A river type water source risk comprehensive assessment and prediction method, system and electronic equipment

Abstract

The application discloses a river type water source risk comprehensive evaluation and prediction method and system and electronic equipment, and the method comprises the following steps: constructing a river type water source risk evaluation index system; establishing a single water source static risk situation evaluation model; establishing a single water source dynamic risk trend evaluation model; and constructing a multi-water source system comprehensive risk simulation prediction model based on system dynamics. The single water source static risk situation evaluation and the dynamic risk trend evaluation are combined innovatively, the static and dynamic combination of the single water source risk evaluation is realized, the risk mutual feedback relationship among the single water sources is further analyzed by using the cause-effect loop diagram and the stock-flow diagram, the simulation prediction of the comprehensive risk level of the river type water source system is realized, the risk factors faced by the river type water source can be comprehensively identified, the system risk can be quantitatively analyzed, the risk management level can be effectively improved, and the safe and stable operation of the river type water source is ensured.

Description

Technical Field

[0001] This invention relates to water source risk assessment and prediction technology, specifically to a comprehensive risk assessment and prediction method, system, and electronic equipment for river-type water sources. Background Technology

[0002] Based on the characteristics of the water source, drinking water sources are divided into surface water sources and groundwater sources. Surface drinking water sources are divided into centralized and decentralized types. Among them, centralized surface water sources are further divided into river type, reservoir type and lake type.

[0003] River-type water sources have the largest water supply scale in my country, serving approximately 43% of the total population served by these sources. River-type water sources are complex systems, and risk assessment and management are prerequisites for ensuring drinking water safety. Current research on risk assessment for river-type water sources fails to adequately meet the needs of high-level safety management in the new era, both in depth and breadth. In terms of breadth, research on water source risk assessment mainly focuses on evaluating single risk sources such as sudden environmental pollution accidents or water quality, water environment, and water quantity, lacking a unified description of various risks and failing to comprehensively reflect the risk status of water sources, making it difficult to identify existing but not readily apparent risk factors. In terms of depth, research mainly focuses on analyzing the risk status of a single water source at a specific point in time, without considering risk development trends. Furthermore, the open nature of river-type water sources means their water quality is easily affected by upstream water and the surrounding environment. Therefore, in-depth research is needed on multi-source system risk simulation and prediction from the perspective of the left and right banks and the entire upstream and downstream areas. Summary of the Invention

[0004] Purpose of the invention: In order to overcome the shortcomings of existing water source risk assessment technologies in terms of depth and breadth, this invention provides a comprehensive risk assessment and prediction method for river-type water sources. This method conducts comprehensive risk assessment and prediction research on river-type water sources from three perspectives: static risk level evaluation of a single water source, dynamic risk trend analysis, and risk simulation prediction of a multi-water source system. It can take into account the risk status and development trend of river-type water source systems and achieve simulation prediction of comprehensive risk levels.

[0005] Technical Solution: To achieve the above objectives, this invention provides a comprehensive risk assessment and prediction method for river-type water source areas, comprising the following steps:

[0006] S1. Construction of Risk Assessment Index System for River-type Water Sources: First, based on the pressure-state-response structural model, various factors affecting the risk status of water sources are analyzed. Second, a preliminary selection of risk assessment indexes is conducted from four aspects: water quality, water quantity, ecology, and management of river-type water sources. Then, after qualitative screening using the A. Gulin method, data standardization is performed, a correlation coefficient matrix is ​​constructed, and the matrix eigenvalues ​​and eigenvectors are calculated to obtain principal components with a cumulative contribution rate greater than a threshold. By calculating the factor loading matrix and principal component contribution rate matrix of each principal component, the correlation between each original index and the principal component and its contribution rate to the principal component are determined. Original indicators that are independent of each other and can objectively represent the evaluation object are obtained, thus completing the quantitative screening of indicators. Then, a reasonableness test is conducted to establish a risk assessment index system for river-type water sources. Next, the connotation of each evaluation index is defined, and the risk threshold of the corresponding index is reasonably determined.

[0007] S2. Static Risk Status Assessment of a Single Water Source: First, the comprehensive weight of each evaluation indicator is determined using the entropy weight-analysis-level combined weighting method. Second, based on grey theory and fuzzy comprehensive evaluation theory, a grey fuzzy comprehensive evaluation model for the static risk status of a single water source is established, including: determining the risk level, forming a sample evaluation matrix, whitening analysis of evaluation indicators, calculation of grey statistics, determining the membership matrix, first-level fuzzy comprehensive risk evaluation and multi-level fuzzy comprehensive risk evaluation, to achieve static risk status assessment of a single water source.

[0008] S3. Dynamic Risk Trend Analysis of Single Water Source Area: Based on the comprehensive weight and membership matrix of each evaluation indicator determined in step S2, and according to set pair analysis and risk similarity and difference inverse evaluation theory, calculate the five-element connection number of each risk evaluation indicator and its corresponding partial connection number; and clarify the risk development trend of each risk evaluation indicator in the same, equal, or opposite state according to the risk connection number trend table, thereby realizing the evaluation of the dynamic risk trend of a single water source area.

[0009] S4. Comprehensive Risk Simulation and Prediction of Multi-Source Water System: Determine the research object and objective of the river-type water source system, establish a risk simulation and prediction model of the river-type water source system based on system dynamics, combine the importance analysis results of each individual water source in the multi-source water system with the quantitative analysis of the influence between water sources, propose the assumption of dynamic risk change, and draw the causal relationship diagram and risk stock flow diagram of the system risk simulation model. Use the static risk status evaluation results of each individual water source in step S2 as boundary conditions to realize the comprehensive risk simulation and prediction of the multi-source water system based on system dynamics.

[0010] Furthermore, in step S1, the A. Gulin method removes redundant indicators based on importance ranking, thereby achieving qualitative screening of evaluation indicators. In step S1, the principal component analysis method further quantitatively screens the original evaluation indicators based on linear correlation and independence.

[0011] Furthermore, the basic idea of ​​the rationality test in step S1 is to calculate the percentage of information contained in the evaluation indicators after quantitative screening relative to the information content of the indicator system before quantitative screening. If this percentage meets the predetermined standard value, it means that the evaluation indicator system after screening can pass the rationality test; otherwise, the evaluation indicator system needs to be readjusted and screened. The rationality test mainly includes the following two basic steps:

[0012] (1) Assuming the number of evaluation indicators before quantitative screening is m, and the number of evaluation indicators is reduced to n after quantitative screening, then the information contribution rate of the evaluation indicator system after quantitative screening is In:

[0013]

[0014] Among them, trS m With trS n These represent the traces of the covariance matrices of the evaluation indicators before and after quantitative screening, respectively.

[0015] (2) If In≥90%, the evaluation index system after screening is relatively reasonable and can pass the rationality test; otherwise, the evaluation index system needs to be adjusted or redesigned.

[0016] Furthermore, in step S1, based on the characteristics of the water source risk and actual business needs, the risk measurement and evaluation standards are divided into five levels: low, relatively low, average, relatively high, and high. To facilitate quantitative calculation, a risk level evaluation set is introduced to determine the corresponding risk level.

[0017] R = {Low risk, Lower risk, Moderate risk, Higher risk, High risk}

[0018] ={(0, 0.2],(0.2, 0.4],(0.4, 0.6],(0.6, 0.8],(0.8, 1]}.

[0019] Furthermore, in step S3, the calculation of the multivariate correlation coefficient and its corresponding partial correlation coefficient for each risk assessment indicator is specifically as follows: The multivariate correlation coefficient is determined based on the risk assessment level of each indicator. The multivariate correlation coefficient refers to the multidimensional development trend of the correlation coefficient difference measurement components, and its general expression form is: u = a + b1i1 + b2i2 + ... + b n-2 i n-2 +cj, where u is called the n-ary connection number, and its first-order partial connection number is:

[0020]

[0021] in,

[0022] Its second-order partial correlation coefficient is:

[0023]

[0024] in,

[0025] And so on, its (n-1)th order partial correlation coefficient is:

[0026]

[0027] in,

[0028] In the formula for the first-order partial correlation coefficient, based on the idea of ​​the dynamic development of each risk assessment indicator system, the definite term 'a' in the n-ary correlation coefficient u was originally located at the level of the uncertain term in the difference measure component b1, and it was obtained from b1 through development and evolution. Used to quantitatively characterize the degree of dynamic development of this evolutionary trend; b1 in u was originally at the level of b2, and it evolved from b2. This is used to quantitatively characterize the degree of dynamic development of this trend; and so on; similarly, b in u... n-2 Originally, it was also at the C level, having evolved from C. Used to quantitatively characterize the degree of dynamic development of this trend; therefore, partial correlation coefficients Based on a positive development trend, the n-ary connection number u = a + b1i1 + b2i2 + ... + b is characterized from a holistic perspective. n-2 i n-2 The evolution of the same / different / anti-definite / uncertain relationship states in +cj. The first-order partial correlation coefficient represents the n-ary correlation coefficient based on a positive development trend.

[0029] Furthermore, the specific process of determining the risk development trend of each evaluation indicator in step S3 is as follows: when the opposing measure component c in the n-ary connection number is not 0, the ratio of the same measure component a to the opposing measure component c is defined as the connection potential or set pair potential of the evaluated set pair under a specific research background, denoted as shi(H)=a / c.

[0030] When a / c > 1, the relationship potential is in a state of same potential under the specific research background, indicating that the two sets being evaluated have the same development trend in their similarity and opposition relationships. When a / c = 1, the relationship potential is in a state of equilibrium under the specific research background, indicating that the two sets being evaluated have a "balanced" state in terms of their similarity and opposition trends. When a / c < 1, the relationship potential is in a state of opposition under the specific research background, indicating that the two sets being evaluated have opposing development trends. By using the relationship potential or set-pair potential, the risk development trend of the system can be divided into similar potential, equilibrium potential, and opposition potential. Furthermore, based on the magnitude of each coefficient, the development trend of the system can be ranked.

[0031] In the study on the comprehensive risk assessment of river-type water source areas, when the risk levels of the evaluation indicators are in the same potential zone, it indicates that the actual risk situation of the evaluated object and the reference set have the same trend, that is, the evaluated object is at a "low" risk level. By analyzing the trend of the evaluation indicators and their partial correlation coefficients, we can find evaluation indicators with the first-order partial correlation coefficients in the opposite trend and give them special attention.

[0032] When the risk level of the evaluation indicator is in the equilibrium zone, it means that the actual risk situation of the evaluated object and the reference set are in a state of "equal strength", that is, the evaluated object is at the "medium" risk level. In this case, close attention should be paid to its correlation coefficient and its first-order partial correlation coefficient. For evaluation indicators with the first-order partial correlation coefficient in the opposite direction, the risk status of the corresponding evaluated object should be made to evolve in the same direction, and the risk level should eventually be transformed into the same zone.

[0033] When the risk level of the evaluation indicator is in the reverse zone, it indicates that the actual risk situation of the evaluated object is in opposition to the reference set, that is, the evaluated object is at a "high" risk level. For evaluation indicators whose current risk level is in the reverse zone and whose first-order partial correlation coefficient is in the reverse zone, the risk status of the sensitive evaluation indicator should be transformed into the equilibrium zone, and the risk level should be reversed to the same zone.

[0034] Furthermore, the specific process of drawing the causal relationship diagram and risk stock flow diagram of the system risk simulation model in step S4 is as follows: Through analysis of the river-type water source system, the system risk assessment objectives, assessment scope, and assessment period are clarified; various risk factors affecting the safety of the water source system are identified; and various data related to the quantitative analysis of evaluation indicators are collected to preliminarily clarify the key risk factors and boundary risk factors of the river-type water source system. Based on the logical relationship between the risk assessment objectives of the river-type water source system and each risk factor, as well as the internal feedback and influence among the risk factors of each individual water source, the causal loop diagram and stock flow diagram of the river-type water source system are drawn. Simultaneously, based on the risk of each individual water source... The system risk increment equation and system risk state variable equation are determined by weighting the risk assessment indicators, thereby establishing a system risk simulation and prediction model for river-type water source areas based on system dynamics. Based on the static risk assessment results of a single water source area, the initial values ​​of boundary risk factors of the river-type water source area system and the relevant parameters in each system equation are determined and continuously adjusted to ensure that the system risk simulation and prediction model for river-type water source areas based on system dynamics reflects the actual risk situation of the water source area to the greatest extent. Through system risk simulation, the system risk level in the future is predicted and analyzed, sensitive risk factors of the river-type water source area system are identified, and corresponding countermeasures are proposed to actively prevent and control the system risk of the water source area.

[0035] Beneficial effects: Compared with the prior art, the significant technical effects of the present invention are:

[0036] (1) Taking into account the four aspects of water quality, water quantity, ecology and management, the risk threshold of the corresponding indicators is determined through qualitative and quantitative screening of indicators, and a relatively comprehensive risk assessment indicator system for river-type water sources is constructed, which can provide important support for improving the risk management system of river-type water sources.

[0037] (2) A comprehensive risk evaluation model for water source areas was established, including a grey fuzzy comprehensive evaluation model and a risk similarity and difference inverse evaluation model based on set pair analysis and five-element connection number. The static risk level and dynamic development trend of a single water source area were analyzed. Furthermore, based on system dynamics, the simulation prediction of the comprehensive risk level of a multi-water source area system was realized. This overcame the shortcomings of traditional risk assessment methods that could not take into account both the deterministic and uncertain risk factors of the system, and expanded the research ideas of risk assessment technology for river-type water source areas. Attached Figure Description

[0038] Figure 1 This is a flowchart of the method of the present invention;

[0039] Figure 2 This is a detailed technical framework diagram of the method of the present invention;

[0040] Figure 3 Qualitative screening results for the risk assessment indicator system of river-type water source areas;

[0041] Figure 4 A causal loop diagram of risks in six river-type water source systems along the Nanjing section of the Yangtze River;

[0042] Figure 5 A causal tree diagram of the comprehensive risk of water sources in the Nanjing section of the Yangtze River;

[0043] Figure 6 The risk causal tree diagram for the Jiangpu-Pukou water source area is shown, where (a) represents water quality change - water pollution risk, (b) represents water quantity change - water shortage risk, (c) represents ecological environment change - ecological environment risk, and (d) represents management safety level change - management safety risk.

[0044] Figure 7 The following is a causal tree diagram of the risks of the Yanziji water source area, in which (a) represents water quality change - water pollution risk, (b) water quantity change - water shortage risk, (c) ecological environment change - ecological environment risk, and (d) management safety level change - management safety risk.

[0045] Figure 8 Risk stock flow diagram of Zihuizhou single water source system;

[0046] Figure 9 A map showing the risk stock flow of the multi-source water system in the Nanjing section of the Yangtze River;

[0047] Figure 10The comprehensive risk level prediction results of the embodiments of the present invention are as follows: (a) is Zihuizhou, (b) is Jiajiang South, (c) is Jiajiang North Estuary, (d) is Yanziji, (e) is Jiangpu-Pukou, (f) is Baguazhou (left branch) Shangba, and (g) is the Yangtze River Nanjing section multi-source system. Detailed Implementation

[0048] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0049] To overcome the shortcomings of existing water source risk assessment technologies in terms of depth and breadth, this invention considers four aspects: water quality, water quantity, ecology, and management. Through qualitative and quantitative screening of indicators, it analyzes and determines the risk thresholds of corresponding indicators, constructing a relatively comprehensive risk assessment indicator system for river-type water sources. This provides important support for improving the risk management system of river-type water sources. Simultaneously, this invention innovatively combines static risk status assessment and dynamic risk trend assessment of a single water source, achieving a dynamic and static integration of risk assessment for a single water source. Furthermore, it uses causal loop diagrams and stock-flow diagrams to analyze the risk feedback relationships between individual water sources, enabling simulation prediction of the comprehensive risk level of the river-type water source system. This allows for the comprehensive identification of risk factors faced by river-type water sources, quantitative analysis of system risks, effective improvement of risk management, and ensuring the safe and stable operation of river-type water sources. This invention provides a comprehensive risk assessment and prediction method for river-type water sources, including: constructing a risk assessment indicator system for river-type water sources; determining the meaning and risk level of evaluation indicators; static risk status assessment of a single water source; dynamic risk trend analysis of a single water source; and simulation prediction of the comprehensive risk of a multi-water source system. Figure 1 and Figure 2 As shown, the specific steps include:

[0050] S1. The construction of the risk assessment indicator system for river-type water sources first follows the principles of scientific rigor, practicality and forward-looking perspective, comprehensiveness and focus on key areas, integrity and hierarchy, comparability and expandability. Initial selection of risk assessment indicators is conducted from four aspects: water quality, water quantity, ecology, and management of river-type water sources. Based on the preliminary construction of the indicator system, qualitative and quantitative screening are used to optimize the indicators, avoiding redundancy, interference, and overlapping effects. After determining the risk assessment indicator system, the connotation of each indicator is defined, and the corresponding risk threshold is reasonably determined. Details are as follows:

[0051] A preliminary risk assessment index system for river-type water source areas was selected. Specifically, based on the definition, connotation, and characteristics of water source safety, and using the Pressure (P)-State (S)-Response (R) structural model, various factors influencing the risk status of water source areas were analyzed. A preliminary risk assessment index system for river-type water source areas was established, consisting of a target layer, a criterion layer, and an index layer, with a multi-layered hierarchical structure. The target layer macroscopically reflects the overall risk status of the water source area. The criterion layer includes four aspects of water source safety, corresponding to water quality, water quantity, ecological environment, and management safety. Each criterion layer is further subdivided into several indicators. The index layer is the smallest evaluation unit in the index system, directly used to measure the magnitude of the corresponding risk status of the water source area. Based on the connotation of water source safety and the PSR structural model, the meaning and risk threshold of each indicator in the index layer were determined, thus establishing a preliminary risk assessment index system for river-type water source areas.

[0052] The selection of risk assessment index system for river-type water source areas includes qualitative screening and quantitative screening. The qualitative screening uses the A. Gulin method to delete indicators with low importance and weak impact on the evaluation target according to the importance ranking, thereby realizing the qualitative screening of evaluation indicators and constructing a three-level hierarchical structure evaluation index system with clear hierarchy and logic.

[0053] For the risk assessment index system of river-type water source areas after qualitative screening, data standardization is carried out, correlation coefficient matrix of time series of each evaluation index is constructed, matrix eigenvalues ​​and eigenvectors are calculated, and principal components with cumulative contribution rates greater than 85% are obtained. By calculating the factor loading matrix and principal component contribution rate matrix of each principal component, the correlation between each original index and the principal component and the magnitude of their contribution rate to the principal component are determined, and original indicators that are independent of each other and can objectively represent the evaluation object are obtained, thereby completing the quantitative screening of indicators.

[0054] Following qualitative and quantitative screening, a rationality test is conducted primarily to examine whether the retained evaluation indicators can reflect the integrity and significance of the water source risk assessment issue. This means that, during the construction of the indicator system, the selected main indicators must be sufficient to express the characteristics of the water source risk assessment system. The basic idea of ​​the rationality test is to calculate the percentage of information contained in the quantitatively screened evaluation indicators relative to the information contained in the indicator system before quantitative screening. If this percentage meets a predetermined standard value, it indicates that the screened evaluation indicator system passes the rationality test; otherwise, the evaluation indicator system needs to be readjusted and screened again. The rationality test mainly includes the following two basic steps:

[0055] (1) Assuming the number of evaluation indicators before quantitative screening is m, and the number of evaluation indicators is reduced to n after quantitative screening, then the information contribution rate of the evaluation indicator system after quantitative screening is In:

[0056]

[0057] In the formula, trS m With trS n These represent the traces of the covariance matrices of the evaluation indicators before and after quantitative screening, respectively.

[0058] (2) If In≥90%, it means that the evaluation index system after screening is relatively reasonable and can pass the rationality test; otherwise, the evaluation index system needs to be adjusted or redesigned.

[0059] Define the connotation of each risk assessment indicator and reasonably determine the corresponding risk threshold. Based on the characteristics of water source risk sources and actual business needs, classify the risk measurement and evaluation standards into five levels: low, relatively low, moderate, relatively high, and high. To facilitate quantitative calculation, a risk level evaluation set corresponding to the risk degree is introduced.

[0060] R = {Low risk, Lower risk, Moderate risk, Higher risk, High risk}

[0061] ={(0, 0.2], (0.2, 0.4], (0.4, 0.6], (0.6, 0.8], (0.8, 1]}

[0062] S2. A static risk assessment model for a single water source area is established. First, the comprehensive weight of each evaluation indicator is determined using the entropy weight-analytic hierarchy process (AHP). Second, a grey fuzzy comprehensive evaluation model for the static risk situation of a single water source area based on combined weighting is established, using grey theory and fuzzy comprehensive evaluation theory. Specifically, this includes: determining the comprehensive weight of each evaluation indicator; calculating the objective and subjective weights of each indicator using the entropy weight method and the AHP method respectively; and further determining the comprehensive weight of each evaluation indicator using the combined weighting method.

[0063] A grey fuzzy comprehensive evaluation model for the static risk situation of a single water source is established based on combined weighting. This model includes steps such as determining risk levels, forming a sample evaluation matrix, whitening analysis of evaluation indicators, calculating grey statistics, determining the membership matrix, and conducting first-level and multi-level fuzzy comprehensive risk evaluations. This achieves a grey fuzzy comprehensive evaluation of the static risk situation of a single water source. Details are as follows:

[0064] (1) Determine the risk level

[0065] Based on the risk characteristics of river-type water source areas, their risk assessment levels are divided into five standards, described by five risk levels: low, relatively low, moderate, relatively high, and high. To facilitate quantitative calculation, a risk value-correspondence risk level assessment set is introduced:

[0066] V = {Low risk level, lower risk level, moderate risk level, higher risk level, high risk level}

[0067] ={0.2, 0.4, 0.6, 0.8, 1}

[0068] (2) Constructing the sample evaluation matrix

[0069] The evaluation vector for each qualitative indicator is determined based on the expert scoring method. Each expert scores the qualitative indicator according to the evaluation principles, resulting in the score d of expert j for indicator i. ij The average of the scores given by the experts is taken, and the initial values ​​of each quantitative indicator are normalized to form a sample evaluation matrix.

[0070] (3) Whitening analysis of indicators

[0071] The number of gray categories is determined according to the risk assessment standard level. In this embodiment, the gray category index set is k = {1, 2, 3, 4, 5}. The whitening function of each indicator is determined according to the number of gray categories. Whitening weight functions of lower limit measure, moderate measure and upper limit measure are used respectively to perform whitening analysis on the evaluation matrix.

[0072] The first gray class (k=1) is defined as the gray class. The whitening weight function is:

[0073]

[0074] The second gray class (k=2) is defined as the gray class. The whitening weight function is:

[0075]

[0076] The third gray class (k=3) is defined as the gray class. The whitening weight function is:

[0077]

[0078] Fourth gray class (k=4), define the gray class The whitening weight function is:

[0079]

[0080] Fifth gray class (k=5), define the gray class The whitening weight function is:

[0081]

[0082] (4) Calculate grey statistics

[0083] Taking indicator D1 as an example, the statistics for each gray category are as follows:

[0084]

[0085]

[0086]

[0087]

[0088]

[0089] (5) Calculate the gray evaluation weights and form the membership matrix.

[0090] For indicator D1, the grey evaluation weight is:

[0091]

[0092]

[0093]

[0094]

[0095]

[0096] Its gray evaluation membership vector r D1 =[r D11 ,r D12 ,r D13 ,r D14 ,r D15 ].

[0097] Similarly, the grey evaluation membership degree r of other evaluation indicators under this criterion layer can be calculated one by one. ij Thus, the risk assessment membership matrix of criterion layer D is obtained. Furthermore, the membership matrix R of each evaluation index in other criterion layers is derived. A R B R C .

[0098] (6) First-level fuzzy comprehensive risk assessment

[0099] From weight set a i ={a i1 ,a i2 ,…,a ij ,…,a in}', (i = 1, 2, ..., m) and membership matrix R i =[r i1 ,r i2 ,…,r ij ,…,rin ], (i = 1, 2, ..., m; j = 1, 2, ..., n) calculate the membership degree c of the i-th criterion layer for each evaluation level. i c i =a i ·R i T .

[0100] For the i-th criterion layer, based on B = C·V T The risk assessment results of the i-th criterion layer can be obtained.

[0101] (7) Multi-level fuzzy comprehensive risk assessment

[0102] From the criterion layer weight set A = {a1, a2, ..., a...} n The second-level fuzzy risk assessment is performed using} and its corresponding membership matrix C, based on E = A·B T The final risk assessment result is obtained.

[0103] S3. A dynamic risk trend assessment model for a single water source area is established. Based on the comprehensive weight and membership matrix of each evaluation indicator determined in step S2, and according to set pair analysis and the theory of risk similarity and difference in reverse evaluation, the multivariate correlation coefficient and its corresponding partial correlation coefficient of each risk evaluation indicator are calculated. Based on the risk correlation coefficient trend table, the risk development trend of each risk evaluation indicator (whether it is in the same, balanced, or opposite trend) is clarified, thereby achieving the assessment of the dynamic risk trend of a single water source area. Details are as follows:

[0104] Calculate the multivariate correlation coefficient and its corresponding partial correlation coefficient for each risk assessment indicator. Specifically, determine the corresponding multivariate correlation coefficient based on the risk assessment level of each indicator. The multivariate correlation coefficient refers to the multidimensional development trend of the correlation coefficient difference measurement components, and its general expression form is: u = a + b1i1 + b2i2 + ... + b n-2 i n-2 +cj, where u is called the n-ary contact number, and its first-order partial contact number is:

[0105]

[0106] in,

[0107] Its second-order partial correlation coefficient is:

[0108]

[0109] in,

[0110] And so on, its (n-1)th order partial correlation coefficient is:

[0111]

[0112] in,

[0113] In equation (2), based on the idea of ​​dynamic development of each risk assessment indicator system, the definite term 'a' in the n-ary connection number u was originally located at the level of the uncertain term in the difference measure component b1, and it was obtained from b1 through development and evolution. Used to quantitatively characterize the degree of dynamic development of this evolutionary trend; b1 in u was originally at the level of b2, and it evolved from b2. Used to quantitatively characterize the degree of dynamic development of this trend; ...; Similarly, b in u n-2 Originally, it was also at the C level, having evolved from C. This is used to quantitatively characterize the degree of dynamic development of this trend. Therefore, partial correlation coefficients... Based on a positive development trend, the n-ary connection number u = a + b1i1 + b2i2 + ... + b is characterized from a holistic perspective. n-2 i n-2 The evolution of the same / different / anti-deterministic / uncertain relationship states in +cj, therefore The first-order partial correlation coefficient represents the n-ary correlation coefficient based on a positive development trend.

[0114] Determine the risk development trend of each evaluation indicator; the specific process is as follows: when the opposing measure component c in the n-ary connection number is not 0, the ratio of the same measure component a to the opposing measure component c is defined as the connection potential or set pair potential of the evaluated set pair under a specific research background, denoted as shi(H)=a / c.

[0115] When a / c > 1, the relationship potential is in a state of same potential under the specific research context, meaning that the two sets being evaluated have the same development trend in their similarity, difference, and opposition relationships. When a / c = 1, the relationship potential is in a state of equilibrium under the specific research context, indicating that the two sets being evaluated have a "balanced" state in terms of their similarity and opposition trends. When a / c < 1, the relationship potential is in a state of opposition under the specific research context, indicating that the two sets being evaluated have opposing development trends. Using the relationship potential (or set pair potential), the risk development trend of the system can be divided into similar potential, equilibrium potential, and opposition potential. Furthermore, the development trend of the system can be ranked based on the magnitude of each coefficient.

[0116] In the comprehensive risk assessment study of river-type water source areas, when the risk levels of evaluation indicators are in the same potential zone, it indicates that the actual risk situation of the evaluated object and the reference set have the same trend, that is, the evaluated object is at a "low" risk level. By analyzing the trend of evaluation indicators and their partial correlation coefficients, evaluation indicators (risk factors) with first-order partial correlation coefficients in the opposite trend are identified and given special attention.

[0117] When the risk level of an evaluation indicator is in the equilibrium zone, it indicates that the actual risk situation of the evaluated object and the reference set are in a state of "evenly matched", that is, the evaluated object is at a "medium" risk level. Close attention should be paid to its correlation coefficient and its first-order partial correlation coefficient. For evaluation indicators with a negative first-order partial correlation coefficient, necessary efforts should be made to promote the risk status of the corresponding evaluated object to evolve in the same direction, ultimately transforming the risk level into the equilibrium zone.

[0118] When the risk level of an evaluation indicator is in the negative trend zone, it indicates that the actual risk situation of the evaluated object is in opposition to that of the reference set, meaning the evaluated object is at a "high" risk level. For evaluation indicators whose current risk level is in the negative trend zone and whose first-order partial correlation coefficient is also in the negative trend zone, it is necessary to further promote the transformation of the risk state of sensitive evaluation indicators towards the equilibrium zone and strive to reverse the risk level to the same trend zone.

[0119] S4. A comprehensive risk simulation model for a multi-source water system based on system dynamics is constructed. The research object and system modeling objective of the river-type source water system are determined. Combining the importance analysis results of each individual source water area in the multi-source water system with the quantitative analysis of the influence between sources water areas, a hypothesis of dynamic risk change is proposed. A causal relationship diagram and a risk stock-flow diagram of the system risk simulation model are drawn. The static risk status evaluation results of each individual source water area in step S2 are used as boundary conditions to achieve comprehensive risk simulation and prediction of the multi-source water system based on system dynamics. Details are as follows:

[0120] The specific process of drawing the causal relationship diagram and risk stock-flow diagram of the system risk simulation model is as follows: Through analysis of the river-type water source system, the system risk assessment objectives, assessment scope, and assessment period are clarified; various risk factors affecting the safety of the water source system are identified; and various data related to the quantitative analysis of evaluation indicators are collected to preliminarily identify the key risk factors and boundary risk factors of the river-type water source system. Based on the logical relationships of mutual feedback and influence between the risk assessment objectives of the river-type water source system and the risk factors of each risk factor and each subsystem (single water source), the causal loop diagram and stock-flow diagram of the river-type water source system are drawn. Simultaneously, based on the risk assessment indicators of each single water source... By assigning weights to determine the mathematical expressions of system equations such as state variable equations and rate equations, a system risk simulation and prediction model for river-type water source areas based on system dynamics is established. Based on the static risk assessment results of a single water source area, initial values ​​of boundary risk factors for the river-type water source area system and relevant parameters in each system equation are determined and continuously adjusted to ensure that the system risk simulation and prediction model for river-type water source areas based on system dynamics reflects the actual risk situation of the studied water source area to the greatest extent possible. Through system risk simulation, the system risk level for a future period is predicted and analyzed, sensitive risk factors of the river-type water source area system are identified, and corresponding countermeasures are proposed to actively prevent and control the risks of the water source area system.

[0121] Based on the above solution, this embodiment applies the solution in a practical way as follows:

[0122] This embodiment uses the Nanjing section of the Yangtze River as an example to illustrate the comprehensive risk assessment and prediction method for river-type water sources described in this invention, and demonstrates the effectiveness and rationality of this invention.

[0123] The Nanjing section of the Yangtze River mainly comprises six river-type water sources: Zihuizhou, Jiajiang South, Jiajiang North Estuary, Yanziji, Jiangpu, and Baguazhou (left branch) Shangba. Over 90% of the total domestic water consumption of residents in Nanjing's main urban area and suburbs comes from this section. Based on risk factor identification and analysis, and through indicator reduction and optimization, a risk assessment index system for river-type water sources was determined. A comprehensive risk assessment and prediction model was constructed to conduct static and dynamic risk assessments of single water sources and simulation predictions of the comprehensive risk level of multiple water source systems. This has significant theoretical and practical application value for improving water source risk management and ensuring the sustainable and healthy development of the economy and society.

[0124] Therefore, we first constructed a risk assessment index system for river-type water sources from four aspects: water quality, water quantity, ecology and management, following the principles of scientificity, practicality and foresight. The preliminary selected indicators for water pollution, water shortage, ecological environment and management safety risks are detailed in Tables 1 to 4.

[0125] Table 1 Water Pollution Risk Indicators

[0126]

[0127]

[0128] Table 2 Water Shortage Risk Indicators

[0129]

[0130] Table 3 Ecological and Environmental Risk Indicators

[0131]

[0132] Table 4 Management Safety Risk Indicators

[0133]

[0134] Through extensive literature review and expert consultation, the A. Gulin method was used for qualitative screening of indicators. Indicators with low importance and weak impact on the evaluation objectives were removed based on their importance ranking, thus achieving qualitative screening of evaluation indicators. On this basis, the various risk assessment indicators were rationally classified and merged to construct a hierarchical, logically clear three-tiered evaluation indicator system, as detailed below. Figure 3 .

[0135] Principal component analysis (PCA) was used to calculate the eigenvalues, eigenvector matrices, and factor loading matrices of the principal components. The factor loading matrix quantitatively represents the correlation between the original indicators and the principal components. It is obtained by rotating the factors using the maximum variance method, which can redistribute the relationship between the principal components and the original variables, making it easier to discover the magnitude and explanatory power of the original variables on each principal component. Based on the calculated eigenvalues ​​and eigenvectors, and following the principle that the cumulative contribution rate is greater than 85%, four principal components were selected. Indicators with factor loadings less than 0.4 in all four principal components were removed, namely, I2, I3, I12, I15, I18, I22, I26, and I28.

[0136] Based on the rationality test, the traces of the covariance matrices of the evaluation index data after qualitative and quantitative screening and the traces of the covariance matrices of the original index data were calculated. The information contribution rate In was calculated to be 90.26%, therefore, the constructed risk assessment index system for river-type water sources is considered relatively reasonable. Through the initial selection, optimization, and rationality test of the index system, a risk assessment index system for river-type water sources was finally established. The index system includes four criterion layers: water quality, water quantity, ecology, and management. The criterion layers are further subdivided into 26 indicators. The detailed framework of the risk assessment index system for river-type water sources is shown in Table 5.

[0137] Table 5 Risk Assessment Indicators and Their Numbers for River-type Water Source Areas

[0138]

[0139]

[0140] By conducting longitudinal comparative analysis of historical data of river-type water sources in the middle and lower reaches of the Yangtze River and horizontal comparative analysis of data of drinking water sources in other parts of the country, and combining national and local standards with expert opinions, the risk classification standards for each indicator were determined. The risk classification standards for water shortage assessment indicators are used as an example for illustration, as detailed in Table 6.

[0141] Table 6 Risk Level Classification of Water Shortage Assessment Indicators

[0142]

[0143] A static risk assessment study was conducted using the Jiangpu-Pukou water source area in the Nanjing section of the Yangtze River as an example.

[0144] Depend on Where j = 1, 2, 3, 4, 5, the membership matrix of the Jiangpu-Pukou water source criterion layer is obtained as follows:

[0145]

[0146] Based on the risk level assessment set U, the matrix V = (0.1, 0.3, 0.5, 0.7, 0.9) for each risk level is determined. Therefore, the comprehensive evaluation result of the criterion layer is:

[0147] B1 = C·V T ={0.3499 0.3526 0.4647 0.3557} T

[0148] Target layer evaluation results:

[0149] E1 = W JP ·B1 T =0.3778

[0150] Based on the risk assessment results of the Jiangpu-Pukou water source area, the risks in four aspects—water pollution, water shortage, ecological environment and management safety—were analyzed in the criterion layer of the evaluation index system. Using the same method, the static risk situation of the Zihuizhou, Jiajiang South, Jiajiang North, Yanziji, and Baguazhou (left branch) upstream dam water sources was also analyzed and evaluated. The static risk assessment results for six river-type water sources in the Nanjing section of the Yangtze River are shown in Table 7.

[0151] Table 7 Risk Assessment Results of Six River-Type Water Sources in the Nanjing Section of the Yangtze River

[0152]

[0153] Static risk assessment of a single water source area is mainly used to analyze the current status of risk assessment indicators, while dynamic assessment can take into account both the certainty and uncertainty of the assessment object, comprehensively considering its similarities, contradictions, and differences. It is mainly used to analyze the evolution trend of risk assessment indicators. Both are based on the same assessment period, with the former focusing on the current risk status and the latter focusing on the risk evolution trend. Based on the calculation results of the combined weights of the above assessment indicators and the risk membership matrix, the current status five-element correlation coefficient and its corresponding first to fourth-order partial correlation coefficients of each risk assessment indicator are calculated based on the partial correlation function. Then, based on the five-element correlation coefficient and the situation zoning table of each order of partial correlation coefficient, the zoning of the partial correlation coefficient of each risk assessment indicator is determined, thereby realizing the dynamic evolution trend analysis of system risk.

[0154] Similarly, this invention takes the Jiangpu-Pukou water source area as an example to conduct a dynamic risk assessment study of a single water source area. Based on the calculation results of the combined weights of each evaluation index in the risk assessment index system and the risk membership matrix, the current status 5-element connection coefficient and its first to fourth order partial connection coefficients of each risk assessment index of the Jiangpu-Pukou water source area are calculated based on the 5-element connection coefficient risk similarity and difference in reverse evaluation model, thereby analyzing its risk dynamic development trend. The specific calculation results are detailed in Table 8.

[0155] Table 8. Current Status of Risk Assessment Indicators in Jiangpu-Pukou Water Source Area: Calculation Results of Five-Element Connection Numbers and Partial Connection Numbers

[0156]

[0157]

[0158] Table 8 (Continued) shows the calculation results of the current status of the five-element correlation coefficients and their partial correlation coefficients for various risk assessment indicators of the Jiangpu-Pukou water source area.

[0159]

[0160]

[0161] Similarly, the same method was used to analyze and evaluate the dynamic risk trends of the water source areas at Zihuizhou, Jiajiang South, Jiajiang North, Yanziji, and Baguazhou (left branch) in the Nanjing section of the Yangtze River. The results of the dynamic risk evaluation of the six river-type water source areas in the Nanjing section of the Yangtze River are shown in Table 9.

[0162] Table 9. Risk Dynamic Assessment Results of Six River-Type Water Sources in the Nanjing Section of the Yangtze River

[0163]

[0164] After identifying the key risk factors (risk assessment indicators) for six river-type water sources in the Nanjing section of the Yangtze River and clarifying the logical relationships between them, a system dynamic risk causal loop diagram and a risk flow stock diagram were established for the six water sources in the Nanjing section of the Yangtze River. Furthermore, each risk factor was connected to the system equations as a whole through the risk stock flow diagram. The system risk causal loop diagram for the six river-type water sources in the Nanjing section of the Yangtze River is shown below. Figure 4 The causal tree diagram of the comprehensive risk of the multi-river water source system in the Nanjing section of the Yangtze River is shown below. Figure 5 .

[0165] The six river-type water sources along the Nanjing section of the Yangtze River are considered as a whole system, divided into six subsystems: two on the north bank and four on the south bank. In analyzing the impact relationships between these subsystems, the principle of "upstream influencing downstream" is followed. The Jiangpu-Pukou (JP) water source on the north bank and the Zihuizhou (ZHZ) water source on the south bank are located upstream of the drinking water sources on both banks of the Yangtze River in Nanjing; therefore, the impact of the other four water sources on these two sources is not considered. The ZHZ water source is located upstream of the Jiajiang South (JS) water source; therefore, the risks of water pollution, water shortage, ecological environment, and management safety at the ZHZ water source are considered. Risks affect the JS water source area as risk factors, and similarly, the JS water source area affects the Jiajiang North (JN) water source area in the same way. Since the Baguazhou (BGZ) (left branch) dam water source area and the Yanziji (YZJ) water source area are located in the Baguazhou left branch and the Baguazhou main river section respectively, and the Yangtze River Jiajiang and main river sections diverge after converging, flowing towards the Baguazhou left branch and the main river section respectively, the BGZ and YZJ water sources are simultaneously affected by the JP and JN water sources. That is, the risk factors of the four criterion layers of the JP and JN water sources affect the BGZ (left branch) dam water source area and the YZJ water source area as risk factors. See the risk causal tree diagram for the Jiangpu-Pukou water source area. Figure 6 See (a)-(d) for the causal tree diagram of the risk of the Yanziji water source area. Figure 7 (a)-(d), Risk Stock Flow Map of a Single Water Source System (Taking Zihuizhou as an Example) Figure 8 See the risk stock and flow map of the multiple water source systems in the Nanjing section of the Yangtze River. Figure 9 The initial values ​​of risk factors at the boundaries of the six river-type water sources in the Nanjing section of the Yangtze River are shown in Table 10 (taking Yanziji water source as an example). The system equations were determined based on the structure of each variable and the causal feedback logic in the system dynamics model. The system equations for Yanziji water source are described in Table 11.

[0166] Table 10 Weights of Risk Factor Combinations at Each Boundary Level of the Yanziji Water Source Area

[0167]

[0168]

[0169] Table 11 Description of the Yanziji Water Source System Equations

[0170]

[0171]

[0172] The equations for each individual water source subsystem were entered into the Vensim_PLE software using the equation editor. After debugging and running the simulation, the results of the comprehensive risk assessment for six river-type water sources along the Nanjing section of the Yangtze River were observed. The model's time unit is monthly. Considering the simulation and prediction of quantitative risk changes over 36 months after 2018, the simulation end time was set to 36 months. Since the risk prevention and management mechanisms for water quality, water quantity, and the ecological environment of the Yangtze River's Nanjing section water sources will tend to improve over time, and the risk changes in the water sources will tend to stabilize, simulation results beyond 3 years are not considered for the time being.

[0173] The comprehensive risk simulation results for four river-type water sources on the south bank of the Yangtze River in Nanjing (Zihuizhou, Jiajiang South, Jiajiang North Estuary, and Yanziji) from 2019 to 2021 are shown below. Figure 10 (a) to (d), the comprehensive risk assessment values ​​for the Zihuizhou, Jiajiang South, and Yanziji water sources in 2021 are all below 0.2, indicating a low-risk level. However, the Jiajiang North Estuary water source, due to its large water supply scale, large population served, and the location of its protection zone within the main urban area, along with diverse land use types and a complex structure of potential risk sources, has a risk assessment value slightly above 0.2 in 2021, placing it at a relatively low-risk level. With the continuous upgrading of comprehensive risk prevention and control measures for water sources along the Nanjing section of the Yangtze River, the risk assessment value of the Jiajiang North Estuary water source shows a downward trend. Based on the trend of risk assessment value changes, it can be predicted that the risk assessment value of the Jiajiang North Estuary water source will also be below 0.2 for some time to come, placing it at a low-risk level. The comprehensive risk simulation results for the two river-type water sources on the north bank of the Nanjing section of the Yangtze River from 2019 to 2021 are shown below. Figure 10 (e) and Figure 10 (f) Since the Jiangpu water source of the Yangtze River already has the Sancha Reservoir in Pukou District as a backup water source within its water supply range, and with the completion of the extension and laying of the water supply network in the past two years, the risk assessment value of the Jiangpu water source can be reduced, keeping it at a low risk level. However, the Baguazhou (left branch) Shangba water source is located in the northeast of the main urban area, in a relatively remote geographical location, and is far from the Sancha Reservoir, Yangku Reservoir, and Xinjizhou Fenghuang Lake (two reservoirs and one lake) emergency backup water sources that are under construction. Therefore, although the risk assessment value of the Baguazhou (left branch) Shangba water source is showing a downward trend, it is still at a low risk level from 2019 to 2021.

[0174] The results of the integrated risk simulation of the multi-source water system in the Nanjing section of the Yangtze River are shown below. Figure 10(g) By 2021, the comprehensive risk assessment value of the six river-type water source systems in the Nanjing section of the Yangtze River will be at the critical value between low risk and relatively low risk. With the completion of the construction of the "two reservoirs and one lake" (Yangku Reservoir in Jiangning District, Sancha Reservoir in Pukou District, and Phoenix Lake in Xinjizhou) emergency backup water source, Nanjing will be built into a "large water tank", and the main urban area, Jiangning and Jiangbei New Area will be able to achieve dual water source supply and have backup water at any time. This will significantly improve the ability of river-type water sources in the Nanjing section of the Yangtze River to cope with sudden water source crises, and keep the comprehensive risk assessment level at a low level.

[0175] Based on the same inventive concept, the present invention provides a comprehensive risk assessment and prediction system for river-type water source areas, comprising:

[0176] The indicator system construction module is used to construct a risk assessment indicator system for river-type water sources. First, the risk assessment indicator system is initially selected from four aspects: water quality, water quantity, ecology and management of river-type water sources. Second, the evaluation indicators are optimized through qualitative and quantitative screening to establish a risk assessment indicator system for river-type water sources. Then, the connotation of each evaluation indicator is defined, and the risk threshold of the corresponding indicator is reasonably determined.

[0177] The static risk situation assessment module is used to assess the static risk status of a single water source. First, the subjective and objective comprehensive weights of each evaluation indicator are determined using the entropy weight-analysis combined weighting method. Second, a gray fuzzy comprehensive evaluation model of the static risk situation of a single water source based on combined weighting is established based on grey theory and fuzzy comprehensive evaluation theory to achieve the static risk status assessment of a single water source.

[0178] The dynamic risk trend analysis module is used to analyze the dynamic risk trends of a single water source: First, based on set pair analysis and the theory of risk similarity and difference in reverse evaluation, the multivariate correlation coefficients of each risk evaluation indicator and their corresponding partial correlation coefficients are calculated; based on the risk correlation coefficient trend table, the risk development trend of each risk evaluation indicator in the same, equal, or opposite state is clarified, thereby realizing the evaluation of the dynamic risk trend of a single water source.

[0179] The multi-source water system risk prediction module is used for comprehensive risk simulation and prediction of multi-source water systems. It identifies the research object and objective of the river-type water source system, establishes a risk simulation and prediction model of the river-type water source system based on system dynamics, combines the results of the importance analysis of each individual water source in the multi-source water system with the quantitative analysis of the influence between water sources, proposes the hypothesis of dynamic risk change, and draws the causal relationship diagram and risk stock flow diagram of the system risk simulation model, so as to realize the comprehensive risk simulation and prediction of multi-source water systems based on system dynamics.

[0180] Based on the same inventive concept, the present invention provides an electronic device comprising a memory and a processor, wherein:

[0181] Memory is used to store computer programs that can run on a processor;

[0182] The processor is used to execute the steps of the above-described method for comprehensive risk assessment and prediction of river-type water source areas when running the computer program.

[0183] Based on the same inventive concept, the present invention provides a storage medium storing a computer program, which, when executed by at least one processor, implements the steps of the above-described method for comprehensive risk assessment and prediction of river-type water source areas.

Claims

1. A comprehensive risk assessment and prediction method for river-type water source areas, characterized in that, Includes the following steps: S1. Construction of Risk Assessment Index System for River-type Water Sources: First, based on the pressure-state-response structural model, various factors affecting the risk status of water sources are analyzed. Second, a preliminary selection of risk assessment indexes is conducted from four aspects: water quality, water quantity, ecology, and management of river-type water sources. Then, after qualitative screening using the A. Gulin method, data standardization is performed, a correlation coefficient matrix is ​​constructed, and the matrix eigenvalues ​​and eigenvectors are calculated to obtain principal components with a cumulative contribution rate greater than a threshold. By calculating the factor loading matrix and principal component contribution rate matrix of each principal component, the correlation between each original index and the principal component and its contribution rate to the principal component are determined. Original indicators that are independent of each other and can objectively represent the evaluation object are obtained, thus completing the quantitative screening of indicators. Then, a reasonableness test is conducted to establish a risk assessment index system for river-type water sources. Next, the connotation of each evaluation index is defined, and the risk threshold of the corresponding index is reasonably determined. S2. Static Risk Status Assessment of a Single Water Source: First, the comprehensive weight of each evaluation indicator is determined using the entropy weight-analysis-level combined weighting method. Second, based on grey theory and fuzzy comprehensive evaluation theory, a grey fuzzy comprehensive evaluation model for the static risk status of a single water source is established, including: determining the risk level, forming a sample evaluation matrix, whitening analysis of evaluation indicators, calculation of grey statistics, determining the membership matrix, first-level fuzzy comprehensive risk evaluation and multi-level fuzzy comprehensive risk evaluation, to achieve static risk status assessment of a single water source. S3. Dynamic Risk Trend Analysis of Single Water Source Area: Based on the comprehensive weight and membership matrix of each evaluation indicator determined in step S2, and according to set pair analysis and risk similarity and difference inverse evaluation theory, calculate the five-element connection number of each risk evaluation indicator and its corresponding partial connection number; and clarify the risk development trend of each risk evaluation indicator in the same, equal, or opposite state according to the risk connection number trend table, thereby realizing the evaluation of the dynamic risk trend of a single water source area. S4. Comprehensive Risk Simulation and Prediction of Multi-Source Water System: Determine the research object and objective of the river-type water source system, establish a risk simulation and prediction model of the river-type water source system based on system dynamics, combine the importance analysis results of each individual water source in the multi-source water system with the quantitative analysis of the influence between water sources, propose the assumption of dynamic risk change, and draw the causal relationship diagram and risk stock flow diagram of the system risk simulation model. Use the static risk status evaluation results of each individual water source in step S2 as boundary conditions to realize the comprehensive risk simulation and prediction of the multi-source water system based on system dynamics.

2. The method for comprehensive risk assessment and prediction of river-type water source areas according to claim 1, characterized in that, In step S1, the A. Gulin method removes redundant indicators based on importance ranking, thereby achieving qualitative screening of evaluation indicators. In step S1, the principal component analysis method further performs quantitative screening of evaluation indicators based on linear correlation and independence.

3. The method for comprehensive risk assessment and prediction of river-type water source areas according to claim 1, characterized in that, The basic idea of ​​the rationality test in step S1 is to calculate the percentage of information contained in the evaluation indicators after quantitative screening relative to the information content of the indicator system before quantitative screening. If this percentage meets the predetermined standard value, it means that the evaluation indicator system after screening can pass the rationality test; otherwise, the evaluation indicator system needs to be readjusted and screened. The rationality test mainly includes the following two basic steps: (1) Assuming the number of evaluation indicators before quantitative screening is m, and the number of evaluation indicators is reduced to n after quantitative screening, then the information contribution rate of the evaluation indicator system after quantitative screening is In: Among them, trS m With trS n These represent the traces of the covariance matrices of the evaluation indicators before and after quantitative screening, respectively. (2) If In≥90%, the evaluation index system after screening is relatively reasonable and can pass the rationality test; otherwise, the evaluation index system needs to be adjusted or redesigned.

4. The method for comprehensive risk assessment and prediction of river-type water source areas according to claim 1, characterized in that, In step S1, based on the characteristics of the water source risk and actual business needs, the risk measurement and evaluation standards are divided into five levels: low, relatively low, average, relatively high, and high. To facilitate quantitative calculation, a risk level evaluation set is introduced to determine the corresponding risk level. R = {Low risk, Lower risk, Moderate risk, Higher risk, High risk} ＝{(0，0.2]，(0.2，0.4]，(0.4，0.6]，(0.6，0.8]，(0.8，1]}。 5. The method for comprehensive risk assessment and prediction of river-type water source areas according to claim 1, characterized in that, Step S3, calculating the multivariate correlation coefficient and its corresponding partial correlation coefficient for each risk assessment indicator, specifically involves determining the corresponding multivariate correlation coefficient based on the risk assessment level of each indicator. The multivariate correlation coefficient refers to the multidimensional development trend of the correlation coefficient difference measurement components, and its general expression form is: u = a + b1i1 + b2i2 + ... + b n-2 i n-2 +cj, where u is called the n-ary connection number, and its first-order partial connection number is: in, Its second-order partial correlation coefficient is: in, And so on, its (n-1)th order partial correlation coefficient is: in, In the formula for the first-order partial correlation coefficient, based on the idea of ​​the dynamic development of each risk assessment indicator system, the definite term 'a' in the n-ary correlation coefficient u was originally located at the level of the uncertain term in the difference measure component b1, and it was obtained from b1 through development and evolution. Used to quantitatively characterize the degree of dynamic development of this evolutionary trend; b1 in u was originally at the level of b2, and it evolved from b2. This is used to quantitatively characterize the degree of dynamic development of this trend; and so on; similarly, b in u... n-2 Originally, it was also at the C level, having evolved from C. Used to quantitatively characterize the degree of dynamic development of this trend; therefore, partial correlation coefficients Based on a positive development trend, the n-ary connection number u = a + b1i1 + b2i2 + ... + b is characterized from a holistic perspective. n-2 i n-2 The evolution of the same / different / anti-definite / uncertain relationship states in +cj. The first-order partial correlation coefficient represents the n-ary correlation coefficient based on a positive development trend.

6. The method for comprehensive risk assessment and prediction of river-type water source areas according to claim 1, characterized in that, The specific process for determining the risk development trend of each evaluation indicator in step S3 is as follows: when the opposing measure component c in the n-ary connection number is not 0, the ratio of the same measure component a to the opposing measure component c is defined as the connection potential or set pair potential of the evaluated set pair under a specific research background, denoted as shi(H)=a / c. When a / c > 1, the relationship potential is in a state of same potential under the specific research background, indicating that the two sets being evaluated have the same development trend in their similarity and opposition relationships. When a / c = 1, the relationship potential is in a state of equilibrium under the specific research background, indicating that the two sets being evaluated have a "balanced" state in terms of their similarity and opposition trends. When a / c < 1, the relationship potential is in a state of opposition under the specific research background, indicating that the two sets being evaluated have opposing development trends. By using the relationship potential or set-pair potential, the risk development trend of the system can be divided into similar potential, equilibrium potential, and opposition potential. Furthermore, based on the magnitude of each coefficient, the development trend of the system can be ranked. In the study on the comprehensive risk assessment of river-type water source areas, when the risk levels of the evaluation indicators are in the same potential zone, it indicates that the actual risk situation of the evaluated object and the reference set have the same trend, that is, the evaluated object is at a "low" risk level. By analyzing the trend of the evaluation indicators and their partial correlation coefficients, we can find evaluation indicators with the first-order partial correlation coefficients in the opposite trend and give them special attention. When the risk level of the evaluation indicator is in the equilibrium zone, it means that the actual risk situation of the evaluated object and the reference set are in a state of "equal strength", that is, the evaluated object is at the "medium" risk level. In this case, close attention should be paid to its correlation coefficient and its first-order partial correlation coefficient. For evaluation indicators with the first-order partial correlation coefficient in the opposite direction, the risk status of the corresponding evaluated object should be made to evolve in the same direction, and the risk level should eventually be transformed into the same zone. When the risk level of the evaluation indicator is in the opposite zone, it indicates that the actual risk situation of the evaluated object is in opposition to the reference set, that is, the evaluated object is at a "high" risk level. For evaluation indicators whose current risk level is in the negative trend zone and whose first-order partial correlation coefficient is in the negative trend zone, the risk status of sensitive evaluation indicators should be transformed into the equilibrium zone, and the risk level should be reversed to the equal trend zone.

7. The method for comprehensive risk assessment and prediction of river-type water source areas according to claim 1, characterized in that, The specific process of drawing the causal relationship diagram and risk stock flow diagram of the system risk simulation model in step S4 is as follows: through the analysis of the river-type water source system, the system risk assessment objectives, assessment scope and assessment period are clarified, various risk factors affecting the safety of the water source system are identified, and various data related to the quantitative analysis of the evaluation indicators are collected to preliminarily clarify the key risk factors and boundary risk factors of the river-type water source system. Based on the risk assessment objectives of river-type water source systems and the logical relationships of mutual feedback and influence among various risk factors and individual water source risk factors, a causal loop diagram and a stock-flow diagram of the river-type water source system are drawn. Simultaneously, based on the weights of the risk assessment indicators for each individual water source, the system risk increment equation and system risk state variable equation are determined, thereby establishing a river-type water source system risk simulation and prediction model based on system dynamics. Based on the static risk assessment results of individual water sources, the initial values ​​of boundary risk factors of the river-type water source system and relevant parameters in each system equation are determined and continuously adjusted to ensure that the river-type water source system risk simulation and prediction model based on system dynamics reflects the actual risk situation of the studied water source to the greatest extent. Through system risk simulation, the system risk level for a future period is predicted and analyzed, sensitive risk factors of the river-type water source system are identified, and corresponding countermeasures are proposed to actively prevent and control the risks of the water source system.

8. A comprehensive risk assessment and prediction system for river-type water source areas, characterized in that, include: The indicator system construction module is used to construct a risk assessment indicator system for river-type water sources. First, the risk assessment indicator system is initially selected from four aspects: water quality, water quantity, ecology and management of river-type water sources. Second, the evaluation indicators are optimized through qualitative and quantitative screening to establish a risk assessment indicator system for river-type water sources. Then, the connotation of each evaluation indicator is defined, and the risk threshold of the corresponding indicator is reasonably determined. The static risk situation assessment module is used to assess the static risk status of a single water source. First, the subjective and objective comprehensive weights of each evaluation indicator are determined using the entropy weight-analysis combined weighting method. Second, a gray fuzzy comprehensive evaluation model of the static risk situation of a single water source based on combined weighting is established based on grey theory and fuzzy comprehensive evaluation theory to achieve the static risk status assessment of a single water source. The dynamic risk trend analysis module is used to analyze the dynamic risk trends of a single water source: First, based on set pair analysis and the theory of risk similarity and difference in reverse evaluation, the multivariate correlation coefficients of each risk evaluation indicator and their corresponding partial correlation coefficients are calculated; based on the risk correlation coefficient trend table, the risk development trend of each risk evaluation indicator in the same, equal, or opposite state is clarified, thereby realizing the evaluation of the dynamic risk trend of a single water source. The multi-source water system risk prediction module is used for comprehensive risk simulation and prediction of multi-source water systems. It identifies the research object and objective of the river-type water source system, establishes a risk simulation and prediction model of the river-type water source system based on system dynamics, combines the results of the importance analysis of each individual water source in the multi-source water system with the quantitative analysis of the influence between water sources, proposes the hypothesis of dynamic risk change, and draws the causal relationship diagram and risk stock flow diagram of the system risk simulation model, so as to realize the comprehensive risk simulation and prediction of multi-source water systems based on system dynamics.

9. An electronic device, characterized in that, Includes memory and processor, wherein: Memory is used to store computer programs that can run on a processor; A processor, configured to, while running the computer program, perform the steps of the method for comprehensive risk assessment and prediction of river-type water source areas as described in any one of claims 1-7.

10. A storage medium, characterized in that, The storage medium stores a computer program, which, when executed by at least one processor, implements the steps of the comprehensive risk assessment and prediction method for river-type water source areas as described in any one of claims 1-7.

Images