Evaluation method and system for agricultural water resource safety guarantee
Through the correlation weight entropy weight method and the fuzzy comprehensive evaluation method, the problem of neglected correlation between indicators in agricultural water resource safety evaluation is solved, and a more accurate water resource safety evaluation is achieved. The correlation and variation between indicators is considered, and the improvement of improvement The objectivity and reliability of the evaluation results are carried out.
Patent Information
- Application Number
- CN202411035323.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-31
- Publication Date
- 2024-11-08
AI Technical Summary
The existing agricultural water resource safety evaluation methods ignore the correlation between indicators, resulting in inaccurate evaluation results, inconsistent evaluation standards and scales, which are greatly affected by subjectiveness.
By using the correlation weight entropy weight method and fuzzy comprehensive evaluation method, the initial evaluation index system is constructed, the index numerical value is normalized and the correlation analysis is performed, the index weight is calculated, and the information sensitivity analysis is performed based on the improved principal components. Finally, the Fuzzy comprehensive evaluation was performed on the triangular membership function and the weighted average synthesis operator.
A more accurate and objective assessment of agricultural water resources safety was achieved, taking into account the correlation and variation between indicators, and improving the accuracy and reliability of evaluation results.
Smart Images

Figure CN118917543A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of agricultural water resource evaluation methods, and relates to an evaluation method for ensuring agricultural water resource security. The present invention also relates to an evaluation system for ensuring agricultural water resource security. Background Art
[0002] Water resource security has always been a highly valued issue. The rational use of agricultural water resources is the key to ensuring the sustainable development of agriculture. Therefore, it is necessary to carry out research on agricultural water resource security evaluation. By evaluating the security status of agricultural water resources, it provides theoretical basis and technical support for managers to propose targeted protection measures.
[0003] In current research, water resource security evaluation in the agricultural sector often ignores the interaction between influencing factors, is greatly influenced by subjective factors, and has inconsistent evaluation standards and scales, making it difficult to accurately evaluate the security status of agricultural water resources. The indicator weight determination method often uses a combination of hierarchical analysis and entropy weight method, taking subjective and objective influences into consideration. However, the entropy weight method assumes that there is no correlation between indicators, while in reality there is correlation between indicators, which will affect the evaluation results. Summary of the invention
[0004] The purpose of the present invention is to provide an evaluation method for agricultural water resources security, which solves the problem in the prior art that the correlation between indicators is not considered, thereby affecting the evaluation results.
[0005] The technical solution adopted by the present invention is an evaluation method for agricultural water resources security, which is specifically implemented according to the following steps:
[0006] Step 1: Construct an evaluation index system, including indicators, indicator values and evaluation criteria;
[0007] Step 2, using the correlation weight entropy weight method to calculate the indicator weight;
[0008] Step 3: Based on the indicator weights, the fuzzy comprehensive evaluation method is used to evaluate the agricultural water resources security status in the evaluation area.
[0009] The present invention is also characterized in that:
[0010] Step 1 is as follows:
[0011] Step 1.1: Build an initial evaluation index system
[0012] The frequency analysis method was used to select indicators with a frequency of ≥5 in the relevant literature in the research area in the past five years, and the selected indicators were divided into three criteria levels: water resources security, social and economic security, and ecological and environmental security according to the characteristics of the indicators, thus obtaining the initial evaluation indicator system;
[0013] Step 1.2, according to the specific situation of the study area, obtain the corresponding indicator value. Each indicator data includes the data of the past N years in units of years, and reselect the indicators selected in step 1.1 according to the acquisition of the indicator value to obtain a reselected evaluation indicator system. Assuming that the evaluation indicator system includes K indicators in total, the indicator data for the kth indicator is: {x 1k ,x 2k ,…,x Nk}, where x Nk It represents the indicator value of the kth indicator in the Nth year, k = 1, 2, 3, ..., K. The data corresponding to all indicators in each year is a sample data, that is, it includes N sample data. For the i-th sample data: {x i1 ,x i2 ,…,x iK}, where x iK Indicates the indicator value corresponding to the Kth indicator in the i-th year, i = 1, 2, 3, ..., N;
[0014] Step 1.3, normalize the selected indicators. For the kth indicator, the normalized indicator data is: 11 ,a 12 ,...,a NK}, where a NK It is the normalized value of the Kth indicator in the Nth year;
[0015] Step 1.4, for the normalized indicators, the correlation analysis method is used to calculate the correlation coefficients between the two indicators in each criterion layer. If the absolute value of the correlation coefficient between two indicators is greater than 0.8, any one of the corresponding two indicators is eliminated. Then, the indicator information sensitivity analysis is performed based on the improved principal component. The information sensitivity of all indicators eliminated by the correlation analysis method is calculated, and they are sorted from large to small according to information sensitivity. Indicators with cumulative information sensitivity ≥ 70% are retained, and each indicator is divided into 5 levels according to the indicator value, namely: unsafe, less safe, basically safe, relatively safe and very safe, to obtain the final evaluation indicator system;
[0016] Step 1.3 is as follows:
[0017] For each indicator data, the standard deviation standardization method is first used to eliminate the influence of outliers on the indicator system, and then the range standardization method is used to eliminate the influence of variable dimensions and variation range, and the data is scaled so that it falls into [0,1];
[0018] Among them, the standard deviation standardization formula is:
[0019]
[0020] Among them, x ik is the index value of the kth index in the multiple-selection index system in the i-th year, i = 1, 2, 3, ..., N, k = 1, 2, 3, ..., K, x i ' k To x ik The data after standardization is is the mean of the original data of the k-th indicator in N years; σ is the standard deviation of the k-th indicator;
[0021] The range standardization method processing formula is:
[0022] If the indicator has a driving effect on the criterion layer, that is, the larger the corresponding indicator value, the safer the criterion layer, then it is considered that the indicator has a driving effect on the criterion layer. At this time, the indicator attribute is positive, and it is processed according to the following formula:
[0023]
[0024] If the indicator has a restrictive effect on the criterion layer, that is, the larger the corresponding indicator value, the less secure the criterion layer, then the indicator is considered to have a restrictive effect on the criterion layer. At this time, the indicator attribute is negative, and it is processed according to the following formula:
[0025]
[0026] Among them, max(x i ' k ) is to convert x ik The maximum value of the data after standard deviation standardization, min(x i ' k ) is to convert x ik The minimum value of the data after standardization, a ik It is the indicator value of the kth indicator in the multiple-selection indicator system in the i-th year after being standardized by the range.
[0027] In step 1.4, the correlation analysis method is used to calculate the correlation coefficients between the two indicators in each criterion layer:
[0028]
[0029] Among them, r is the correlation coefficient matrix between indicators, r kk′ is the correlation coefficient of the kth and k′th indicators, a ik is the value of the kth indicator in the i-th year after normalization, a ik′ is the value of the k′th indicator in the i-th year after normalization, is the average value of the kth indicator in N years after normalization, is the average value of the k′th indicator in N years after normalization, k, k′=1,2,…,K, and the measurement standard of the correlation coefficient is: 0<|r kk′ |≤0.3, irrelevant; 0.3<|r kk′ |≤0.8, weak correlation; 0.8<|r kk′ |≤1, strong correlation;
[0030] The calculation formula for indicator information sensitivity analysis is as follows:
[0031]
[0032] r T r-λ k E K |=0
[0033]
[0034] Among them, τ m is the cumulative information sensitivity of the first m indicators; is the tth principal component Z t For the kth index X k The partial derivative of kt is the load coefficient of the kth index on the tth principal component; k is the eigenvalue of the kth index; e kt is the eigenvector corresponding to the eigenvalue of the kth index; η k is the variance contribution rate of the kth indicator; β k is the information sensitivity of the kth indicator; is the information sensitivity of the kth indicator calculated according to the ranking; r is the correlation coefficient matrix between indicators, E K is the K-order unit matrix, Ω t is the cumulative variance contribution rate of the first t indicators, Ω t When the sensitivity is ≥85%, the first t indicators are retained as principal components, t is the number of principal components (t∈[1,K]), and m is the number of indicators when the cumulative information sensitivity is ≥70% (m∈[1,K]).
[0035] Step 2 is as follows:
[0036] For the entropy weight ω of the vth index v , which is the weight of the vth indicator, where v = 1, 2, ..., m, and is calculated according to the following formula:
[0037]
[0038] Among them, e v is the entropy value of the vth indicator, and m is the number of indicators in the final evaluation system.
[0039] The entropy value e of the vth indicator v Calculate according to the following formula:
[0040]
[0041]
[0042] X0=x ov
[0043] Among them, f iv is the proportion of the normalized value of the vth indicator in the i-th year to the normalized total value of the vth indicator in the N-th year; N is the number of samples, N = 12; X0 is the reference sequence, if it is a positive indicator, x 0v is the maximum value of the original data of the same indicator v in different sample data. If it is a negative indicator, x 0v is the minimum value of the original data of the same indicator v in different sample data; r 0v is the vth index a in the reference sequence 0v and the vth indicator a in the i-th year iv The correlation between 0v is the vth index x in the reference sequence 0v Normalized value; a iv is the normalized value of the vth indicator in the final indicator system in the i-th year; r v is the correlation between the reference sequence index and the vth index; ξ 0i (v) is the difference between the vth indicator in the i-th year and the vth indicator in the reference sequence a 0v The correlation coefficient between them; ρ is the resolution coefficient, which is taken as 0.5.
[0044] Step 3 is as follows:
[0045] Step 3.1, establish the evaluation factor set U and the comment set V;
[0046] Step 3.2, using the triangular membership function to determine the fuzzy relationship matrix R of each sample from the evaluation factor set U to the comment set V;
[0047] Step 3.3: Use weighted average synthesis operator to perform fuzzy synthesis to obtain the fuzzy vector B of each criterion layer. g (g = 1, 2, 3), combined with the evaluation level division, the maximum membership principle is adopted, and the fuzzy vector B is taken g The evaluation level where the maximum value of all components in is taken as the evaluation result of the criterion layer, and the numerical vector V of the evaluation level is g With the fuzzy vector B g Multiply them together to get the score F of the evaluation area on the evaluation target.
[0048] The evaluation factor set U in step 3.1 is composed of the indicators in the final indicator system, that is, U = {U1, U2, U3, ..., Um}, where U1, U2, U3, ..., Um represent m evaluation indicators respectively;
[0049] The comment set is a collection of evaluation results, which are represented by different levels and values. Five levels are selected: unsafe, less safe, basically safe, relatively safe, and very safe. V = {unsafe, less safe, basically safe, relatively safe, very safe} = {1, 2, 3, 4, 5}.
[0050] Step 3.2 is as follows:
[0051] The fuzzy relationship matrix R is:
[0052]
[0053] If the indicator attribute is positive:
[0054]
[0055] If the indicator attribute is negative:
[0056]
[0057] Among them, r vj is the membership degree of the vth indicator at the jth level, v = 1, 2, ..., m; j is the jth level in the comment set, j = 1, 2, 3, 4, 5; x1, x3, x5, x7 are the division nodes of the evaluation level data interval except the two end points, x2, x4, x6 are the midpoints of the level intervals x1 ~ x3, x3 ~ x5, x5 ~ x7 respectively; x is the value of the evaluation sample on each indicator, q is the number of levels, q = 5;
[0058] According to the calculation formula of R, the fuzzy relationship matrix is divided into three matrices R1, R2, and R3 according to the criterion layer;
[0059]
[0060]
[0061] Among them, R1, R2, and R3 represent the fuzzy relationship matrices of the three criterion layers of water resources security, social and economic security, and ecological and environmental security, respectively. c=m, a is the number of indicators in the water resources security criterion layer, ba is the number of indicators in the social and economic security criterion layer, and cba is the number of indicators in the ecological and environmental security criterion layer.
[0062] In step 3.3:
[0063] The fuzzy vector B of each criterion layerg Calculate according to the following formula:
[0064] B g =A g ·R g
[0065] Among them, when g = 1, 2, and 3, they represent three criterion levels, that is, g is the water resources security or social and economic security or ecological and environmental security level;
[0066] A g is the set of all indicator weights in the corresponding criterion layer; R g is the fuzzy relation matrix of the corresponding criterion layer; A1={ω1,ω2,...,ω a},A2={ω a+1 ,ω a+2 ,...,ω b},A3={ω b+1 ,ω b+2 ,...,ω c};
[0067] The calculation formula of the score F of the evaluation area on the evaluation target is as follows:
[0068]
[0069] Where: V g is the numerical vector V of the evaluation level g =(1,2,3,4,5).
[0070] The second technical solution adopted by the present invention is: an evaluation system for agricultural water resources security, used to implement an evaluation method for agricultural water resources security, including a data acquisition and transmission module, a data storage module, an algorithm model module and a client connected in sequence, the data acquisition and transmission module is used to collect index data and transmit the data to the storage module, and the data acquisition and transmission module collects index data by manual entry and Web page data source crawling to realize data collection;
[0071] The data storage module is used to store the index data collected by the data collection and transmission module and the intermediate data calculated by the algorithm model module;
[0072] The algorithm model module is used to realize normalization processing, weight assignment, and use fuzzy comprehensive evaluation method to evaluate the agricultural water resources security status of the evaluation area according to the indicator weight;
[0073] The client is used for visualization of agricultural water resources security assessment results and for use by managers and grassroots users, including web and mobile terminals.
[0074] The beneficial effects of the present invention are:
[0075] On the basis of the subjective selection of index preliminary and re-selection, the present invention further optimizes the index system from the objective perspective of improving the sensitivity analysis of index information of the principal component, so as to determine a more accurate and complete agricultural water resources security evaluation system. At the same time, the index weights are assigned by the correlation weight entropy weight method, which not only considers the variation degree of the index values, but also considers the correlation degree between the indicators, and can objectively and truly reflect the importance of the indicators. Finally, the fuzzy comprehensive evaluation method is used for evaluation, which takes into account fuzzy phenomena and fuzzy concepts such as water resources security level, so as to make the evaluation results more accurate. BRIEF DESCRIPTION OF THE DRAWINGS
[0076] Figure 1 It is a flow chart of the evaluation method for agricultural water resources security of the present invention. DETAILED DESCRIPTION
[0077] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.
[0078] Example 1
[0079] The evaluation method used for agricultural water resources security is as follows: Figure 1 As shown, the specific steps are as follows:
[0080] Step 1: Construct an evaluation index system, including indicators, indicator values and evaluation criteria; specifically:
[0081] Step 1.1: Build an initial evaluation index system
[0082] The frequency analysis method was used to select indicators with a frequency of ≥5 in the relevant literature in the research area in the past five years, and the selected indicators were divided into three criteria levels: water resources security, social and economic security, and ecological and environmental security according to the characteristics of the indicators, thus obtaining the initial evaluation indicator system;
[0083] Step 1.2, according to the specific situation of the study area, obtain the corresponding indicator value. Each indicator data includes the data of the past N years in units of years, and reselect the indicators selected in step 1.1 according to the acquisition of the indicator value to obtain a reselected evaluation indicator system. Assuming that the evaluation indicator system includes K indicators in total, the indicator data for the kth indicator is: {x 1k ,x 2k ,...,x Nk}, where x Nk It represents the indicator value of the kth indicator in the Nth year, k = 1, 2, 3, ..., K. The data corresponding to all indicators in each year is a sample data, that is, it includes N sample data. For the i-th sample data: {x i1 ,x i2,...,x iK}, where x iK Indicates the indicator value corresponding to the Kth indicator in the i-th year, i = 1, 2, 3, ..., N;
[0084] Step 1.3, normalize the selected indicators. For the kth indicator, the normalized indicator data is: 11 ,a 12 ,…,a NK}, where a NK It is the normalized value of the Kth indicator in the Nth year; specifically:
[0085] For each indicator data, the standard deviation standardization method is first used to eliminate the influence of outliers on the indicator system, and then the range standardization method is used to eliminate the influence of variable dimensions and variation range, and the data is scaled so that it falls into [0,1];
[0086] Among them, the standard deviation standardization formula is:
[0087]
[0088] Among them, x ik is the index value of the kth index in the multiple-selection index system in the i-th year, i = 1, 2, 3, ..., N, k = 1, 2, 3, ..., K, x i ' k To x ik The data after standardization is is the mean of the original data of the k-th indicator in N years; σ is the standard deviation of the k-th indicator;
[0089] The range standardization method processing formula is:
[0090] If the indicator has a driving effect on the criterion layer, that is, the larger the corresponding indicator value, the safer the criterion layer, then it is considered that the indicator has a driving effect on the criterion layer. At this time, the indicator attribute is positive, and it is processed according to the following formula:
[0091]
[0092] If the indicator has a restrictive effect on the criterion layer, that is, the larger the corresponding indicator value, the less secure the criterion layer, then the indicator is considered to have a restrictive effect on the criterion layer. At this time, the indicator attribute is negative, and it is processed according to the following formula:
[0093]
[0094] Among them, max(x i ' k ) is to convert x ikThe maximum value of the data after standard deviation standardization, min(x i ' k ) is to convert x ik The minimum value of the data after standardization, a ik The index value of the kth index in the multiple-selection index system in the i-th year after the range is normalized;
[0095] Step 1.4, for the normalized indicators, the correlation analysis method is used to calculate the correlation coefficients between the two indicators in each criterion layer. If the absolute value of the correlation coefficient between two indicators is greater than 0.8, any one of the corresponding two indicators is eliminated. Then, the indicator information sensitivity analysis is performed based on the improved principal component. The information sensitivity of all indicators eliminated by the correlation analysis method is calculated, and they are sorted from large to small according to information sensitivity. Indicators with cumulative information sensitivity ≥ 70% are retained, and each indicator is divided into 5 levels according to the indicator value, namely: unsafe, less safe, basically safe, relatively safe and very safe, to obtain the final evaluation indicator system;
[0096] The correlation analysis method is used to calculate the correlation coefficients between the two indicators in each criterion layer:
[0097]
[0098] Among them, r is the correlation coefficient matrix between indicators, r kk′ is the correlation coefficient of the kth and k′th indicators, a ik is the value of the kth indicator in the i-th year after normalization, a ik′ is the value of the k′th indicator in the i-th year after normalization, is the average value of the kth indicator in N years after normalization, is the average value of the k′th indicator in N years after normalization, k, k′=1,2,…,K, and the measurement standard of the correlation coefficient is: 0<|r kk′ |≤0.3, irrelevant; 0.3<|r kk′ |≤0.8, weak correlation; 0.8<|r kk′ |≤1, strong correlation;
[0099] The calculation formula for indicator information sensitivity analysis is as follows:
[0100]
[0101] r T r-λ k E K |=0
[0102]
[0103] Among them, τ m is the cumulative information sensitivity of the first m indicators; is the tth principal component Z t For the kth index X k The partial derivative of kt is the load coefficient of the kth index on the tth principal component; k is the eigenvalue of the kth index; e kt is the eigenvector corresponding to the eigenvalue of the kth index; η k is the variance contribution rate of the kth indicator; β k is the information sensitivity of the kth indicator; is the information sensitivity of the kth indicator calculated according to the ranking; r is the correlation coefficient matrix between indicators, E K is the K-order unit matrix, Ω t is the cumulative variance contribution rate of the first t indicators, Ω t When the sensitivity is ≥85%, the first t indicators are retained as principal components, t is the number of principal components (t∈[1,K]), and m is the number of indicators when the cumulative information sensitivity is ≥70% (m∈[1,K]).
[0104] Step 2: Calculate the index weight using the correlation weight entropy weight method; specifically:
[0105] For the entropy weight ω of the vth index v , which is the weight of the vth indicator, where v = 1, 2, ..., m, and is calculated according to the following formula:
[0106]
[0107] Among them, e v is the entropy value of the vth indicator, and m is the number of indicators in the final evaluation system.
[0108] The entropy value e of the vth indicator v Calculate according to the following formula:
[0109]
[0110] X0=x ov
[0111] Among them, f iv is the proportion of the normalized value of the vth indicator in the i-th year to the normalized total value of the vth indicator in the N-th year; N is the number of samples, N = 12; X0 is the reference sequence, if it is a positive indicator, x 0v is the maximum value of the original data of the same indicator v in different sample data. If it is a negative indicator, x 0v is the minimum value of the original data of the same indicator v in different sample data; r 0vis the vth index a in the reference sequence 0v and the vth indicator a in the i-th year iv The correlation between 0v is the vth index x in the reference sequence 0v Normalized value; a iv is the normalized value of the vth indicator in the final indicator system in the i-th year; r v is the correlation between the reference sequence index and the vth index; ξ 0i (v) is the difference between the vth indicator in the i-th year and the vth indicator in the reference sequence a 0v The correlation coefficient between them; ρ is the resolution coefficient, which is taken as 0.5.
[0112] Step 3: Based on the indicator weights, the fuzzy comprehensive evaluation method is used to evaluate the agricultural water resources security status in the evaluation area, specifically:
[0113] Step 3.1, establish the evaluation factor set U and the comment set V; wherein the evaluation factor set U is the indicator composition of the final indicator system of the evaluation factor set, that is, U = {U1, U2, U3, ..., Um}, wherein U1, U2, U3, ..., Um represent m evaluation indicators respectively;
[0114] The comment set is a set of evaluation results, which are expressed by different levels and values. Five levels are selected: unsafe, less safe, basically safe, relatively safe, and very safe. V = {unsafe, less safe, basically safe, relatively safe, very safe} = {1, 2, 3, 4, 5};
[0115] Step 3.2, use the triangular membership function to determine the fuzzy relationship matrix R of each sample from the evaluation factor set U to the comment set V; specifically:
[0116] The fuzzy relationship matrix R is:
[0117]
[0118] If the indicator attribute is positive:
[0119]
[0120] If the indicator attribute is negative:
[0121]
[0122] Among them, r vjis the membership degree of the vth indicator at the jth level, v = 1, 2, ..., m; j is the jth level in the comment set, j = 1, 2, 3, 4, 5; x1, x3, x5, x7 are the division nodes of the evaluation level data interval except the two end points, x2, x4, x6 are the midpoints of the level intervals x1 ~ x3, x3 ~ x5, x5 ~ x7 respectively; x is the value of the evaluation sample on each indicator, q is the number of levels, q = 5;
[0123] According to the calculation formula of R, the fuzzy relationship matrix is divided into three matrices R1, R2, and R3 according to the criterion layer;
[0124]
[0125] Among them, R1, R2, and R3 represent the fuzzy relationship matrices of the three criterion layers of water resources security, social and economic security, and ecological and environmental security, respectively, c = m, a is the number of indicators in the water resources security criterion layer, ba is the number of indicators in the social and economic security criterion layer, and cba is the number of indicators in the ecological and environmental security criterion layer;
[0126] Step 3.3: Use weighted average synthesis operator to perform fuzzy synthesis to obtain the fuzzy vector B of each criterion layer. g (g = 1, 2, 3), combined with the evaluation level division, the maximum membership principle is adopted, and the fuzzy vector B is taken g The evaluation level where the maximum value of all components in is taken as the evaluation result of the criterion layer, and the numerical vector V of the evaluation level is g With the fuzzy vector B g Multiply them together to get the score F of the evaluation area on the evaluation target, where:
[0127] The fuzzy vector B of each criterion layer g Calculate according to the following formula:
[0128] B g =ω g ·R g
[0129] Among them, when g = 1, 2, and 3, they represent three criterion levels, that is, g is the water resources security or social and economic security or ecological and environmental security level;
[0130] ω g is the set of all indicator weights in the corresponding criterion layer; R g is the fuzzy relationship matrix corresponding to the criterion layer;
[0131] The calculation formula of the score F of the evaluation area on the evaluation target is as follows:
[0132]
[0133] Where: Vg is the numerical vector V of the evaluation level g =(1,2,3,4,5).
[0134] Example 2
[0135] An evaluation system for agricultural water resource security, used to implement the evaluation method for agricultural water resource security of embodiment 1, comprising a data acquisition and transmission module, a data storage module, an algorithm model module and a client connected in sequence, the data acquisition and transmission module is used to collect index data and transmit the data to the storage module, and the data acquisition and transmission module collects index data by manual entry and Web page data source crawling to realize data collection;
[0136] The data storage module is used to store the index data collected by the data collection and transmission module and the intermediate data calculated by the algorithm model module;
[0137] The algorithm model module is used to implement normalization processing, weight assignment, and use fuzzy comprehensive evaluation method to evaluate the agricultural water resources security status of the evaluation area according to the indicator weight. The processing process of normalization processing, weight assignment, and use fuzzy comprehensive evaluation method to evaluate the agricultural water resources security status of the evaluation area according to the indicator weight in Example 1 is carried in the algorithm model module;
[0138] The client is used for visualization of agricultural water resources security assessment results and for use by managers and grassroots users, including web and mobile terminals.
[0139] Example 3
[0140] Based on Example 1, this example takes the agricultural water resources security assessment in Xi'an in 2022 as an example, and is implemented in the following steps:
[0141] The evaluation method for agricultural water resources security is implemented in the following steps:
[0142] Step 1: Construct an evaluation index system, including indicators, indicator values and evaluation criteria; specifically:
[0143] Step 1.1: Build an initial evaluation index system
[0144] The frequency analysis method was used to select indicators with a frequency ≥ 5 in the relevant literature on northern river basins in the past five years, and the selected indicators were divided into three criterion levels: water resources security, social and economic security, and ecological and environmental security according to the characteristics of the indicators, thus obtaining an initial evaluation indicator system. Among them, 14 indicators were selected at the water resources security level, 10 indicators were selected at the social and economic security level, and 7 indicators were selected at the ecological and environmental security level, totaling 31 indicators. The initial indicator system is shown in Table 1.
[0145] Table 1 Initial indicator system for agricultural water resources security assessment
[0146]
[0147]
[0148] Step 1.2, according to the specific situation of the study area, obtain the corresponding indicator value. Each indicator data is in units of years and includes the data of the past N years. According to the acquisition of the indicator value, the indicators selected in step 1.1 are re-selected to obtain a re-selected evaluation indicator system. When re-selecting, the indicators with difficult data acquisition and redundant meanings in the initial indicator system for agricultural water resources security evaluation are increased or decreased to construct a re-selected indicator system. The specific increase and decrease process at the water resources security level is as follows: the per capita water resources, surface water resources, groundwater resources, and per mu water resources of cultivated land have overlapping meanings. The per capita water resources indicator representing the water resources for human activities and the per mu water resources indicator representing the water resources for agricultural production activities are retained. The per-mu water resources indicator of cultivated land with quantity was excluded, and the surface water resources and groundwater resources indicators were eliminated; the data of total water supply and water supply modulus indicators were not complete, so they were eliminated; the per-mu irrigation water consumption of farmland and the farmland irrigation water consumption both represented the water consumption of agricultural production activities, and the former was more representative, so the per-mu irrigation water consumption indicator of farmland was retained, and the farmland irrigation water consumption indicator was eliminated; the surface water resources development and utilization rate and the groundwater resources development and utilization rate indicators were merged into the water resources development and utilization rate indicator, and the remaining indicators were retained. After re-selection, there were 8 indicators at the water resources security level, 11 indicators at the social and economic security level, and 8 indicators at the ecological and environmental security level, totaling 27 indicators. The re-selected indicator system is shown in Table 2.
[0149] Table 2 Agricultural water resources security evaluation multiple-selection indicator system
[0150]
[0151] Step 1.3, normalize the selected indicators. For the kth indicator, the normalized indicator data is: 11 ,a 12 ,…,a Nk}, where a NK The normalized value of the indicator value of the kth indicator in the Nth year is firstly used to eliminate the influence of outliers on the indicator system, and then the range standardization method is used to eliminate the influence of the variable dimension and variation range, and the data is scaled so that it falls into [0,1]. The normalized data is corrected using constants to avoid errors caused by zero values;
[0152] Step 1.4, for the normalized indicators, the correlation analysis method is used to calculate the correlation coefficients between the two indicators in each criterion layer. If the absolute value of the correlation coefficient between two indicators is greater than 0.8, any one of the corresponding two indicators is eliminated. Then, the indicator information sensitivity analysis is performed based on the improved principal component. The information sensitivity of all indicators eliminated by the correlation analysis method is calculated, and they are sorted from large to small according to information sensitivity. Indicators with cumulative information sensitivity ≥ 70% are retained, and each indicator is divided into 5 levels according to the indicator value, namely: unsafe, relatively unsafe, basically safe, relatively safe and very safe, to obtain the final evaluation index system. After screening, there are 7 indicators at the water resources security level, 11 indicators at the social and economic security level, and 6 indicators at the ecological and environmental security level, totaling 24 indicators;
[0153] The calculation of the correlation coefficient between two indicators takes the eight indicators at the water resources security level as an example. The correlation relationship of indicators at this level is shown in Table 3.
[0154] Table 3 Correlation between indicators under water resource criteria layer
[0155]
[0156] The correlation between the water production coefficient and the water production modulus is 0.943>0.8, and the relationship is strongly correlated. The water production modulus can better reflect the regional water resources situation, so the water production coefficient indicator is eliminated.
[0157] The information sensitivity of the 24 indicators after correlation analysis was calculated. The greater the information sensitivity, the more important the indicator. The indicators were sorted from large to small to calculate the cumulative information sensitivity after sorting. The indicators with cumulative information sensitivity ≥ 70% were retained, and the remaining indicators were eliminated.
[0158] The sensitivity and cumulative information content of water resource criterion layer indicators are shown in Table 4:
[0159] Table 4 Sensitivity and cumulative information content of indicators at the water resource criterion level
[0160] index Sensitivity Sensitivity ranking Cumulative Information Content Precipitation 30.1963 Average irrigation water consumption per mu of farmland (31.8312) 0.1572 Water production modulus 24.9532 Water resources development and utilization rate (30.3933) 0.3072 Water resources per capita 26.9671 Average water resources per mu of cultivated land (30.2306) 0.4565 Water resources development and utilization rate 30.3933 Precipitation (30.1963) 0.6056 Average water resources per mu of cultivated land 30.2306 Per capita domestic water consumption (27.9581) 0.7436 Average irrigation water consumption per mu of farmland 31.8312 Per capita water resources (26.9671) 0.8768 Per capita domestic water consumption 27.9581 Water production modulus (24.9532) 1.0000
[0161] The cumulative information content of the indicators "water resources development and utilization rate" and "average water resources per mu of cultivated land" are 0.6056<0.7 and 0.7436>0.7 respectively, so the indicators "average irrigation water consumption per mu of farmland", "water resources development and utilization rate", "average water resources per mu of cultivated land", "precipitation" and "per capita domestic water consumption" are selected, and the indicators "water production modulus" and "per capita water resources" are screened out.
[0162] Similarly, other criterion layer indicators preferably refer to the water resources layer.
[0163] Finally, the final evaluation index system was selected from the multiple-selection index system, and the agricultural water resources security evaluation index system is shown in Table 5. Agricultural water resources security is the target layer, water resources security, social and economic security, and ecological and environmental security are the criterion layer, and precipitation, water resources development and utilization rate, per mu of cultivated land water resources, per mu of farmland irrigation water consumption, per capita domestic water consumption, crop planting area, field channel integrity rate, field channel matching rate, irrigation water utilization coefficient, water fee collection rate, single water production, population density, total wastewater discharge, industrial wastewater discharge compliance rate, forest coverage rate, drinking water quality compliance rate, and agricultural fertilizer application intensity are the indicator layer.
[0164] Table 5 Agricultural water resources security evaluation index system
[0165]
[0166]
[0167] The indicators are divided into five levels, namely unsafe, relatively unsafe, basically safe, relatively safe and very safe. The specific classification of the indicators is shown in Table 5.
[0168] Table 5 Grading of agricultural water resources security evaluation indicators
[0169] Indicator name unit Very safe Safer Basic safety Less safe Unsafe Precipitation mm ≥80 60~80 40~60 20~40 <20 Water resources development and utilization rate % ≤0.1 0.1~0.2 0.2~0.4 0.4~0.6 >0.6 Average water resources per mu of cultivated land <![CDATA[m 3 / acre]]> ≥40 30~40 20~30 10~20 <10 Average irrigation water consumption per mu of farmland <![CDATA[m 3 / acre]]> ≤10 10~20 20~40 40~60 >60 Per capita domestic water consumption Liter / day ≤50 50~80 80~120 120~150 >150 Crop planting area Ten thousand acres ≥600 300~600 200~300 100~200 <100 Field channel matching rate % ≥0.9 0.8~0.9 0.7~0.8 0.6~0.7 <0.6 Field channel integrity rate % ≥0.85 0.77~0.85 0.69~0.77 0.6~0.69 <0.6 Irrigation water utilization coefficient % ≥0.7 0.65~0.7 0.6~0.65 0.55~0.6 <0.55 Water fee collection rate % ≥0.8 0.6~0.8 0.4~0.6 0.2~0.4 <0.2 Water production per unit area <![CDATA[kg / m 3 ]]> ≥3.0 2.5~3.0 2.0~2.5 1.5~2.0 <1.5 Population Density <![CDATA[persons / km 2 > ≤70 70~140 140~250 250~300 >300 Water supply network leakage rate % ≤0.06 0.06~0.08 0.08~0.1 0.1~0.2 >0.2 Total amount of waste water discharged 100 million t ≤0.08 0.08~0.25 0.25~0.58 0.58~0.8 >0.8 Industrial wastewater discharge compliance rate % ≥0.95 0.90~0.95 0.80~0.90 0.60~0.80 <0.60 Forest Coverage % ≥0.7 0.6~0.7 0.4~0.6 0.2~0.4 <0.2 Drinking water quality compliance rate % ≥0.95 0.90~0.95 0.80~0.90 0.60~0.80 <0.60 Intensity of agricultural fertilizer application <![CDATA[kg / hm 2 ]]> ≤15 15`20 20~25 25~30 >30
[0170] Step 2, using the correlation weight entropy weight method to calculate the indicator weight;
[0171] According to the calculation formula of the indicator weight in Example 1, the indicator weight is divided into three weight sets according to the criterion layer:
[0172] A1=(ω1,ω2,…,ω5)=(0.0747,0.1806,0.0236,0.0384,0.0160);
[0173] A2=(ω6,ω7,…,ω 13 )=(0.0873,0.0455,0.0177,0.0676,0.0454,0.0421,0.0089,0.0189);
[0174] A3=(ω 14 ,ω 15 ,…,ω 18 )=(0.0339,0.0372,0.1272,0.0810,0.0540)
[0175] Step 3: Based on the indicator weights, the fuzzy comprehensive evaluation method is used to evaluate the agricultural water resources security status in the evaluation area, specifically:
[0176] Step 3.1, the evaluation factor set is composed of the indicators in the indicator system, with a total of 18 evaluation factors, U = {precipitation, water resource development and utilization rate, per mu water resource of cultivated land, per mu irrigation water consumption of farmland, per capita domestic water consumption, crop planting area, field channel integrity rate, field channel matching rate, irrigation water utilization coefficient, water fee collection rate, unit water production, population density, total wastewater discharge, industrial wastewater discharge compliance rate, forest coverage rate, drinking water quality compliance rate, agricultural fertilizer application intensity};
[0177] The comment set is a set of evaluation results, which are expressed by different levels and values. Five levels are selected: unsafe, less safe, basically safe, relatively safe, and very safe. V = {unsafe, less safe, basically safe, relatively safe, very safe} = {1, 2, 3, 4, 5};
[0178] Step 3.2, using the triangular membership function to determine the fuzzy relationship matrix R of each sample from the evaluation factor set U to the comment set V;
[0179] The degree of membership of the indicators calculated in Example 1 is shown in Table 6:
[0180] Table 6 The degree of membership of indicators in the hierarchy
[0181]
[0182]
[0183] The fuzzy relationship matrix is divided into three sets according to the criterion layer:
[0184]
[0185] Step 3.3, according to Example 1, the fuzzy vectors of the three criterion layers are calculated as follows: B1=(0.1152, 0, 0.2510, 0.0633, 0.2241);
[0186] B2=(0.1464,0.0042,0.0028,0.1023,0.0554);
[0187] B3=(0.1169,0.0181,0.1367,0.0092,0).
[0188] Calculate the total score of the Agricultural Water Resources Security Assessment System:
[0189] F=(0.1152×5+0×4+0.2510×3+0.0633×2+0.2241×1)+…+(0.1169×5+0.0181×4+0.1367×3+0.0092×2+0×1)=3.7826
[0190] Therefore, the evaluation results of the criterion layers in the agricultural water resources security evaluation of Xi'an in 2022 are: the water resources security layer is "relatively safe", the social and economic security layer is "very safe", and the ecological and environmental security layer is "basic safety"; the comprehensive score is 3.7826, which falls between [3,4], indicating that the agricultural water resources security level in Xi'an in 2022 is "basic safety" and tends to be "relatively safe".
Claims
1. An evaluation method for agricultural water resources security, characterized in that: The specific steps are as follows: Step 1: Construct an evaluation index system, including indicators, indicator values and evaluation criteria; Step 2, using the correlation weight entropy weight method to calculate the indicator weight; Step 3: Based on the indicator weights, the fuzzy comprehensive evaluation method is used to evaluate the agricultural water resources security status in the evaluation area.
2. The evaluation method for agricultural water resources security according to claim 1, characterized in that: The step 1 is specifically as follows: Step 1.1: Construct an initial evaluation index system The frequency analysis method was used to select indicators with a frequency of ≥5 in the relevant literature in the research area in the past five years, and the selected indicators were divided into three criteria levels: water resources security, social and economic security, and ecological and environmental security according to the characteristics of the indicators, thus obtaining the initial evaluation indicator system; According to the specific situation of the study area, the corresponding indicator value is obtained. Each indicator data includes the data of the past N years in units of years. According to the acquisition of the indicator value, the indicators selected in step 1.1 are re-selected to obtain the re-selected evaluation indicator system. Assuming that the evaluation indicator system includes K indicators in total, the indicator data for the kth indicator is: {x 1k ,x 2k ,...,x Nk }, where x Nk It represents the indicator value of the kth indicator in the Nth year, k = 1, 2, 3, ..., K. The data corresponding to all indicators in each year is a sample data, that is, it includes N sample data. For the i-th sample data: {x i1 ,x i2 ,...,x iK }, where x iK Indicates the indicator value corresponding to the Kth indicator in the i-th year, i = 1, 2, 3, ..., N; Step 1.3, normalize the selected indicators. For the kth indicator, the normalized indicator data is: 11 ,a 12 ,...,a NK }, where a NK It is the normalized value of the Kth indicator in the Nth year; Step 1.4, for the normalized indicators, the correlation analysis method is used to calculate the correlation coefficients between the two indicators in each criterion layer. If the absolute value of the correlation coefficient between two indicators is >0.8, any one of the corresponding two indicators is eliminated, and then the indicator information sensitivity analysis is performed based on the improved principal component. The information sensitivity of all indicators eliminated by the correlation analysis method is calculated, and they are sorted from large to small according to information sensitivity. Indicators with cumulative information sensitivity ≥70% are retained, and each indicator is divided into 5 levels according to the indicator value, namely: unsafe, relatively unsafe, basically safe, relatively safe and very safe, to obtain the final evaluation index system.
3. The evaluation method for agricultural water resources security according to claim 2, characterized in that: The step 1.3 is specifically as follows: For each indicator data, the standard deviation standardization method is first used to eliminate the influence of outliers on the indicator system, and then the range standardization method is used to eliminate the influence of variable dimensions and variation range, and the data is scaled so that it falls into [0,1]; Among them, the standard deviation standardization formula is: Among them, x ik is the index value of the kth index in the multiple-selection index system in the i-th year, i = 1, 2, 3, ..., N, k = 1, 2, 3, ..., K, x i ' k To x ik The data after standardization is is the mean of the original data of the k-th indicator in N years; σ is the standard deviation of the k-th indicator; The range standardization method processing formula is: If the indicator has a driving effect on the criterion layer, that is, the larger the corresponding indicator value, the safer the criterion layer, then it is considered that the indicator has a driving effect on the criterion layer. At this time, the indicator attribute is positive, and it is processed according to the following formula: If the indicator has a restrictive effect on the criterion layer, that is, the larger the corresponding indicator value, the less secure the criterion layer, then the indicator is considered to have a restrictive effect on the criterion layer. At this time, the indicator attribute is negative, and it is processed according to the following formula: Among them, max(x i ' k ) is to convert x ik The maximum value of the data after standard deviation standardization, min(x i ' k ) is to convert x ik The minimum value of the data after standardization, a ik It is the indicator value of the kth indicator in the multiple-selection indicator system in the i-th year after being standardized by the range.
4. The evaluation method for agricultural water resources security according to claim 3, characterized in that: In step 1.4, the correlation analysis method is used to calculate the correlation coefficients between the two indicators in each criterion layer respectively, specifically: Among them, r is the correlation coefficient matrix between indicators, r kk′ is the correlation coefficient of the kth and k′th indicators, a ik is the value of the kth indicator in the i-th year after normalization, a ik′ is the value of the k′th indicator in the i-th year after normalization, is the average value of the kth indicator in N years after normalization, is the average value of the k′th indicator in N years after normalization, k, k′=1,2,…,K, and the measurement standard of the correlation coefficient is: 0<|r kk′ |≤0.3, irrelevant; 0.3<|r kk′ |≤0.8, weak correlation; 0.8<|r kk′ |≤1, strong correlation; The calculation formula for indicator information sensitivity analysis is as follows: Among them, τ m is the cumulative information sensitivity of the first m indicators; is the tth principal component Z t For the kth index X k The partial derivative of kt is the load coefficient of the kth index on the tth principal component; k is the eigenvalue of the kth index; e kt is the eigenvector corresponding to the eigenvalue of the kth index; η k is the variance contribution rate of the kth indicator; β k is the information sensitivity of the kth indicator; is the information sensitivity of the kth indicator calculated according to the ranking; r is the correlation coefficient matrix between indicators, E K is the K-order unit matrix, Ω t is the cumulative variance contribution rate of the first t indicators, Ω t When the sensitivity is ≥85%, the first t indicators are retained as principal components, t is the number of principal components (t∈[1,K]), and m is the number of indicators when the cumulative information sensitivity is ≥70% (m∈[1,K]).
5. The evaluation method for agricultural water resources security according to claim 4, characterized in that: The step 2 is specifically as follows: For the entropy weight ω of the vth index v , which is the weight of the vth indicator, where v = 1, 2, ..., m, and is calculated according to the following formula: Among them, e v is the entropy value of the vth indicator, and m is the number of indicators in the final evaluation system.
6. The evaluation method for agricultural water resources security according to claim 5, characterized in that: The entropy value e of the vth indicator v Calculate according to the following formula: X0=x ov Among them, f iv is the proportion of the normalized value of the vth indicator in the i-th year to the normalized total value of the vth indicator in the N-th year; N is the number of samples, N = 12; X0 is the reference sequence, if it is a positive indicator, x 0v is the maximum value of the original data of the same indicator v in different sample data. If it is a negative indicator, x 0v is the minimum value of the original data of the same indicator v in different sample data; r 0v is the vth index a in the reference sequence 0v and the vth indicator a in the i-th year iv The correlation between 0v is the vth index x in the reference sequence 0v Normalized value; a iv is the normalized value of the vth indicator in the final indicator system in the i-th year; r v is the correlation between the reference sequence index and the vth index; ξ 0i (v) is the difference between the vth indicator in the i-th year and the vth indicator in the reference sequence a 0v The correlation coefficient between them; ρ is the resolution coefficient, which is taken as 0.
5.
7. The evaluation method for agricultural water resources security according to claim 6, characterized in that: The step 3 is specifically as follows: Step 3.1, establish the evaluation factor set U and the comment set V; Step 3.2, using the triangular membership function to determine the fuzzy relationship matrix R of each sample from the evaluation factor set U to the comment set V; Step 3.3: Use weighted average synthesis operator to perform fuzzy synthesis to obtain the fuzzy vector B of each criterion layer. g (g = 1, 2, 3), combined with the evaluation level division, the maximum membership principle is adopted, and the fuzzy vector B is taken g The evaluation level where the maximum value of all components in is taken as the evaluation result of the criterion layer, and the numerical vector V of the evaluation level is g With the fuzzy vector B g Multiply them together to get the score F of the evaluation area on the evaluation target.
8. The evaluation method for agricultural water resources security according to claim 7, characterized in that: The evaluation factor set U in step 3.1 is composed of the indicators in the final indicator system, that is, U = {U1, U2, U3, ..., Um}, where U1, U2, U3, ..., Um represent m evaluation indicators respectively; The comment set is a collection of evaluation results, which are represented by different levels and values. Five levels are selected: unsafe, less safe, basically safe, relatively safe, and very safe. V = {unsafe, less safe, basically safe, relatively safe, very safe} = {1, 2, 3, 4, 5}.
9. The evaluation method for agricultural water resources security according to claim 8, characterized in that: The step 3.2 is specifically as follows: The fuzzy relationship matrix R is: If the indicator attribute is positive: If the indicator attribute is negative: Among them, r vj is the membership degree of the vth indicator at the jth level, v = 1, 2, ..., m; j is the jth level in the comment set, j = 1, 2, 3, 4, 5; x1, x3, x5, x7 are the division nodes of the evaluation level data interval except the two end points, x2, x4, x6 are the midpoints of the level intervals x1 ~ x3, x3 ~ x5, x5 ~ x7 respectively; x is the value of the evaluation sample on each indicator, q is the number of levels, q = 5; According to the calculation formula of R, the fuzzy relationship matrix is divided into three matrices R1, R2, and R3 according to the criterion layer; Among them, R1, R2, and R3 represent the fuzzy relationship matrices of the three criterion layers of water resources security, social and economic security, and ecological and environmental security, respectively, c = m, a is the number of indicators in the water resources security criterion layer, ba is the number of indicators in the social and economic security criterion layer, and cba is the number of indicators in the ecological and environmental security criterion layer; In step 3.3: The fuzzy vector B of each criterion layer g Calculate according to the following formula: B g =A g ·R g Among them, when g = 1, 2, and 3, they represent three criterion levels, that is, g is the water resources security or social and economic security or ecological and environmental security level; A g is the set of all indicator weights in the corresponding criterion layer; R g is the fuzzy relationship matrix corresponding to the criterion layer; A1={ω1,ω2,…,ω a },A2={ω a+1 ,oh a+2 ,…,oh b },A3={ω b+1 ,oh b+2 ,…,oh c }; The calculation formula of the score F of the evaluation area on the evaluation target is as follows: Where: V g is the numerical vector V of the evaluation level g =(1,2,3,4,5).
10. An evaluation system for agricultural water resources security, characterized in that: Used to implement the evaluation method for agricultural water resources security in claim 9, comprising a data acquisition and transmission module, a data storage module, an algorithm model module and a client connected in sequence, wherein the data acquisition and transmission module is used to collect index data and transmit the data to the storage module, and the data acquisition and transmission module collects index data by manual entry and Web page data source crawling to realize data collection; The data storage module is used to store the index data collected by the data collection and transmission module and the intermediate data calculated by the algorithm model module; The algorithm model module is used to implement normalization processing, weight assignment, and use fuzzy comprehensive evaluation method to evaluate the agricultural water resources security status of the evaluation area according to the indicator weight; The client is used for visualization of agricultural water resources security assessment results and use by managers and grassroots users, including a Web terminal and a mobile terminal.
Citation Information
Patent Citations
Unified evaluation method for agricultural water utilization efficiency and ecological environment effect
CN113065786A
