Community disaster resilience assessment method based on improved entropy weight-CRITIC method

Abstract

The present invention discloses a community disaster resilience assessment method based on the improved entropy weight CRITIC method. By constructing a quantitative indicator system, different representative communities are selected, and the specific values ​​of the underlying indicators are obtained through field investigations. A community disaster resilience assessment model is established and the data is standardized. Then, the indicators are combined and weighted using the improved entropy weight CRITIC method, and the grade standards of each indicator of the evaluation object are divided. The community disaster resilience evaluation standard is established according to the critical value of the quantitative evaluation indicator, and the grades are divided based on the TOPSIS method. The overview and classification of the four resilience grades of I, II, III, and IV are briefly described based on previous data and specific standards. Finally, through example calculations, the closeness of the indicators of each evaluation object is obtained, and its resilience grade is obtained, so that it can be analyzed whether the specific actual situation of each evaluation object is consistent with the calculated resilience grade.

Description

Technical Field

[0001] The present invention belongs to the field of resilient cities and relates to a community disaster resilience assessment method based on an improved entropy weight-CRITIC method. Background Art

[0002] In recent years, natural disasters (such as earthquakes, floods, and mudslides) have had an increasingly greater impact on China's urban development. How to improve the city's ability to respond to disasters and the city's own resilience has become the focus of scholars' research. As the basic unit of the city, the community is the basic place for urban residents to carry out activities and the basis for ensuring urban safety. There are relatively few studies on community resilience in my country involving disaster prevention, and the disaster response system is not yet perfect, showing a certain degree of vulnerability in the face of disasters. Research on disaster risks and response strategies based on resilience theory is of great significance to promoting the safe development of cities.

[0003] With the progress of society, people gradually realize that only by building resilient communities can the daily lives of residents be better maintained. "Resilience" is the basic place for urban residents to carry out activities and the basis for ensuring urban safety. There are relatively few studies on community resilience in my country involving disaster prevention. The disaster response system is not perfect enough and shows a certain degree of vulnerability in the face of disasters. Research on disaster risks and response strategies based on resilience theory is of great significance to promoting the safe development of cities. Summary of the invention

[0004] The purpose of the present invention is to provide a community disaster resilience assessment method based on the improved entropy weight-CRITIC method, which can have a better assessment effect on the resilience of the community.

[0005] In order to achieve the above task, the present invention adopts the following scheme, which specifically includes the following steps:

[0006] A community disaster resilience assessment method based on the improved entropy weight-CRITIC method is characterized by comprising the following steps:

[0007] Step 1: Construct a quantitative indicator system, establish a criterion layer from the four aspects of society, economy, organization and infrastructure, and then set up an indicator layer, point out the link of resilience and disaster prevention to which each indicator belongs, and select the measurement method for each indicator.

[0008] Step 2: Select different representative communities and obtain specific values ​​of underlying indicators through field investigations.

[0009] Step 3: Establish a community disaster resilience evaluation model. Assuming there are m evaluation objects and n evaluation indicators, establish the original data evaluation matrix X;

[0010] X=(x ij ) m×n

[0011] Step 4: Standardize the data;

[0012] Step 5: Calculate the indicator value x″ ij The weight under index j gives the matrix P:

[0013] P=(p ij ) m×n

[0014]

[0015] Step 6: Calculate the entropy value of j and get the matrix E:

[0016] E=(e1,e2,……,e n )

[0017]

[0018] in,

[0019] Step 7: Calculate the difference coefficient of index j and get the matrix G:

[0020] G=(g1,g2,……,g n )

[0021] g j =1-e j

[0022] Step 8: Find the weight of each indicator and get the matrix W:

[0023] W=(ω1,ω2,……,ω n )

[0024]

[0025] Step 9: Improve the CRITIC weighting method and calculate the coefficient of variation v of each indicator j :

[0026]

[0027] Step 10: Get the correlation coefficient matrix R, which is calculated as follows:

[0028] R=(r jk ) m×n

[0029]

[0030] Step 11: Calculate the comprehensive information of the indicators and obtain the matrix C:

[0031] C=(c1,c2,……,c n )

[0032]

[0033] Step 12: Determine the weight of each indicator and obtain matrix B:

[0034] B=(β1,β2,……,β n )

[0035]

[0036] Step 13: Determine the combined weights. After the above steps, the weight values ​​calculated by the improved entropy weight method and the CRITIC method are obtained respectively. Finally, a Lagrange multiplier method is used to solve the combined weights to obtain the matrix Z:

[0037] Z=(z1,z2,……,z n )

[0038]

[0039] Step 14: Assess the disaster resilience of the community, grade the resilience assessment indicators into four levels: I, II, III, and IV, and assign corresponding scores to each level.

[0040] Step 15: Based on the indicator data of each evaluation object, a score matrix is ​​obtained, which is then multiplied by the comprehensive weight to obtain the score of community disaster resilience of each plan, so as to compare the resilience scores; in addition, from another perspective, the scores of the four criterion layers of each evaluation object can be obtained, so as to make horizontal and vertical comparisons.

[0041] Step 16: Use the TOPSIS method to calculate the toughness level of each evaluation object.

[0042] Step 17: Determine the community disaster resilience level of each evaluation object based on the obtained closeness value.

[0043] Specifically, the data standardization process described in step 4 is performed in the following manner:

[0044] Step 4-1: Assume that the index value after standardization is:

[0045]

[0046] Among them, x′ ij Indicates the index value after standardization; represents the average value of the jth index; s j represents the standard deviation of the j-th indicator;

[0047] Step 4-2: In order to eliminate negative values, the coordinates can be translated to obtain the matrix X″:

[0048] x″ ij = x′ ij +k

[0049] Where k is the value of the coordinate translation.

[0050] Furthermore, the specific implementation method of calculating the toughness grade of each evaluation object by using the TOPSIS method in step 16 is:

[0051] Step 16-1: To eliminate the impact of dimension, the indicator matrix of the evaluation object is dimensionless, and the values ​​are normalized using the following method to obtain the standardized matrix P;

[0052] The larger the better indicators are:

[0053]

[0054] The smaller the better indicators are:

[0055]

[0056] Step 16-2: Construct the weighted standardized decision matrix q ij :

[0057] q ij =z j ·p ij

[0058] Among them, zj is the comprehensive index obtained by the above calculation.

[0059] Step 16-3: Find the positive and negative ideal solutions:

[0060]

[0061]

[0062] Step 16-4: Calculate the distance between each evaluation index and the positive and negative ideal solutions:

[0063]

[0064]

[0065] Step 16-5: Calculate the relative closeness between each evaluation object and the optimal value and the worst value:

[0066]

[0067] The community disaster resilience assessment method based on the improved entropy weight-CRITIC method of the present invention has the following beneficial effects:

[0068] 1) The improved entropy weight-CRITIC method is used to combine and weight the indicators. It is a combination of two objective weighting methods. The entropy weight method's judgment on the importance of relevant indicators depends entirely on the data changes of the relevant indicators. The improved CRITIC method comprehensively measures the indicator weights based on the comparative strength of the indicators and the conflicts between the indicators. This can avoid the shortcomings of a single weighting method and eliminate the influence of human subjective factors in traditional weighting methods such as AHP, greatly improving the accuracy and objectivity of the evaluation indicator weights.

[0069] 2) The grading standards of each indicator of the evaluation object are divided into levels I-IV, and then the actual specific values ​​are graded and assigned scores (1-4 points). Then the scoring matrix is ​​listed, and the comprehensive resilience score of each evaluation object and the criterion-level resilience score are obtained, so that horizontal and vertical comparisons of various communities can be carried out.

[0070] 3) The community disaster resilience evaluation standard was established based on the critical value of the quantitative evaluation index, and the grades were divided based on the TOPSIS method. The overview and classification of the four resilience levels of I, II, III, and IV were briefly described based on previous literature and specific standards. Finally, through example calculations, the closeness of the indicators of each evaluation object was obtained, and its resilience level was obtained, so that it can be analyzed whether the specific actual situation of each evaluation object is consistent with the calculated resilience level. DETAILED DESCRIPTION

[0071] The present invention is further described in detail below in conjunction with specific embodiments.

[0072] Part I:

[0073] This embodiment provides a community disaster resilience assessment method based on the improved entropy weight-CRITIC method. The complete steps include:

[0074] Step 1: Construct a quantitative indicator system. This embodiment establishes a completely quantitative evaluation indicator system, taking community resilience as the target layer, and establishing a criterion layer from four aspects: social resilience, economic resilience, organizational resilience, and infrastructure resilience. There are a total of 18 indicator layers, and the link of resilience and disaster prevention to which each indicator belongs is pointed out, and the measurement method of each indicator is selected.

[0075]

[0076] Step 2: Select four representative communities, named P1, P2, P3 and P4. Obtain the specific values ​​of 18 underlying indicators through field investigation.

[0077] Step 3: Establish a community disaster resilience evaluation model and create the original data evaluation matrix X;

[0078]

[0079] Step 4: Standardize the data according to the formula; use the following processing method:

[0080] Step 4-1: Assume that the index value after standardization is:

[0081]

[0082] Among them, x′ ij Indicates the index value after standardization; represents the average value of the jth index; s j represents the standard deviation of the j-th indicator;

[0083] Step 4-2: In order to eliminate negative values, the coordinates can be translated to obtain the matrix X″:

[0084] x″ ij = x′ ij +k

[0085] Where k is the value of the coordinate translation.

[0086] Step 5: Calculate the indicator value x″ ij The weight under index j, we get the matrix P;

[0087] P=(p ij ) m×n

[0088]

[0089]

[0090] Step 6: Calculate the entropy value of j and get the matrix E:

[0091] E=(e1,e2,……,e n )

[0092]

[0093] in

[0094] Step 7: Calculate the difference coefficient of index j and get the matrix G:

[0095] G=(g1,g2,……,g n )

[0096] g j =1-e j

[0097] Step 8: Calculate the weight of each indicator to obtain the matrix W and the indicator weight of the criterion layer;

[0098] W=(0.0545 0.0631 0.0565 0.0642 0.0683 0.0490 0.0498 0.0523 0.05610.0566 0.0521 0.0468 0.0517 0.0521 0.0490 0.0528 0.0678 0.0573)

[0099]

[0100] Step 9: Improve the CRITIC weighting method and calculate the coefficient of variation v of each indicator j :

[0101]

[0102] Step 10: Get the correlation coefficient matrix R, which is calculated as follows:

[0103] R=(r jk ) m×n

[0104]

[0105] Step 11: Calculate the comprehensive information of the indicators and obtain the matrix C;

[0106] C=(c1,c2,……,c n )

[0107]

[0108] Step 12: Determine the weight of each indicator and obtain the matrix B;

[0109] B=(0.0529 0.0193 0.0529 0.0396 0.0187 0.0172 0.0350 0.0446 0.01010.0005 0.0370 0.4493 0.0605 0.0370 0.0172 0.0293 0.0704 0.0082)

[0110] Step 13: Determine the combined weights. After the above steps, the weight values ​​calculated by the improved entropy weight method and the CRITIC method are obtained respectively. Finally, a Lagrange multiplier method is used to solve the combined weights to obtain the matrix Z:

[0111] Z=(0.0650 0.0423 0.0662 0.0610 0.0433 0.0352 0.0506 0.0585 0.02880.0067 0.0532 0.1755 0.0678 0.0532 0.0352 0.0476 0.0837 0.0263)

[0112] Step 14: The toughness evaluation index is graded into four levels: I, II, III, and IV, and each level is assigned a corresponding score: I (4 points), II (3 points), III (2 points), and IV (1 point). The grade classification and grade value results are shown in the following table.

[0113] In addition, the classification of toughness levels Ⅰ, Ⅱ, Ⅲ and Ⅳ is as follows:

[0114] Level I resilience: The community has a developed economy; all kinds of facilities are intact; the social security system is complete and can provide a safety barrier for residents when disasters occur; the level of organizational management is high, and the community can maintain normal order when disasters occur. At the same time, it can learn from experience and lessons after the disaster and immediately recover to a better state than before the disaster.

[0115] Level II resilience: The community's economic level is average; various facilities are relatively good, and the basic structure can be guaranteed not to be damaged when a disaster occurs; the social security system is relatively complete; the organizational management level is average, and normal order can be restored in a timely manner after a disaster occurs.

[0116] Level III resilience: The community has relatively weak economic strength, average facilities, and an average social security system. The level of organizational management is relatively poor and is easily affected to a certain extent by disasters, and it takes a long time to recover after a disaster.

[0117] Level IV resilience: The community's economy is backward; the level of various facilities is low; the social security system is imperfect; the level of organizational management is low, which makes it easy to be affected by disasters and difficult to recover after disasters. Step 14: According to the indicator data of each evaluation object, a score matrix is ​​obtained, and then multiplied by the comprehensive weight to obtain the score of the community disaster prevention resilience of each plan, so as to compare the resilience scores; in addition, from another perspective, the scores of the four criterion layers of each evaluation object can be obtained, so as to make horizontal and vertical comparisons.

[0118]

[0119]

[0120] Step 15: Finally, we get the score matrix of the four communities.

[0121]

[0122] The community disaster resilience scores of the four communities are obtained by multiplying the score matrix with the combined weight Z obtained above:

[0123] Y=ZA=(2.5921 2.2708 2.2298 3.0220)

[0124] Multiplying the matrix blocks with the comprehensive matrix Z can obtain the scores of the four community criteria level indicators;

[0125]

[0126] From the data in the analysis table, we can see that in horizontal comparison, P1 community has the highest organizational resilience and the lowest economic resilience, P2 community has the highest organizational resilience and the lowest social resilience, P3 community has the highest economic resilience and the lowest social resilience, and P4 community has the highest organizational resilience and the lowest social resilience. In vertical analysis, in terms of social resilience, P1 community has the highest score, P3 community has the highest economic resilience score, P4 community has the highest organizational resilience score, and P4 community has the highest infrastructure resilience.

[0127] Step 16: Use the TOPSIS method to calculate the toughness level of each evaluation object.

[0128] Step 16-1: To eliminate the impact of dimension, the indicator matrix of the evaluation object is dimensionless, and the values ​​are normalized using the following method to obtain the standardized matrix P;

[0129]

[0130] Step 16-2: Construct a weighted standardized decision matrix Q;

[0131]

[0132] z j It is the comprehensive index obtained by the above calculation.

[0133] Step 16-3: Find positive and negative ideal solutions;

[0134] q + =(0.0000 0.0423 0.0662 0.0610 0.0433 0.0352 0.0506 0.0585 0.0288 0.0067 0.0532 0.0000 0.0678 0.0532 0.0352 0.0476 0.0000 0.0263)

[0135] q -=(0.0650 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0035 0.0000 0.0000 0.0000 0.0000 0.0000 0.0837 0.0000)

[0136] Step 16-4: Calculate the distance between each evaluation index and the positive and negative ideal solutions;

[0137] Step 16-5: Calculate the relative closeness between each evaluation object and the optimal value and the worst value;

[0138]

[0139] Calculate each group of data And obtain the toughness grade classification standard;

[0140]

[0141] Step 17: Determine the community disaster resilience level of each evaluation object. Through the above calculation, the disaster resilience level of P1 and P2 communities is Level III, and the disaster resilience level of P3 community is Level IV, which is at a medium-low level. Only the resilience level of P4 community is Level II, which is consistent with the comprehensive disaster resilience evaluation score obtained previously.

[0142] Part II:

[0143] Through the above methods and steps, resilience can be improved by improving the shortcomings of different communities. In order to improve the disaster resilience of the community, the P1 community should increase its efforts to improve social resilience and economic resilience, such as increasing medical service capabilities to ensure that people have access to medical and health measures; the P2 community has a low infrastructure resilience score and should improve the level of its disaster prevention facilities and housing facilities; the P3 community should also improve its social resilience indicators; the P4 community has a higher level of comprehensive resilience than the other three communities, but it should also work hard to improve its social resilience.

Claims

1. A community disaster resilience assessment method based on the improved entropy weight-CRITIC method, characterized by The specific steps include: Step 1: Construct a quantitative indicator system, establish a criterion layer from the four aspects of society, economy, organization and infrastructure, and then set up an indicator layer, and point out the link of resilience and disaster prevention to which each indicator belongs, and select the measurement method for each indicator; Step 2: Select different representative communities and obtain specific values ​​of underlying indicators through field investigations; Step 3: Establish a community disaster resilience evaluation model. Assuming there are m evaluation objects and n evaluation indicators, establish the original data evaluation matrix X; X=(x ij ) m×n Step 4: Standardize the data; Step 4-1: Assume that the index value after standardization is: Among them, x' ij Indicates the index value after standardization; represents the average value of the jth index; s j represents the standard deviation of the j-th indicator; Step 4-2: To eliminate negative values, translate the coordinates to obtain the matrix X': x” ij =x’ ij +k Where k is the value of the coordinate translation; Step 5: Calculate the indicator value x" ij The weight under index j gives the matrix P: P=(p ij ) m×n Step 6: Calculate the entropy value of j and get the matrix E: E=(e1,e2,……,e n ) in, Step 7: Calculate the difference coefficient of index j and get the matrix G: G=(g1,g2,……,g n ) g j =1-e j Step 8: Find the weight of each indicator and get the matrix W: W=(ω1,ω2,……,ω n ) Step 9: Improve the CRITIC weighting method and calculate the coefficient of variation v of each indicator j : Step 10: Get the correlation coefficient matrix R, which is calculated as follows: R＝(r jk ) m×n Step 11: Calculate the comprehensive information of the indicators and obtain the matrix C: C=(c1,c2,……,c n ) Step 12: Determine the weight of each indicator and obtain matrix B: B＝(β1,β2,……，β n ) Step 13: Determine the combined weights. After the above steps, the weight values ​​calculated by the improved entropy weight method and the CRITIC method are obtained respectively. Finally, a Lagrange multiplier method is used to solve the combined weights to obtain the matrix Z: Z=(z1,z2,……,z n ) Step 14: Assess the disaster resilience of the community, classify the resilience assessment indicators into four levels: I, II, III, and IV, and assign corresponding scores to each level; Level I resilience: The community has a developed economy; various facilities are intact; the social security system is complete and can provide a safety barrier for residents when disasters occur; the level of organizational management is high, and the community can maintain normal order when disasters occur. At the same time, after the disaster, lessons can be learned and immediately restored to a better state than before the disaster; Level II resilience: The community has an average economic level; various facilities are relatively good, and the basic structure can be guaranteed not to be damaged when a disaster occurs; the social security system is relatively complete; the organizational management level is average, and normal order can be restored in time after a disaster occurs; Level III resilience: The community has relatively weak economic strength; various facilities are average; the social security system is average; the level of organizational management is relatively poor, and it is easily affected to a certain extent by disasters, and it takes a long time to recover after the disaster; Level IV resilience: The community is economically backward; the level of various facilities is low; the social security system is imperfect; the level of organizational management is low, which makes it easy to be affected by disasters and difficult to recover after disasters; Step 15: Based on the indicator data of each evaluation object, a score matrix is ​​obtained, which is then multiplied by the comprehensive weight to obtain the community disaster resilience score of each plan, so as to compare the resilience scores. In addition, from another perspective, the scores of the four criterion layers of each evaluation object are also obtained, so as to make horizontal and vertical comparisons; Step 16: Calculate the toughness grade of each evaluation object using the TOPSIS method; Step 16-1: To eliminate the impact of dimension, the indicator matrix of the evaluation object is dimensionless, and the values ​​are normalized using the following method to obtain the standardized matrix P; The larger the better indicators are: The smaller the better indicators are: Step 16-2: Construct the weighted standardized decision matrix Q: q ij =z j ·p ij Among them, z j is the comprehensive index obtained by the above calculation; Step 16-3: Find the positive and negative ideal solutions: Step 16-4: Calculate the distance between each evaluation index and the positive and negative ideal solutions: Step 16-5: Calculate the relative closeness between each evaluation object and the optimal value and the worst value: Step 17: Determine the community disaster resilience level of each evaluation object based on the obtained closeness value.

Images