Step S13b, divide the flow change process into m time periods based on a predetermined step size. For each time period, use the MK trend analysis method in sequence to calculate the MK statistic and the corresponding significance of the runoff change in the current time period, collect flood event time periods, and the flood event time period is a period of flow change process that includes no less than a consecutive period of significant flow increase and no less than b consecutive period of significant flow decrease. The process is a flood. a is the rising time threshold of different types of floods, and b is the decline time threshold of different types of floods, and n flood processes are obtained.
[0025] According to one aspect of the present application, step S2 further comprises:
[0026] Step S21, based on the extracted n floods, construct a risk factor sequence with a length of n, wherein the risk factors include indicators describing flood volume characteristics: total flood volume, flood peak flow; indicators describing flood time characteristics: flood duration, flood peak occurrence time; indicators describing flood process changes: peak rise rate, peak fall rate; indicators describing flood morphology: coefficient of variation, skewness coefficient, a total of 8 indicators;
[0027] Step S22, construct a one- and two-dimensional coupled hydrodynamic-water quality model, simulate the pollutant concentration values at the water diversion outlet of the lake under the action of n flood processes, and construct a risk value sequence with a length of n based on the pollutant concentration values.
[0028] According to one aspect of the present application, step S22 is further:
[0029] Step S22a, constructing a one- and two-dimensional coupled hydrodynamic water quality model of virtual rivers and lakes, taking the flow process of each flood, the flow of water diversion projects into and out of the regulating lake, and the status of pollution sources in the flood season as model inputs, and simulating the change process of lake pollutant concentration under the action of each flood during water diversion;
[0030] Step S22b, setting the pollutant concentration at the water outlet of the regulating lake as the water quality risk index of the water, extracting the pollutant concentration values at the water outlet at the end of each flood, and obtaining the risk value sequence of the water quality of the water.
[0031] According to one aspect of the present application, step S3 further comprises:
[0032] Step S31: Based on the obtained risk factor indicator sequence and risk value indicator sequence, construct the Kendall correlation matrix of the two, calculate the correlation coefficient, and take the correlation coefficient greater than 0.5 as the threshold to determine the risk factor set significantly correlated with the water quality risk of the water diversion, wherein the risk factor set includes x flood characteristic indicators, where x is a positive integer;
[0033] Step S32: Use the Copula function to construct an x+1-dimensional joint probability distribution of x key risk factors and risk value indicators to obtain a water quality risk analysis model.
[0034] According to one aspect of the present application, step S32 is further:
[0035] Step S32a, constructing a distribution function set, and using the distribution functions in the distribution function set to perform distribution fitting on the risk factor indicators and the risk value indicators one by one;
[0036] Step S32b, respectively calculating and normalizing the Akaike information criterion, the Bayesian information criterion and the root mean square error of each fitted distribution function, taking the average of the three as the comprehensive comparison index for risk factor distribution fitting, and selecting the distribution with the smallest comprehensive comparison index as the optimal distribution;
[0037] Step S32c, based on the commonly used Copula function, construct an x+1 dimensional joint distribution of x risk factor variables and risk value variables, use the maximum likelihood estimation method to estimate the parameters of the Copula function, fit the x+1 dimensional joint distribution, calculate the Akaike information criterion and Bayesian information criterion indicators of the fitted candidate distribution, respectively, use the average values of the Akaike information criterion and the Bayesian information criterion indicators as the comprehensive evaluation index of the degree of fitting, compare the candidate Copula functions of each joint distribution, and select the candidate joint distribution with the smallest comprehensive evaluation index as the optimal joint distribution, which is the water quality analysis model.
[0038] According to one aspect of the present application, step S4 is further:
[0039] Step S41, constructing a water quality risk event set for water diversion, setting the upper and lower limits of the intervals of Class IV water, Class V water, and Class V-lower water as risk value indicators for the occurrence of water quality risk events for water diversion, and determining the risk value indicators for each water quality risk event for water diversion and its occurrence;
[0040] Step S42, select one risk factor at a time, keep the other x-1 risk factors unchanged, change the value of the risk factor, input the water quality risk analysis model, output the probability of occurrence of the risk event, perform the above calculations on each risk factor and each risk event of different intensity in turn, and obtain the change of the probability of occurrence of water quality risk of water diversion of different intensities with the risk factor;
[0041] Step S43: for each risk factor and each risk event, the flow of one of the secondary river divisions is closed one by one, and the probability of occurrence of the corresponding risk event is calculated again. The probability difference between the probability of occurrence of the risk event and the probability of occurrence of the risk event before the closure is calculated, which is the risk contribution of the secondary river division.
[0042] Step S44, calculating the risk contribution rate for each secondary river partition in turn, where the risk contribution rate is the ratio of the contribution of the secondary river partition to the sum of the risk contributions of all secondary river partitions.
[0043] According to one aspect of the present application, step S5 is further:
[0044] Step S51, selecting a risk event from the constructed water quality risk event set for water diversion, and extracting a certain time period during the occurrence of the selected risk event;
[0045] Step S52: in the extracted time period, calculate the three-dimensional joint probability distribution of the water quality risk occurrence probability of the water diversion as a function of the risk factor and time through step S42;
[0046] Step S53: for each risk factor and selected risk event, the flow of one of the secondary river divisions is closed one by one, and the probability of occurrence of the corresponding risk event in the extracted time period is calculated again, and the probability difference between the probability of occurrence of the risk event and the probability of occurrence of the risk event before the closure is calculated, that is, the change of the risk contribution of the secondary river division over time;
[0047] Step S54, calculate the change in risk contribution rate for each secondary river partition in turn, the risk contribution rate is the ratio of the contribution of the secondary river partition to the sum of the risk contributions of all secondary river partitions, and the change in risk contribution rate is the change in the ratio of the contribution of the secondary river partition to the sum of the risk contributions of all secondary river partitions over time.
[0048] According to another aspect of the present application, a system for analyzing water quality risks associated with flood events in inter-basin water diversion projects is provided, comprising:
[0049] at least one processor; and
[0050] a memory communicatively connected to at least one of the processors; wherein,
[0051] The memory stores instructions that can be executed by the processor, and the instructions are used to be executed by the processor to implement the method for analyzing water quality risks associated with flood events in inter-basin water diversion projects as described in any of the above technical solutions.
[0052] Beneficial effects: The present invention provides a new technical idea. Based on the hydrological runoff situation theory, hydrodynamic-water quality model and Copula function, a risk analysis method that comprehensively considers the risk physical process, random process and multi-factor action attributes is constructed. The water quality risk analysis method associated with flood events in inter-basin water diversion projects can truly reflect the regional characteristics, uncertainty and multi-factor action attributes of risks, and has the advantages of both data-driven models and physical mechanism models, which improves the accuracy of quantitative risk analysis. Combined with the actual needs of water diversion during the current flood season, it is possible to more scientifically and reasonably identify and predict the water quality risks associated with the regulating and storing lakes along the flood season operation of water diversion projects, which has certain guiding significance for risk management and control decisions of water diversion projects. BRIEF DESCRIPTION OF THE DRAWINGS
[0053] Figure 1 It is a flow chart of the present invention.
[0054] Figure 2 It is a flow chart of step S1 of the present invention.
[0055] Figure 3 It is a flow chart of step S2 of the present invention.
[0056] Figure 4 It is a flow chart of step S3 of the present invention.
[0057] Figure 5 It is a flow chart of step S4 of the present invention. DETAILED DESCRIPTION
[0058] like Figure 1 According to one aspect of the present application, a method for analyzing water quality risks associated with flood events in inter-basin water diversion projects is provided, which is characterized by comprising the following steps:
[0059] Step S1, collecting spatial geography and water diversion project data, obtaining and analyzing the topological relationship between river and lake systems and water diversion projects, clustering rivers, and forming at least three-level river and lake systems;
[0060] Step S2, constructing a risk factor sequence and a risk value sequence of water quality risk for the river and lake system;
[0061] Step S3, selecting key risk factors from the risk factor sequence, and constructing a water quality risk analysis model based on the key risk factors and risk values;
[0062] Step S4, setting a risk scenario set and analyzing the associated water quality risks of the water diversion project when encountering flood events during water diversion through the water quality risk analysis model, and calculating the risk contribution rate of each secondary river zone.
[0063] Step S5, select a risk event from the risk event set, and select a time period during the occurrence of the risk event, analyze the associated water quality risks of the water diversion project in the selected time period when encountering a flood event during water diversion through the water quality risk analysis model, and calculate the change in the risk contribution rate of each secondary river division in the time period.
[0064] The present invention takes into account the advantages of data-driven models and physical mechanism models. It considers physical properties when constructing river and lake systems, collecting flood data, and constructing risk factor sequences and risk value sequences. It considers multi-factor action properties and randomness when constructing single-factor distribution and multi-factor joint probability distribution. It comprehensively proposes a risk analysis method that can simultaneously reflect the physical properties, multi-factor action properties and randomness of risks, improves the accuracy of quantitative risk analysis, and points out the direction for risk management and control by tracing the risk contribution rate of secondary river zoning.
[0065] Therefore, in this embodiment, a risk analysis method is constructed which comprehensively reflects the physical properties, multi-factor action properties and randomness of risks. By constructing a topological structure of the spatial relationship between rivers and lakes and clustering rivers, at least a three-level river-lake system is formed. A risk factor sequence and a risk value sequence of water quality risks are constructed, and key risk factors are screened out from the risk factor sequence. A water quality risk analysis model is then constructed based on the key risk factors and risk values. Then, the variable analysis method is used to obtain the probability of occurrence of water quality risks for water diversions of different intensities as the risk factors change, and the risk contribution rate of each secondary river division is calculated. Compared with the commonly used water quality risk analysis methods, the water quality risk analysis method used in this embodiment can accurately reflect the randomness of risk events and the interaction of multiple risk factors, and can also accurately calculate the impact of each secondary river division on the occurrence of risk events, providing data support for predicting risk changes after future engineering scheduling adjustments.
[0066] According to one aspect of the present application, step S1 further comprises:
[0067] Step S11, analyzing the topological relationship of the geographical location of rivers and lakes and the correlation of runoff;
[0068] Step S11a, analyzing the topological relationship of the geographical location of rivers and lakes and the correlation of runoff;
[0069] Step S11b, construct the topological relationship between lakes, first-, second- and third-level rivers, and calculate the comprehensive correlation coefficient of the direct runoff correlation index of the second-level rivers.
[0070] Considering that different correlation test methods may yield different results for the same set of data, in this embodiment, the Pearson, Spearman and Kendall correlation coefficients are used to comprehensively calculate the direct runoff correlation index of the secondary rivers to obtain a comprehensive correlation coefficient.
[0071] Step S12: clustering rivers to construct a virtual river-lake system;
[0072] Step S12a, based on the water flow propagation time, the flow of all third-level rivers is accumulated to the second-level rivers that have a direct subordinate topological relationship with them;
[0073] Step S12b: for each first-level river, read the second-level rivers that have a direct topological relationship with it, and cluster several second-level rivers whose comprehensive correlation coefficient is greater than a threshold into a group.
[0074] Step S12c, repeat steps S12a-S12b to complete the clustering of the third-level and second-level rivers, and construct a virtual river-lake system with R1 first-level rivers, R2 second-level river divisions, and 1 lake, where R1 and R2 are positive integers.
[0075] In this embodiment, the topological relationship of the geographical location of each river channel is taken into account, and the Kendall correlation analysis method is used to calculate the runoff correlation between each river. The rivers with strong runoff correlation are merged and clustered in turn to form secondary river partitions, and the risk contribution rate of each secondary river partition can be accurately analyzed later. When clustering the rivers, only the rivers entering the lake upstream are selected, and the downstream rivers are not analyzed and clustered. Specifically, in a certain embodiment, the hydrological elements of the upstream hydrological stations of the seven rivers entering the lake around the regulating lake from 2007 to 2018 are collected. Since the water quality risk associated with short-flow rapid floods in the east of the lake is mainly considered, the rivers in the west of the lake are not clustered, and the four main rivers in the east of the lake are clustered as virtual rivers in the east of the lake.
[0076] Step S13: Collect basin hydrological data, calculate the flow time series of the virtual river, and extract n flood processes.
[0077] Step S13a, collecting basin hydrological data, calling the measured flow of the clustered rivers, and calculating the flow time series of the virtual river;
[0078] Step S13b, divide the flow change process into m time periods based on a predetermined step size. For each time period, use the MK trend analysis method in sequence to calculate the MK statistic and the corresponding significance of the runoff change in the current time period, collect flood event time periods, and the flood event time period is a period of flow change process that includes no less than a consecutive period of significant flow increase and no less than b consecutive period of significant flow decrease. The process is a flood. a is the rising time threshold of different types of floods, and b is the decline time threshold of different types of floods, and n flood processes are obtained.
[0079] The method of selecting the flood process is usually to extract it by analyzing the 12-hour interval flow data. This method cannot guarantee the accuracy of the flood rise and fall time. Therefore, this embodiment uses the MK trend analysis method to calculate the MK statistic and the corresponding significance of the runoff change in the current period to collect the flood event period and then extract the flood process.
[0080] The MK trend analysis method is a rank-related, non-parametric statistical method. The method does not require samples to obey a specific statistical distribution during implementation, and has good fault tolerance for individual observation outliers. It is often used to test the trend of time series changes. The trend and degree of change of the traffic time series are judged according to the MK statistic. When the MK statistic is greater than 0, the time series has an increasing trend. When the MK statistic is equal to 0, it indicates that the time series has no changing trend. When the MK statistic is less than 0, the time series has a decreasing trend. The absolute value of the MK statistic represents the degree of increase or decrease. In this embodiment, when the absolute value of the MK statistic is greater than 1.64, its significance reaches more than 90%. At this time, the trend of the traffic time series change is significant, and this period is regarded as a period of significant increase in traffic or a period of significant decrease in traffic.
[0081] Obtaining the accurate time points of flood rise and fall can make the subsequent calculation of pollutant concentration values at the water diversion outlet at the end of the flood more accurate, providing accurate data for subsequent analysis and calculation.
[0082] According to one aspect of the present application, step S2 further comprises:
[0083] Step S21, based on the extracted n floods, construct a risk factor sequence with a length of n, wherein the risk factors include indicators describing flood volume characteristics: total flood volume, flood peak flow; indicators describing flood time characteristics: flood duration, flood peak occurrence time; indicators describing flood process changes: peak rise rate, peak fall rate; indicators describing flood morphology: coefficient of variation, skewness coefficient, a total of 8 indicators;
[0084] Indicators describing flood characteristics are selected to construct an initial risk factor set. In the initial risk factor set, there are inevitably different degrees of correlation between the selected factors. There is a lot of redundancy and interference in the information reflected by highly correlated factors, which makes the trend of the same factor data linearly enhanced or weakened. If the physical mechanism model is directly used for simulation, it is easy to cause distortion of the risk analysis results. Therefore, in this embodiment, principal component analysis is used to transform and reduce the dimension of the initial risk factor set indicator system as a whole to obtain the first subset; then the BP neural network is used to delete the indicators in the initial risk factor set one by one to obtain the second subset. The two screening methods have their own advantages and disadvantages, so the union of the first and second subsets is solved to obtain the risk factor set, including indicators describing the characteristics of flood volume: total flood volume and peak flow; indicators describing the time characteristics of flood: flood duration and peak time; indicators describing the changes in flood process: peak rise rate and peak fall rate; indicators describing flood morphology: coefficient of variation and skewness coefficient, a total of 8 indicators.
[0085] Step S22, construct a one- and two-dimensional coupled hydrodynamic-water quality model, simulate the pollutant concentration values at the water diversion outlet of the lake under the action of n flood processes, and construct a risk value sequence with a length of n based on the pollutant concentration values.
[0086] According to one aspect of the present application, step S22 is further:
[0087] Step S22a, constructing a one- and two-dimensional coupled hydrodynamic water quality model of virtual rivers and lakes, taking the flow process of each flood, the flow of water diversion projects into and out of the regulating lake, and the status of pollution sources in the flood season as model inputs, and simulating the change process of lake pollutant concentration under the action of each flood during water diversion;
[0088] Step S22b, setting the pollutant concentration at the water outlet of the regulating lake as the water quality risk index of the water, extracting the pollutant concentration values at the water outlet at the end of each flood, and obtaining the risk value sequence of the water quality of the water.
[0089] According to one aspect of the present application, step S3 further comprises:
[0090] Step S31: Based on the obtained risk factor indicator sequence and risk value indicator sequence, construct the Kendall correlation matrix of the two, calculate the correlation coefficient, and take the correlation coefficient greater than 0.5 as the threshold to determine the risk factor set significantly correlated with the water quality risk of the water diversion, wherein the risk factor set includes x flood characteristic indicators, where x is a positive integer;
[0091] In this embodiment, a Kendall correlation matrix analysis was performed on the characteristic sequences of 13 flood events and the water quality risk value sequences of water diversion, among which the risk factor peak flow rate had the highest correlation with the risk value total phosphorus, and the risk factor peak total volume had the highest correlation with the risk value chemical oxygen demand. Therefore, the peak flow rate was identified as a significantly correlated risk factor for the total phosphorus water quality risk of water diversion, and the total flood volume was identified as a significantly correlated risk factor for the chemical oxygen demand water quality risk of water diversion.
[0092] Step S32: Use the Copula function to construct an x+1-dimensional joint probability distribution of x key risk factors and risk value indicators to obtain a water quality risk analysis model.
[0093] According to one aspect of the present application, step S32 is further:
[0094] Step S32a, constructing a distribution function set, and using the distribution functions in the distribution function set to perform distribution fitting on the risk factor indicators and the risk value indicators one by one;
[0095] Step S32b, respectively calculating and normalizing the Akaike information criterion, the Bayesian information criterion and the root mean square error of each fitted distribution function, taking the average of the three as the comprehensive comparison index for risk factor distribution fitting, and selecting the distribution with the smallest comprehensive comparison index as the optimal distribution;
[0096] When selecting distribution functions for distribution fitting of risk factor indicators and risk value indicators, considering that the accuracy of the analysis results of each distribution function for different indicators is different, in this embodiment, 9 distribution functions are used to perform distribution fitting of variables peak flow, total flood volume, total phosphorus and chemical oxygen content, respectively. These 9 distribution functions are Normal distribution, Exponential distribution, Logistic distribution, Gamma distribution, Weibull distribution, Cauchy distribution, Gumbel distribution, Log-Normal distribution and Log-Logistic distribution. The obtained data are then calculated by Akaike information criterion, Bayesian information criterion and root mean square error respectively, and the results are normalized. The average value of the three is taken as the comprehensive comparison index for risk factor distribution fitting, and the distribution with the smallest comprehensive comparison index is selected as the optimal distribution. Finally, Log-Normal distribution is selected for distribution fitting of peak flow, Cauchy distribution is selected for distribution fitting of total flood volume, Weibull distribution is selected for distribution fitting of total phosphorus, and Weibull distribution is selected for distribution fitting of chemical oxygen demand.
[0097] Step S32c, based on the commonly used Copula function, construct an x+1 dimensional joint distribution of x risk factor variables and risk value variables, use the maximum likelihood estimation method to estimate the parameters of the Copula function, fit the x+1 dimensional joint distribution, calculate the Akaike information criterion and Bayesian information criterion indicators of the fitted candidate distribution, respectively, use the average values of the Akaike information criterion and the Bayesian information criterion indicators as the comprehensive evaluation index of the degree of fitting, compare the candidate Copula functions of each joint distribution, and select the candidate joint distribution with the smallest comprehensive evaluation index as the optimal joint distribution, which is the water quality analysis model.
[0098] When the Copula function is selected to construct the x+1-dimensional joint distribution of x risk factor variables and the risk value variable, the joint distribution constructed by a single Copula function has different accuracy for the results. The water quality analysis model obtained by using multiple Copula functions for joint distribution and comprehensive evaluation of the results has higher calculation accuracy. Therefore, in this embodiment, five Copula functions are selected to respectively construct the two-dimensional joint distribution of the total flood peak amount and total phosphorus, and the two-dimensional joint distribution of the total flood amount and chemical oxygen demand. The five Copula functions are: Gaussian, Student's T, Clayton, Frank and Gumbel. The maximum likelihood estimation is used to estimate the parameters of the Copula function, and the indicators of the candidate joint distribution are calculated by the Akaike information criterion and the Bayesian information criterion respectively. The average values of the indicators of the Akaike information criterion and the Bayesian information criterion are used as the comprehensive evaluation indicators of the degree of fitting. The candidate joint distribution with the smallest comprehensive evaluation index is selected as the optimal joint distribution. Finally, the Gumbel Copula function is selected to perform a binary joint distribution of the total flood peak amount and total phosphorus, and the Student's The TCopula function performs a binary joint distribution on the total flood volume and the chemical oxygen demand, which is the water quality analysis model selected in this embodiment and used for subsequent calculations.
[0099] According to one aspect of the present application, step S4 is further:
[0100] Step S41, constructing a water quality risk event set for water diversion, setting the upper and lower limits of the intervals of Class IV water, Class V water, and Class V-lower water as risk value indicators for the occurrence of water quality risk events for water diversion, and determining the risk value indicators for each water quality risk event for water diversion and its occurrence;
[0101] In this embodiment, since the water quality requirement for water diversion is Class III water, according to the surface water environmental quality standard GB3838-2002, the water quality risk intensity of water diversion can be divided into three risk intensities: Class IV water, Class V water and Class V water. The risk index value is set to the upper and lower limits of the interval of Class IV water, Class V water and Class V water, and the risk factors are grouped and valued within their sample value range. Specifically, the risk indicators total phosphorus and chemical oxygen demand are set to the upper and lower limits of the interval of Class IV water, Class V water and Class V water, respectively, where the total phosphorus content of Class IV water is greater than 0.05mg / L and not higher than 0.1mg / L, the chemical oxygen demand is greater than 20mg / L and not higher than 30mg / L, the total phosphorus content of Class V water is greater than 0.1mg / L and not higher than 0.2mg / L, the chemical oxygen demand is greater than 30mg / L and not higher than 40mg / L, and the total phosphorus content of Class V water is greater than 0.2mg / L. g / L, chemical oxygen demand is greater than 40mg / L, and then the risk factors of total peak volume and total flood volume are grouped and valued within their sample value range. Taking Class IV water as an example, the specific settings are as follows: the total phosphorus and chemical oxygen demand water quality risk scenarios are set by keeping the total phosphorus values unchanged at 0.05mg / L and 0.1mg / L respectively, and the total peak volume increases from 1 to 15 at an interval of 1 and is combined with the total phosphorus; keeping the chemical oxygen demand unchanged at 20mg / L and 30mg / L respectively, the total flood volume increases from 0 to 120mm at an interval of 10mm and is combined with the chemical oxygen demand to obtain the water diversion water quality risk event set for subsequent analysis.
[0102] Step S42, select one risk factor at a time, keep the other x-1 risk factors unchanged, change the value of the risk factor, input the water quality risk analysis model, output the probability of occurrence of the risk event, perform the above calculations on each risk factor and each risk event of different intensity in turn, and obtain the change of the probability of occurrence of water quality risk of water diversion of different intensities with the risk factor;
[0103] Although the risk analysis method based on the physical mechanism model in the existing risk analysis method can better reflect the physical mechanism of risk occurrence, it is unable to accurately reflect the randomness of risk events and the interaction properties of multiple risk factors due to the limited artificially set risk scenarios and the fact that each risk factor is often analyzed separately. Therefore, in this embodiment, the variable analysis method is used to input the risk factor into the water quality risk analysis model to obtain the probability of occurrence of risk events when one of the risk factors is changed and the other risk factors remain unchanged, and the above calculation is performed on each risk factor and each risk event of different intensities in turn. The obtained probability of occurrence of water quality risks of water diversion of different intensities changes with the risk factor can accurately reflect the probability of occurrence of risk events under the interaction of multiple risk factors, and more accurately reflect the randomness of risk events.
[0104] Step S43: for each risk factor and each risk event, the flow of one of the secondary river divisions is closed one by one, and the probability of occurrence of the corresponding risk event is calculated again. The probability difference between the probability of occurrence of the risk event and the probability of occurrence of the risk event before the closure is calculated, which is the risk contribution of the secondary river division.
[0105] When analyzing the impact of water inflow from each secondary river partition on the occurrence of lake risks, this embodiment closes the flow of the secondary river partition in the virtual river-lake system, keeps other conditions unchanged, and obtains the probability k1 of the water quality risk scenario at the water diversion outlet. It is compared with the probability k when the flow of the secondary river partition is not closed, and the difference ∆k 1= k- k1 is the risk contribution of the secondary river zone to the occurrence of risk events, and then the contribution values of R2 secondary river zones are simulated and calculated in sequence (∆k1, ∆k2, ∆k3, ∆k4, ∆k5, ... ∆k R2 ).
[0106] Step S44, calculating the risk contribution rate for each secondary river partition in turn, where the risk contribution rate is the ratio of the contribution of the secondary river partition to the sum of the risk contributions of all secondary river partitions.
[0107] In this embodiment, by calculating the risk contribution of each secondary river division, it is possible to know the impact of a flood event on the change in pollutant concentration at the water diversion outlet when the secondary river division is affected, and to accurately predict the risk changes brought about by the changes in each secondary river division after future engineering scheduling adjustments. This solves the problem of the existing risk analysis method based on data-driven risk analysis, in which the analysis process mainly relies on historical long-term data, lacks consideration of the physical mechanism of risk occurrence, and is particularly unable to predict the risk changes after future engineering scheduling adjustments, thereby improving the accuracy of quantitative risk analysis.
[0108] According to one aspect of the present application, step S5 is further:
[0109] Step S51, selecting a risk event from the constructed water quality risk event set for water diversion, and extracting a certain time period during the occurrence of the selected risk event;
[0110] Step S52: in the extracted time period, calculate the three-dimensional joint probability distribution of the water quality risk occurrence probability of the water diversion as a function of the risk factor and time through step S42;
[0111] Step S53: for each risk factor and selected risk event, the flow of one of the secondary river divisions is closed one by one, and the probability of occurrence of the corresponding risk event in the extracted time period is calculated again, and the probability difference between the probability of occurrence of the risk event and the probability of occurrence of the risk event before the closure is calculated, that is, the change of the risk contribution of the secondary river division over time;
[0112] Step S54, calculate the change in risk contribution rate for each secondary river partition in turn, the risk contribution rate is the ratio of the contribution of the secondary river partition to the sum of the risk contributions of all secondary river partitions, and the change in risk contribution rate is the change in the ratio of the contribution of the secondary river partition to the sum of the risk contributions of all secondary river partitions over time.
[0113] In order to obtain the changes in water quality risks associated with inter-basin water diversion projects during the entire process of floods, in this embodiment, the changes in the probability of occurrence of water quality risks in the entire flood process are obtained through the three-dimensional joint probability distribution of water quality risk occurrence probability, risk factors and time changes, and the changes in the risk contribution rates of all secondary river divisions are calculated. The contribution rate of each secondary river division to the occurrence of risk events in the constructed virtual river and lake system during the flood event is obtained, which provides data support for predicting the risk changes after future project scheduling adjustments and improves the accuracy of risk quantitative analysis.
[0114] According to another aspect of the present application, a water quality risk analysis system associated with flood events in inter-basin water diversion projects is provided, which is characterized by comprising:
[0115] at least one processor; and
[0116] a memory communicatively connected to at least one of the processors; wherein,
[0117] The memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement any of the above-mentioned methods for analyzing water quality risks associated with flood events in water diversion projects across river basins.
[0118] The preferred embodiments of the present invention are described in detail above; however, the present invention is not limited to the specific details in the above embodiments. Within the technical concept of the present invention, various equivalent transformations can be made to the technical solutions of the present invention, and these equivalent transformations all belong to the protection scope of the present invention.
Claims
1. A method for analyzing water quality risks associated with flood events in inter-basin water transfer projects, characterized in that: The steps include: Step S1, collecting spatial geography and water diversion project data, obtaining and analyzing the topological relationship between river and lake systems and water diversion projects, clustering rivers, and forming at least three-level river and lake systems; Step S2, constructing a risk factor sequence and a risk value sequence of water quality risk for the river and lake system; Step S3, selecting key risk factors from the risk factor sequence, and constructing a water quality risk analysis model based on the key risk factors and risk values; Step S4, setting a risk event set and analyzing the associated water quality risk of flood events encountered by the water diversion project during water diversion through the water quality risk analysis model, and calculating the risk contribution rate of each secondary river zone; Step S5, selecting a risk event from the risk event set, and selecting a time period during the occurrence of the risk event, analyzing the associated water quality risk of the water diversion project in the selected time period when encountering a flood event during water diversion through the water quality risk analysis model, and calculating the change in the risk contribution rate of each secondary river division in the time period; The step S1 is further as follows: Step S11, analyzing the topological relationship of the geographical location of rivers and lakes and the correlation of runoff; Step S12: clustering rivers to construct a virtual river and lake system; Step S13, collecting basin hydrological data, calculating the flow time series of the virtual river, and extracting n flood processes; The step S11 is further as follows: Step S11a, analyzing the topological relationship of the geographical location of rivers and lakes and the correlation of runoff; Step S11b, constructing the topological relationship between lakes, first-, second-, and third-level rivers, and comprehensively calculating the direct runoff correlation index of the second-level rivers to obtain a comprehensive correlation coefficient; The step S12 is further as follows: Step S12a, based on the water flow propagation time, the flow of all third-level rivers is accumulated to the second-level rivers that have a direct subordinate topological relationship with them; Step S12b: for each first-level river, read the second-level rivers that have a direct topological relationship with it, and cluster several second-level rivers whose comprehensive correlation coefficient is greater than a threshold into a group; Step S12c, repeating steps S12a-S12b, completing the clustering of the third-level and second-level rivers, and constructing a virtual river-lake system with R1 first-level rivers, R2 second-level river divisions, and 1 lake, where R1 and R2 are positive integers; The step S13 is further as follows: Step S13a, collecting basin hydrological data, calling the measured flow of the clustered rivers, and calculating the flow time series of the virtual river; Step S13b, dividing the flow change process into m time periods based on a predetermined step length, and for each time period, sequentially using the MK trend analysis method to calculate the MK statistic and corresponding significance of the runoff change in the current time period, collecting the flood event period, and taking the flow change process of the flood event period as a flood, and obtaining n flood processes, wherein the flood event period is a period of time including no less than a consecutive flow significant increase period and no less than b consecutive flow significant decrease period, wherein a is the rise time threshold of different types of floods, and b is the fall time threshold of different types of floods; The step S2 is further as follows: Step S21: Based on the extracted n floods, construct a risk factor sequence with a length of n, wherein the risk factors include indicators describing flood volume characteristics: total flood volume, flood peak flow; indicators describing flood time characteristics: flood duration, flood peak occurrence time; Indicators describing changes in flood processes: peak rise rate, peak fall rate; indicators describing flood morphology: coefficient of variation, coefficient of skewness, a total of 8 indicators; Step S22, constructing a one- and two-dimensional coupled hydrodynamic-water quality model, simulating the pollutant concentration values at the water diversion outlet of the lake under the action of n flood processes, and constructing a risk value sequence with a length of n based on the pollutant concentration values; The step S22 is further as follows: Step S22a, constructing a one- and two-dimensional coupled hydrodynamic water quality model of virtual rivers and lakes, taking the flow process of each flood, the flow of water diversion projects into and out of the regulating lake, and the status of pollution sources in the flood season as model inputs, and simulating the change process of lake pollutant concentration under the action of each flood during water diversion; Step S22b, setting the pollutant concentration at the water outlet of the regulating lake as the water quality risk index of the water, extracting the pollutant concentration value at the water outlet at the end of each flood, and obtaining the risk value sequence of the water quality of the water; The step S4 is further as follows: Step S41, constructing a water quality risk event set for water diversion, setting the upper and lower limits of the intervals of Class IV water, Class V water, and Class V-lower water as risk value indicators for the occurrence of water quality risk events for water diversion, and determining the risk value indicators for each water quality risk event for water diversion and its occurrence; Step S42, select one risk factor at a time, keep the other x-1 risk factors unchanged, change the value of the risk factor, input the water quality risk analysis model, output the probability of occurrence of the risk event, perform the above calculations on each risk factor and each risk event of different intensity in turn, and obtain the change of the probability of occurrence of water quality risk of water diversion of different intensities with the risk factor; Step S43: for each risk factor and each risk event, the flow of a secondary river partition is closed one by one, and the probability of occurrence of the corresponding risk event is calculated again. The probability difference between the probability of occurrence of the corresponding risk event after closing the flow of a secondary river partition and the probability of occurrence of the risk event before closing is calculated, which is the risk contribution of the secondary river partition; Step S44, calculating the risk contribution rate for each secondary river partition in turn, where the risk contribution rate is the ratio of the contribution of the secondary river partition to the sum of the risk contributions of all secondary river partitions.
2. The method for analyzing water quality risks associated with flood events in inter-basin water diversion projects according to claim 1, characterized in that: The step S3 is further as follows: Step S31: Based on the obtained risk factor indicator sequence and risk value indicator sequence, construct the Kendall correlation matrix of the two, calculate the correlation coefficient, and take the correlation coefficient greater than 0.5 as the threshold to determine the risk factor set significantly correlated with the water quality risk of the water diversion, wherein the risk factor set includes x flood characteristic indicators, where x is a positive integer; Step S32: Use Copula function to construct an x+1-dimensional joint probability distribution of x key risk factors and risk value indicators to obtain a water quality risk analysis model; The step S32 is further as follows: Step S32a, constructing a distribution function set, and using the distribution functions in the distribution function set to perform distribution fitting on the risk factor indicators and the risk value indicators one by one; Step S32b, respectively calculating and normalizing the Akaike information criterion, the Bayesian information criterion and the root mean square error of each fitted distribution function, taking the average of the three as the comprehensive comparison index for risk factor distribution fitting, and selecting the distribution with the smallest comprehensive comparison index as the optimal distribution; Step S32c, based on the commonly used Copula function, construct an x+1 dimensional joint distribution of x risk factor variables and risk value variables, use the maximum likelihood estimation method to estimate the parameters of the Copula function, fit the x+1 dimensional joint distribution, calculate the Akaike information criterion and Bayesian information criterion indicators of the fitted candidate distribution, respectively, use the average values of the Akaike information criterion and the Bayesian information criterion indicators as the comprehensive evaluation index of the degree of fitting, compare the candidate Copula functions of each joint distribution, and select the candidate joint distribution with the smallest comprehensive evaluation index as the optimal joint distribution, which is the water quality risk analysis model.
3. The method for analyzing water quality risks associated with flood events in inter-basin water diversion projects according to claim 1, characterized in that: The step S5 is further as follows: Step S51, selecting a risk event from the constructed water quality risk event set for water diversion, and extracting a certain time period during the occurrence of the selected risk event; Step S52: in the extracted time period, calculate the three-dimensional joint probability distribution of the water quality risk occurrence probability of the water diversion as a function of the risk factor and time through step S42; Step S53: for each risk factor and selected risk event, the flow of a secondary river partition is closed one by one, and the probability of occurrence of the corresponding risk event in the extracted time period is calculated again, and the probability difference between the probability of occurrence of the corresponding risk event after closing the flow of a secondary river partition and the probability of occurrence of the risk event before closing is calculated, that is, the change of the risk contribution of the secondary river partition over time; Step S54, calculate the change in risk contribution rate for each secondary river partition in turn, the risk contribution rate is the ratio of the contribution of the secondary river partition to the sum of the risk contributions of all secondary river partitions, and the change in risk contribution rate is the change in the ratio of the contribution of the secondary river partition to the sum of the risk contributions of all secondary river partitions over time.
4. The water quality risk analysis system for flood incidents associated with inter-basin water diversion projects is characterized by: include: at least one processor; as well as a memory communicatively connected to at least one of the processors; wherein, The memory stores instructions executed by the processor, and the instructions are used to be executed by the processor to implement the method for analyzing water quality risks associated with flood events in inter-basin water diversion projects as described in any one of claims 1 to 3.
Citation Information
Patent Citations
Important water functional area water quality risk assessment method based on multi-source data coupling
CN110765419A
Non-uniform multi-source flood encountering risk analysis method
CN111080059A
River network area water quality section pollution contribution rate accounting method based on water quality model
CN111598754A
Method for identifying and controlling associated risk of river and lake water system communication
CN115115269A
Regional river and lake water system communication combined dispatching and water quality safety guarantee method and system
CN115375198A
