A method for constructing a forest fire warning system, a decision method and a device thereof

Abstract

The application discloses a method for constructing a forest fire warning system, a decision method and a device thereof, wherein the method for constructing the forest fire warning system comprises obtaining a weather condition set, a date set and a relationship set under different regions; an incomplete T-spherical fuzzy relation set is generated according to the weather condition set, the date set and the relationship set; a completion method is constructed for the incomplete T-spherical fuzzy relation set; a fire warning standard set is generated according to the T-spherical fuzzy relation set; and the forest fire warning system is constructed according to the weather condition set, the date set, the relationship set, the incomplete T-spherical fuzzy relation set and the fire warning standard set. The above scheme can cope with the incompleteness and uncertainty of the relationship between weather factors and the possibility of fire occurrence under the actual forest fire warning background, improve the accuracy and problem processing efficiency of the forest fire warning, and enable forest firefighters to intuitively understand the instant relevant information of the forest fire, so as to make a timely judgment on the current situation.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of data processing, and particularly relates to a method for constructing a forest fire warning system, a decision method and a device thereof. BACKGROUND

[0002] Forest fires have serious harm to ecological environment, biological survival and social economy, so it is necessary to effectively evaluate the risk of forest fire occurrence. Due to a large number of influencing factors and high complexity of problem solving process, the existing forest fire warning system often cannot comprehensively describe the uncertainty information, and cannot process the incomplete information in the collected data, and does not consider the limited rationality of decision makers. There are many influencing factors in the forest fire warning problem, which can be regarded as multiple indexes, and the potential burning area can be regarded as multiple decision makers, so it is a reasonable choice to transform the forest fire warning problem into a multi-attribute group decision problem for processing, which can solve the problem that the existing system cannot process to a certain extent. SUMMARY

[0003] In view of the technical problem that there are many influencing factors in the above forest fire warning problem, the present application provides a method for constructing a forest fire warning system, a decision method and a device thereof, which are used to cope with the incompleteness and uncertainty of the relationship between weather factors and fire occurrence possibility under the actual forest fire warning background.

[0004] In order to solve the above technical problem, the technical scheme adopted by the present application is as follows:

[0005] A method for constructing a forest fire warning system, characterized in that it comprises the following steps:

[0006] S1.1, obtaining a weather condition set, a date set corresponding to the weather condition set, and a relationship set under different regions corresponding to the weather condition set and the date set;

[0007] S1.2, generating an incomplete T-spherical fuzzy relation set according to the weather condition set, the date set and the relationship set under different regions;

[0008] S1.3, constructing a completion method to transform the incomplete T-spherical fuzzy relation set into a complete T-spherical fuzzy relation set;

[0009] S1.4, generating a fire warning standard set according to the complete T-spherical fuzzy relation set;

[0010] S1.5, constructing a forest fire warning system according to the weather condition set, the date set, the relationship set under different regions, the incomplete T-spherical fuzzy relation set and the fire warning standard set.

[0011] The method for generating an incomplete T-spherical fuzzy relation set based on the weather condition set, the date set, and the relation sets under different regions is as follows:

[0012] The initial data representing different indicators of the weather condition set are transformed into membership degrees in T-spherical fuzzy numbers according to the linear normalization formula, wherein the linear normalization formula is expressed as:

[0013]

[0014] in, Represents the normalized first The first in each region The date and the Membership degree of each indicator Represents the first in the initial data The first in each region The date and the The relationship between the indicators Relationships representing different regions are concentrated in the first Each region Representative Date Collection No. Date, Representing the concentrated weather conditions One indicator;

[0015] The non-membership degree and hesitation degree in the T-spherical fuzzy number are generated based on the membership degree and transformation formula in the aforementioned T-spherical fuzzy number, wherein the transformation formula is:

[0016]

[0017]

[0018] in, Represents the normalized first The first in each region The date and the Non-membership degree of each indicator Represents the normalized first The first in each region The date and the Hesitation level of each indicator;

[0019] For missing values ​​in the initial data of the weather condition set, their membership degree, non-membership degree, and hesitation degree are assigned the values ​​-1, 0, and 0, respectively;

[0020] The incomplete T-spherical fuzzy relation set is generated based on the membership degree, the non-membership degree, and the hesitation degree.

[0021] A decision-making method for constructing a forest fire early warning system includes the following steps:

[0022] S2.1, input the weather condition set, the date set corresponding to the weather condition set, and the relationship set under different regions corresponding to the weather condition set and the date set into the forest fire warning system for data initialization processing, and generate an incomplete T-spherical fuzzy relationship set after processing;

[0023] S2.2, calculate the similarity classes of all dates in the date set by using a similarity-based incomplete T-spherical fuzzy processing method, and generate the conditional probability and the adjustable conditional probability of all dates in the date set according to the definition of the similarity class;

[0024] S2.3, convert the incomplete T-spherical fuzzy relationship set into a complete T-spherical fuzzy relationship set by using a similarity-based completion method;

[0025] S2.4, calculate the index weight set in the weather condition set according to the complete T-spherical fuzzy relationship set and the entropy weight method, and further calculate the region weight set in the relationship set under different regions based on the weight of the index in the weather condition set;

[0026] S2.5, define the multi-granularity T-spherical fuzzy membership, and calculate the threshold value of all regions in the relationship set under different regions according to the definition and the relative loss function;

[0027] S2.6, integrate the threshold values of all regions to obtain a final threshold value according to the T-spherical fuzzy regret theory;

[0028] S2.7, generate the fire occurrence possibility ranking results of all dates in the date set according to the decision rule, the score function, the adjustable multi-granularity T-spherical fuzzy conditional probability, and the final threshold value.

[0029] The method for generating the conditional probability and the adjustable conditional probability of all dates in the date set is:

[0030] The similarity classes of all dates in the date set are calculated by using a similarity-based incomplete T-spherical fuzzy processing method, and the similarity-based incomplete T-spherical fuzzy processing method is described as follows:

[0031] First, the similarity and the difference between each date in the date set are calculated, and the calculation formula is:

[0032]

[0033] wherein, and represent the first region in the first the similarity and difference between the first date and the second date under the weather condition index, representing missing values, representing the domain of the weather condition index in the th region, representing the number of different values in the domain;

[0034] Next, the T-spherical fuzzy similarity relation is defined based on the above calculation results, and the definition is as follows:

[0035]

[0036] wherein, is the T-spherical fuzzy similarity relation between the first day and the second day in the th region, is the number of weather condition indexes;

[0037] According to the T-spherical fuzzy similarity relation, the th order T-spherical fuzzy similarity relation and the similarity class are further defined, and the definition is as follows:

[0038]

[0039] wherein, is the th order T-spherical fuzzy similarity relation between the first date and the second date in the th region, satisfies and , is the similarity class of the first region in the relation set under the different regions;

[0040] According to the definition of the similarity class, the conditional probability and the adjustable conditional probability of all dates in the date set are generated, and the conditional probability is defined as follows:

[0041]

[0042] wherein, represents the pre-decision result, i.e. preliminary ranking of alternative schemes under a certain decision preference, and selecting the top ranked multiple alternative schemes as the pre-decision result, represents the conditional probability of the th date in the th region;

[0043] For all conditional probabilities, the ascending order arrangement can obtain the adjustable conditional probability , which represents the adjustable conditional probability of the region ranked in the th place. ​

[0044] The method for transforming the incomplete T-spherical fuzzy relation set into a complete T-spherical fuzzy relation set using the similarity-based completion algorithm is as follows:

[0045] Step 1: Numerical scanning is performed on the incomplete T-spherical fuzzy relation set to establish a missing value marking matrix When there is a missing value in the i-th row and j-th column, , , otherwise, ;

[0046] Step 2: Define the marking matrix , representing the i-th row and the j-th row have no missing values or the missing values of both are in the same position, and in other cases ;

[0047] Step 3: Based on the similarity-based calculation method, the similarity matrix is obtained , where the initial value is 1;

[0048] Step 4: The similarity matrix is scanned by row, and the maximum similarity is obtained by calculating the score function (values of 0 and 1 are not considered). The maximum value needs to satisfy and at least one value is 1 and ; if there are multiple maximum similarities obtained, the first maximum similarity obtained is selected for calculation;

[0049] Step 5: Based on the above definition, the incomplete T-spherical fuzzy relation set is transformed into a complete T-spherical fuzzy relation set using the following steps:

[0050] Step 5.1: ;

[0051] Step 5.2: If and , then ,

[0052] If and , then ,

[0053] If and , then go to Step 5.3;

[0054] Step 5.3: ;

[0055] Step 5.4: If , then go to Step 5.2,​

[0056] If , then , then ;

[0057] Step 6: Recalculate the similarity of and and obtain by replacing the corresponding values in ; ;

[0058] Step 7: Scan the first row and the first row in the incomplete T-spherical fuzzy relation set to check for missing values and update the flag matrix ;

[0059] Step 8: If there are still missing values in the T-spherical fuzzy relation set, go to Step 5; otherwise, obtain the complete T-spherical fuzzy relation set, and the completion method ends.

[0060] The method for calculating the index weight set in the weather condition set according to the complete T-spherical fuzzy relation set and the entropy weight method is:

[0061] The entropy weight method first calculates the information entropy of each index in the weather condition set , where ; when , let , and based on the definition of information entropy, the index weight in the weather condition set can be further obtained ;

[0062] On this basis, the regional weight can be further obtained, and the calculation steps are similar to the index weight in the weather condition set described above; first, calculate the information entropy of each region , where , and then the regional weight can be obtained .

[0063] The multi-granularity T-spherical fuzzy membership is defined as follows:

[0064]

[0065] where is the multi-granularity T-spherical fuzzy membership, is the relation value of the first region, the first index, and the first date in the complete T-spherical fuzzy relation set, is the standard set under the first index.

[0066] According to the definition and the relative loss function, the threshold of all regions in the relationship set in different regions is calculated, and the threshold calculation formula is as follows:

[0067] ,

[0068] ,

[0069] Wherein is the risk aversion coefficient, representing the degree of risk aversion, The greater the risk aversion is lower. After calculation, the threshold matrix is And .

[0070] The method for integrating the thresholds of all regions according to the T-sphere fuzzy regret theory to obtain the final threshold is:

[0071] Step 1: using the utility function Calculate the utility value of , and arrange it in ascending order;

[0072] Step 2: calculate the joy-regret value of The joy-regret function is , wherein ,. When , it represents the joy value, and when , it represents the regret value;

[0073] Step 3: calculate the perceived utility value of , and the perceived utility function is ;

[0074] Step 4: calculate the overall perceived utility value of the threshold in the th region ;

[0075] Step 5: by comparing the overall perceived utility values of all regions, the maximum value is obtained, and the threshold of the region corresponding to the maximum value is selected as the final threshold.

[0076] The method for generating the fire occurrence possibility sorting result of all dates in the date set is:

[0077] ,

[0078] ,

[0079] ,

[0080] Wherein , , The positive domain, the negative domain and the boundary domain are represented by +, -, and 0, respectively.

[0081] The score function is represented as , wherein , The comparison rule based on the score function is as follows:

[0082] ,

[0083] ,

[0084] .

[0085] A device for constructing a forest fire warning system, the device comprising:

[0086] an acquisition module, configured to acquire a set of weather conditions, a set of dates corresponding to the set of weather conditions, and a set of relationships in different regions corresponding to the set of weather conditions and the set of dates;

[0087] a first generation module, configured to generate an incomplete T-spherical fuzzy relation set according to the set of weather conditions, the set of dates, and the set of relationships in different regions;

[0088] a conversion module, configured to convert the incomplete T-spherical fuzzy relation set into a complete T-spherical fuzzy relation set by constructing a completion algorithm;

[0089] a second generation module, configured to generate a set of fire warning standards according to the complete T-spherical fuzzy relation set;

[0090] a construction module, configured to construct a forest fire warning system according to the set of weather conditions, the set of dates, the set of relationships in different regions, the incomplete T-spherical fuzzy relation set, and the set of fire warning standards;

[0091] The decision device comprises:

[0092] a processing module, configured to input the set of weather conditions, the set of dates corresponding to the set of weather conditions, and the set of relationships in different regions corresponding to the set of weather conditions and the set of dates into the forest fire warning system for data initialization processing, and generate an incomplete T-spherical fuzzy relation set after processing;

[0093] a third generation module, configured to calculate similarity classes of all dates in the set of dates by using an incomplete T-spherical fuzzy processing method based on similarity, and generate conditional probabilities and adjustable conditional probabilities of all dates in the set of dates according to the definition of the similarity classes;

[0094] The first calculation module is used for calculating a set of index weights in the set of weather conditions according to the complete T-sphere type fuzzy relation set and entropy weight method, and further calculating a set of region weights in the relation set in different regions based on the weights of indexes in the set of weather conditions;

[0095] The second calculation module is used for defining multi-granularity T-sphere type fuzzy membership, and calculating thresholds of all regions in the relation set in different regions according to the definition and a relative loss function;

[0096] The integration module is used for integrating the thresholds of all regions to obtain a final threshold according to T-sphere type fuzzy regret theory;

[0097] The fourth generation module is used for generating a fire occurrence possibility ranking result of all dates in the set of dates according to a decision rule, a score function, the adjustable multi-granularity T-sphere type fuzzy conditional probability and the final threshold.

[0098] Compared with the prior art, the present application has the beneficial effects that:

[0099] The present application uses a method based on fuzzy sets, which can cope with the incompleteness and uncertainty of the relationship between weather factors and fire occurrence possibility in the actual forest fire warning background, and improves the accuracy and problem processing efficiency of forest fire warning, and enables forest firefighters to intuitively understand the instant information related to forest fires, so as to make timely judgments on the current situation. BRIEF DESCRIPTION OF DRAWINGS

[0100] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for those skilled in the art, other drawings can be derived from the provided drawings without creative labor.

[0101] The structures, proportions, sizes, etc. shown in the specification are only used to cooperate with the content disclosed in the specification, to enable those skilled in the art to understand and read, and are not used to limit the limiting conditions under which the present application can be implemented, so they do not have technical significance. Any modification of structure, change of proportion relationship or adjustment of size, without affecting the effects and purposes that the present application can produce, should still fall within the scope of the technical content disclosed by the present application.

[0102] Figure 1 The flowchart of the method for constructing a forest fire warning system provided by Embodiment 1 of the present application;

[0103] Figure 2Flow chart of the similarity-based completion algorithm of the present application;

[0104] Figure 3 Detailed flow chart of step 5 in the similarity-based completion algorithm of the present application;

[0105] Figure 4 Flow chart of the decision-making method of the forest fire warning system constructed in Embodiment 1 of the present application;

[0106] Figure 5 Flow chart of the T-sphere fuzzy regret theory in the decision-making method of the present application;

[0107] Figure 6 Another flow chart of the decision-making method in Embodiment 1 of the present application;

[0108] Figure 7 Structural schematic diagram of a method for constructing a forest fire warning system provided by Embodiment 2 of the present application.

[0109] Figure 8 Structural schematic diagram of a decision-making device based on a forest fire warning system in Embodiment 2 of the present application. DETAILED DESCRIPTION

[0110] In order to make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments of the present application. These descriptions are only for further illustrating the features and advantages of the present application, but not for limiting the claims of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0111] The specific embodiments of the present application will be described in further detail below in combination with the drawings and embodiments. The following embodiments are used to illustrate the present application, but not to limit the scope of the present application.

[0112] The present application proposes a forest fire warning system construction method based on fuzzy sets, and uses the constructed forest fire warning system to solve the actual forest fire warning problem. Compared with the existing forest fire warning methods, the present application has the characteristics of using a method based on fuzzy sets, which can cope with the incompleteness and uncertainty of the relationship between weather factors and fire occurrence possibility in the actual forest fire warning background, and improve the accuracy and problem processing efficiency of forest fire warning, and enable forest firefighters to intuitively understand the real-time information related to forest fires, so as to make timely judgments on the current situation.

[0113] In addition, the application also provides a decision method based on the forest fire warning system, which selects three-branch decision model as a basic framework, adopts similarity and adjustable conditional probability to fuse different indexes of weather conditions, so that the final result is more accurate and reasonable, and on the other hand, for the threshold of different regions, the decision method selects to use T-spherical fuzzy regret theory for integration, which can solve the problem of bounded rationality in decision-making, and finally, based on the decision rule of three-branch decision, all dates are classified and sorted, and the concept of delayed decision is introduced in three-branch decision, so that the final decision result is more in line with human thinking and cognitive characteristics. The method for constructing a forest fire warning system, the decision method and the device thereof are described below with reference to the accompanying drawings.

[0114] Example 1

[0115] Figure 1 A flowchart of a method for constructing a forest fire warning system provided by Embodiment 1 of the application.

[0116] The method for constructing a forest fire warning system provided by Embodiment 1 of the application is shown in Figure 1 The method comprises the following steps:

[0117] Step 110, acquiring a weather condition set, a date set corresponding to the weather condition set, and a relationship set under different regions corresponding to the weather condition set and the date set;

[0118] Step 120, generating an incomplete T-spherical fuzzy relationship set according to the weather condition set, the date set and the relationship set under different regions;

[0119] Step 130, constructing a completion algorithm to convert the incomplete T-spherical fuzzy relationship set into a complete T-spherical fuzzy relationship set;

[0120] Step 140, generating a fire warning standard set according to the complete T-spherical fuzzy relationship set;

[0121] Step 150, constructing a forest fire warning system according to the weather condition set, the date set, the relationship set under different regions, the incomplete T-spherical fuzzy relationship set and the fire warning standard set.

[0122] In an embodiment of the application, the forest fire warning system can be represented as wherein, is a date set, and represents 61 alternative dates, is a weather condition set, wherein , , , respectively represent temperature, relative humidity, wind speed, rainfall, is a set of relations in four different regions, is a set of forest fire warning criteria.

[0123] In an embodiment of the present application, the process of generating the incomplete T-spherical fuzzy relation set in step 120 is as follows:

[0124] The initial data representing different indicators in the set of weather conditions are converted into membership degrees in T-spherical fuzzy numbers according to the linear normalization formula, which is expressed as:

[0125]

[0126] wherein, represents the membership degree of the i-th date and the j-th indicator in the i-th region after normalization, represents the relation of the i-th date and the j-th indicator in the i-th region in the initial data, represents the i-th region in the set of relations in different regions, represents the i-th date in the set of dates, represents the j-th indicator in the set of weather conditions; The non-membership degree and hesitation degree in T-spherical fuzzy numbers are generated according to the membership degrees in T-spherical fuzzy numbers and the conversion formula, which is:

[0127]

[0128]

[0129]

[0130] wherein, represents the non-membership degree of the i-th date and the j-th indicator in the i-th region after normalization, represents the hesitation degree of the i-th date and the j-th indicator in the i-th region after normalization;

[0131] For the missing values existing in the initial data of the set of weather conditions, the membership degree, non-membership degree and hesitation degree thereof are assigned as -1, 0 and 0.

[0132] ​​​​​​​​​​​​​​​The incomplete T-spherical fuzzy relation set is generated based on the membership degree, the non-membership degree, and the hesitation degree.

[0133] In one embodiment of this application, such as Figure 2 As shown, an incomplete T-spherical fuzzy relation set is transformed into a complete T-spherical fuzzy relation set using a completion algorithm:

[0134] Step 131: Scan the incomplete T-spherical fuzzy relation set and establish a missing value label matrix. When the first When there are missing values ​​in the row, ,otherwise, .

[0135] Step 132: Define the tag matrix , represent lines and The row has no missing values ​​or the missing values ​​of both exist in the same position; otherwise... ;

[0136] Step 133: Obtain the similarity matrix based on the similarity calculation method. ,in The initial value is 1;

[0137] Step 134: Scan the similarity matrix row by row, and obtain the maximum similarity by calculating the score function (ignoring values ​​of 0 and 1). This maximum value needs to satisfy... and At least one of them has a value of 1 and If there are multiple maximum similarity scores, the first maximum similarity score is selected for calculation.

[0138] Step 135: Complete the data. The detailed completion process for step 135 is as follows: Figure 3 As shown.

[0139] Step 136: ;

[0140] Step 137: Recalculate separately and and Similarity, and by replacement Obtain the corresponding value in ;

[0141] Step 138: Scan the first... in the currently incomplete T-spherical fuzzy relation set. row and number Rows are checked for missing values ​​and the label matrix is ​​updated. ;

[0142] Step 139: If there are still missing values ​​in the T-spherical fuzzy relation set, proceed to step 135; otherwise, a complete T-spherical fuzzy relation set is obtained, and the completion algorithm ends.

[0143] In one embodiment of this application, the fire early warning standard set generation process in step 140 is as follows:

[0144]

[0145] The results of the fire early warning standard set calculated using this formula are shown in the table below:

[0146]

[0147] Table 1 Standard Set

[0148] Figure 4 This is a flowchart of a decision-making method based on a forest fire early warning system.

[0149] This application provides a decision-making method based on a forest fire early warning system, such as... Figure 4 As shown, this decision-making method includes the following steps:

[0150] Step 210: Input the weather condition set, the date set corresponding to the weather condition set, and the relation set corresponding to different regions under the weather condition set and date set into the forest fire early warning system for data initialization processing, and generate an incomplete T-spherical fuzzy relation set after processing.

[0151] Step 220: Calculate the similarity class of all dates in the date set using a similarity-based incomplete T-spherical fuzzy processing method, and generate the conditional probability and adjustable conditional probability of all dates in the date set according to the definition of the similarity class. The similarity-based incomplete T-spherical fuzzy processing method is described as follows:

[0152] First, the similarity and difference between the dates in the date set are calculated, where the calculation formula is:

[0153]

[0154] in, and Representing the first The first in the region The similarity and difference between the first and second dates under each weather condition indicator. Represents missing values. Representing the The first in the region A domain of weather condition indicators, This represents the number of distinct values ​​in the domain;

[0155] Next, the T-spherical fuzzy similarity relationship is calculated:

[0156]

[0157] in, For the first T-spherical fuzzy similarity relationship between the first and second days in a region The number of weather condition indicators;

[0158] Further calculations were performed based on the T-spherical fuzzy similarity relation. Level T-spherical fuzzy similarity relations and similarity classes:

[0159]

[0160]

[0161] in, For the first The first and second dates in each region Level T-spherical fuzzy similarity relation satisfy and , This refers to the similarity class of the first region in the relation set across the different regions.

[0162] The similarity calculation results are shown in Table 2 below. Due to the large amount of data, only the similarity calculation results for the first 5 candidate dates are shown here:

[0163]

[0164] Table 2 Similarity class

[0165] Based on the definition of similarity classes, the conditional probability and adjustable conditional probability of all dates in the date set are calculated. The formula for calculating the conditional probability is as follows:

[0166]

[0167] in, This represents the preliminary decision outcome, which involves initially ranking the alternatives under a certain decision preference and selecting the top-ranked alternatives as the preliminary decision outcome. Representing the The first in the region The conditional probability of a date;

[0168] The adjustable conditional probability can be obtained by sorting all conditional probabilities in ascending order. , represents the adjustable conditional probability of the region ranked in the first place, the application assigns the parameter based on the risk-neutral decision preference .

[0169] Step 230: using a similarity-based completion algorithm to convert the incomplete T-spherical fuzzy relation set into a complete T-spherical fuzzy relation set, the completion algorithm is shown in Figure 2 and Figure 3 .

[0170] Step 240: according to the complete T-spherical fuzzy relation set and the entropy weight method, the index weight set in the weather condition set is calculated, and further based on the weight of the index in the weather condition set, the region weight set in the relation set under different regions is calculated, and the entropy weight method is described as follows:

[0171] The entropy weight method first calculates the information entropy of each index in the weather condition set , wherein . When , let , based on the definition of information entropy, the index weight in the weather condition set can be further obtained , and the calculation results of the weather condition index weight are shown in Table 3:

[0172]

[0173] Table 3 Weather condition index weight set

[0174] On this basis, the region weight can be further obtained, and the calculation steps are similar to the index weight in the weather condition set. First, calculate the information entropy of each region , wherein . Next we can get the region weight , and the calculation results of the region weight are shown in Table 4:

[0175]

[0176] Table 4 Region weight set

[0177] Step 250: define the multi-granularity T-spherical fuzzy membership, and calculate the threshold of all regions in the relation set under different regions according to the definition and the relative loss function, wherein the multi-granularity T-spherical fuzzy membership calculation formula is as follows:

[0178]

[0179] , wherein is the multi-granularity T-spherical fuzzy membership, To complete the T-spherical fuzzy relation set of the relation value of the first index and the first date in the first region, the relation value of the first index and the first date in the second region, and the relation value of the first index and the first date in the third region, a standard set of the first index. According to the definition and the relative loss function, the threshold values of all regions in the relation set in different regions are calculated, and the threshold value calculation formula is as follows:

[0180]

[0181]

[0182]

[0183] wherein is a risk-aversion coefficient, representing the degree of risk aversion, the greater the degree of risk aversion. After calculation, the threshold matrix is and . In this application, the parameter is assigned a value of based on the risk-neutral decision preference.

[0184] Step 260: obtaining the final threshold value according to the integration of the threshold values of all regions by T-spherical fuzzy regret theory, wherein the T-spherical fuzzy regret theory is as shown in Figure 5 and described as follows:

[0185] First, the utility value of is calculated using the utility function , and is arranged in ascending order;

[0186] Next, the joy-regret value of is calculated. The joy-regret function is , wherein , when , represents the joy value, and when , represents the regret value;

[0187] Then, the perceived utility value of is calculated, and the perceived utility function is ;

[0188] Next, the overall perceived utility value of the threshold value in the first region is calculated;

[0189] Finally, the maximum value of the overall perceived utility value of all regions is obtained by comparison, and the threshold value of the region corresponding to the maximum value is selected as the final threshold value.

[0190] In this application, the parameter in the regret theory is assigned a value of​​​​​​ and .

[0191] Step 270: generating the ranking results of fire occurrence likelihood of all dates in the date set according to the decision rule, the score function, the adjustable multi-granularity T-sphere type fuzzy conditional probability and the final threshold, wherein the decision rule is defined as follows:

[0192] ,

[0193] ,

[0194] ,

[0195] wherein , , respectively represent the positive domain, the negative domain and the boundary domain.

[0196] The score function is expressed as , wherein , The comparison rule based on the score function is as follows:

[0197] ,

[0198] ,

[0199] .

[0200] The classification results are shown in Table 5 as follows:

[0201]

[0202] The final ranking results of all alternative dates are as follows:

[0203]

[0204] In summary, the embodiment of the present application is based on a decision-making method of a forest fire warning system. First, the weather condition set, the date set corresponding to the weather condition set, and the relationship set in different regions corresponding to the weather condition set and the date set are input into the forest fire warning system for data initialization processing, and an incomplete T-spherical fuzzy relationship set is generated after processing. Then, the similarity-based incomplete T-spherical fuzzy processing method is used to calculate the similarity classes of all dates in the date set, and the conditional probability and the adjustable conditional probability of all dates in the date set are generated according to the definition of the similarity class. Next, the similarity-based completion method is used to convert the incomplete T-spherical fuzzy relationship set into a complete T-spherical fuzzy relationship set. Next, the index weight set in the weather condition set is calculated according to the complete T-spherical fuzzy relationship set and the entropy weight method, and the region weight set in the relationship set in different regions is further calculated based on the weight of the index in the weather condition set. Next, the multi-granularity T-spherical fuzzy membership is defined, and the threshold values of all regions in the relationship set in different regions are calculated according to the definition and the relative loss function, and the threshold values of all regions are integrated to obtain the final threshold value according to the T-spherical fuzzy regret theory. Finally, the fire occurrence possibility ranking results of all dates in the date set are generated according to the decision rule, the score function, the adjustable multi-granularity T-spherical fuzzy conditional probability, and the final threshold value. Therefore, the similarity-based completion method proposed in the present application can reasonably and accurately process incomplete information, so that the constructed forest fire warning system can cope with the case of missing values in the initial data. The weight set calculated by the entropy weight method accurately represents the differences between different indicators in different regions and weather conditions, and the threshold values are integrated using the T-spherical fuzzy regret theory to compensate for the limitations of decision makers, so that the final forest fire warning result is more objective.

[0205] Example 2

[0206] To achieve the above embodiment, the embodiment of the present application proposes a device for constructing a forest fire warning system, as shown in Figure 7 The device comprises:

[0207] The acquisition module 10 is configured to acquire a weather condition set, a date set corresponding to the weather condition set, and a relationship set in different regions corresponding to the weather condition set and the date set.

[0208] The first generation module 20 is configured to generate an incomplete T-spherical fuzzy relationship set according to the weather condition set, the date set, and the relationship set in different regions.

[0209] The conversion module 30 is configured to convert the incomplete T-spherical fuzzy relationship set into a complete T-spherical fuzzy relationship set by constructing a completion method.

[0210] The second generating module 40 is configured to generate a fire warning standard set according to the complete T-sphere fuzzy relation set.

[0211] The constructing module 50 is configured to construct a forest fire warning system according to the weather condition set, the date set, the relation set in different regions, the incomplete T-sphere fuzzy relation set and the fire warning standard set.

[0212] Since the embodiments of the present application perform the method of constructing the forest fire warning system in the embodiment 1, the technical effects achieved by the embodiments of the present application are similar to those achieved by the above method. To avoid repetition, details are not described herein.

[0213] In order to achieve the above-mentioned embodiments, the embodiments of the present application propose a decision-making device based on a forest fire warning system, as shown in Figure 8 The decision-making device comprises:

[0214] The processing module 50 is configured to input the weather condition set, the date set corresponding to the weather condition set, and the relation set in different regions corresponding to the weather condition set and the date set into the forest fire warning system for data initialization processing, and generate an incomplete T-sphere fuzzy relation set after processing.

[0215] The third generating module 60 is configured to calculate the similarity classes of all dates in the date set by using the incomplete T-sphere fuzzy processing method based on similarity, and generate the conditional probability and the adjustable conditional probability of all dates in the date set according to the definition of the similarity class.

[0216] The first calculating module 70 is configured to calculate the index weight set in the weather condition set according to the complete T-sphere fuzzy relation set and the entropy weight method, and further calculate the region weight set in the relation set in different regions based on the weight of the index in the weather condition set.

[0217] The second calculating module 80 is configured to define the multi-granularity T-sphere fuzzy membership degree, and calculate the threshold value of all regions in the relation set in different regions according to the definition and the relative loss function.

[0218] The integrating module 90 is configured to integrate the threshold values of all regions to obtain a final threshold value by using the T-sphere fuzzy regret theory.

[0219] The fourth generating module 100 is configured to generate the fire occurrence possibility ranking result of all dates in the date set according to the decision-making rule, the score function, the adjustable multi-granularity T-sphere fuzzy conditional probability and the final threshold value.

[0220] Since the embodiments of the present application execute the decision-making method based on the forest fire warning system in embodiment 1, the technical effects achieved by the embodiments of the present application are similar to the technical effects achieved by the above method, and to avoid repetition, details are not described herein.

[0221] In order to achieve the above-mentioned embodiments, the embodiments of the present application propose a computer device, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the method for constructing the forest fire warning system and the decision-making method based on the forest fire warning system in embodiment 1 are implemented.

[0222] Since the embodiments of the present application execute the method for constructing the forest fire warning system and the decision-making method based on the forest fire warning system in embodiment 1, to avoid repetition, the technical effects achieved by the embodiments of the present application are not described herein.

[0223] In the description of the present specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example" or "some examples" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the present specification and the features of the different embodiments or examples without contradiction.

[0224] Any process or method descriptions in flow charts or described elsewhere herein can be understood as representing code modules, segments, or portions of code which include one or more executable instructions for implementing specific logic functions (or steps) in the process, and that the various embodiments of the preferred implementation include other alternative implementations, in which the order of steps can differ (including simultaneously or in reverse order) depending upon the functionality involved. These descriptions and representations are used by those skilled in the art of software formation to most effectively convey the substance of their work to others skilled in the art.

[0225] The logic and / or steps represented in the flowcharts and / or described herein, for example, can be considered as a sequence of instructions to implement logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, apparatus, or device, such as a computer-based system, processor- based system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For purposes of this specification, a "computer-readable medium" can be any apparatus that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. The computer-readable medium can be a computer- readable storage medium or a computer-readable signal medium. The computer- readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, and a portable compact disc read-only memory (CD-ROM). In addition, the computer-readable medium can even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, for example, via optical scanning of the paper or other medium, then compiled, interpreted, or otherwise processed in a suitable manner, if necessary, and then stored in a computer memory.

[0226] It should be understood that aspects of the application can be implemented in hardware, software, firmware or combinations thereof. In the above embodiments, the various steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. As such, if implemented in hardware and in another embodiment, any of the following technologies, known in the art, or their combinations can be used: discrete logic circuitry having logic gates for implementing logic functions on data signals, application specific integrated circuits having appropriate combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), and the like.

[0227] Those of skill in the art would understand that the various steps carried out in the above-mentioned embodiment methods can be carried out by programs instructing relevant hardware, and the programs can be stored in a computer-readable storage medium. When the programs are executed, they include one or a combination of the steps of the method embodiments.

[0228] In addition, each of the function units in the various embodiments of the present application can be integrated in one processing module, or each unit can be physically present separately, or two or more units can be integrated in one module. The integrated module can be realized in the form of hardware or in the form of a software function module. When the integrated module is realized in the form of a software function module and sold or used as an independent product, it can also be stored in a computer readable storage medium. The above only describes the preferred embodiments of the present application in detail, but the present application is not limited to the above embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the spirit of the present application, and all changes shall be included in the protection scope of the present application.

Claims

1. A method of constructing a forest fire warning system, characterized by: Includes the following steps: S1.1 Obtain a weather condition set, a date set corresponding to the weather condition set, and a relationship set for different regions corresponding to the weather condition set and the date set; S1.2 Generate an incomplete T-spherical fuzzy relation set based on the weather condition set, the date set, and the relation sets under different regions; S1.3 Construct a completion method to transform the incomplete T-spherical fuzzy relation set into a complete T-spherical fuzzy relation set; S1.

4. Generate a fire early warning standard set based on the complete T-spherical fuzzy relation set; S1.5 Construct a forest fire early warning system based on the weather condition set, the date set, the relation set under different regions, the incomplete T-spherical fuzzy relation set, and the fire early warning standard set.

2. The method for constructing a forest fire early warning system according to claim 1, characterized in that: The method for generating an incomplete T-spherical fuzzy relation set based on the weather condition set, the date set, and the relation sets under different regions is as follows: The initial data representing different indicators of the weather condition set are transformed into membership degrees in T-spherical fuzzy numbers according to the linear normalization formula, wherein the linear normalization formula is expressed as: in, Represents the normalized first The first in each region The date and the Membership degree of each indicator Represents the first in the initial data The first in each region The date and the The relationship between the indicators Relationships representing different regions are concentrated in the first Each region Representative Date Collection No. Date, Representing the concentrated weather conditions One indicator; The non-membership degree and hesitation degree in the T-spherical fuzzy number are generated based on the membership degree and transformation formula in the aforementioned T-spherical fuzzy number, wherein the transformation formula is: in, Represents the normalized first The first in each region The date and the Non-membership degree of each indicator Represents the normalized first The first in each region The date and the Hesitation level of each indicator; For missing values ​​in the initial data of the weather condition set, their membership degree, non-membership degree, and hesitation degree are assigned the values ​​-1, 0, and 0, respectively; The incomplete T-spherical fuzzy relation set is generated based on the membership degree, the non-membership degree, and the hesitation degree.

3. A decision-making method for constructing a forest fire early warning system, characterized in that: Includes the following steps: S2.1 Input the weather condition set, the date set corresponding to the weather condition set, and the relation set under different regions corresponding to the weather condition set and the date set into the forest fire early warning system for data initialization processing, and generate an incomplete T-spherical fuzzy relation set after processing; S2.2 Calculate the similarity class of all dates in the date set using the incomplete T-spherical fuzzy processing method based on similarity, and generate the conditional probability and adjustable conditional probability of all dates in the date set according to the definition of the similarity class; S2.

3. Use a similarity-based completion method to transform the incomplete T-spherical fuzzy relation set into a complete T-spherical fuzzy relation set; S2.

4. Calculate the index weight set of the weather condition set based on the complete T-spherical fuzzy relation set and the entropy weight method, and further calculate the regional weight set of the relation set under different regions based on the weight of the weather condition set index. S2.5 Define the multi-granularity T-spherical fuzzy membership degree, and calculate the threshold of all regions in the relation set under the different regions based on this definition and the relative loss function; S2.

6. The thresholds for all regions are integrated using T-spherical fuzzy regret theory to obtain the final threshold. S2.

7. Based on the decision rules, scoring function, adjustable multi-granularity T-spherical fuzzy conditional probability, and final threshold, generate the fire occurrence probability ranking results for all dates in the date set.

4. The decision-making method for constructing a forest fire early warning system according to claim 3, characterized in that: The method for generating the conditional probabilities and adjustable conditional probabilities of all dates in the date set is as follows: The similarity classes of all dates in the date set are calculated using a similarity-based incomplete T-spherical fuzzing method, which is described below: First, the similarity and difference between the dates in the date set are calculated, where the calculation formula is: in, and Representing the first The first in the region The similarity and difference between the first and second dates under each weather condition indicator. Represents missing values. Representing the The first in the region A domain of weather condition indicators, This represents the number of distinct values ​​in the domain; Next, based on the above calculation results, the T-spherical fuzzy similarity relation is defined as follows: in, For the first The T-spherical fuzzy similarity relationship between day 1 and day 2 in each region. The number of weather condition indicators; Further definition based on T-spherical fuzzy similarity relation The T-level spherical fuzzy similarity relation and similarity class are defined as follows: in, For the first The first and second dates in each region Level T-spherical fuzzy similarity relation satisfy and , This refers to the similarity class of the first region in the relation set across the different regions. Based on the definition of similarity classes, generate conditional probabilities and adjustable conditional probabilities for all dates in the date set. The conditional probabilities are defined as follows: in, This represents the preliminary decision outcome, which involves initially ranking the alternatives under a certain decision preference and selecting the top-ranked alternatives as the preliminary decision outcome. Representing the The first in the region The conditional probability of a date; The adjustable conditional probability can be obtained by sorting all conditional probabilities in ascending order. , representing the number Adjustable conditional probability of a region.

5. The decision-making method for constructing a forest fire early warning system according to claim 4, characterized in that: The method for transforming an incomplete T-spherical fuzzy relation set into a complete T-spherical fuzzy relation set using a similarity-based completion algorithm is as follows: Step 1: Perform a numerical scan on the incomplete T-spherical fuzzy relation set and establish a missing value label matrix. When the first When there are missing values ​​in the row, ,otherwise, ; Step 2: Define the tag matrix , represent lines and The row has no missing values ​​or the missing values ​​of both exist in the same position; otherwise... ; Step 3: Based on the aforementioned similarity calculation method, obtain the similarity matrix. ,in The initial value is 1; Step 4: Scan the similarity matrix row by row, and calculate the maximum similarity by using a score function (ignoring values ​​of 0 and 1). This maximum similarity must satisfy the following condition. and At least one of them has a value of 1 and If there are multiple maximum similarity scores, the first maximum similarity score is selected for calculation. Step 5: Based on the above definition, the incomplete T-spherical fuzzy relation set is transformed into a complete T-spherical fuzzy relation set using the following steps: Step 5.1: ; Step 5.2: If and ,but , if and ,but , if and Then proceed to step 5.3; Step 5.3: ; Step 5.4: If Then proceed to step 5.

2. if , but , but ; Step 6: Recalculate separately and and Similarity, and by replacement Obtain the corresponding value in ; Step 7: Scan the first digit of the currently incomplete T-spherical fuzzy relation set. row and number Rows are checked for missing values ​​and the label matrix is ​​updated. ; Step 8: If there are still missing values ​​in the T-spherical fuzzy relation set, go to step 5; otherwise, obtain the complete T-spherical fuzzy relation set, and the completion method ends.

6. The decision-making method for constructing a forest fire early warning system according to claim 3, characterized in that: The method for calculating the index weight set in the weather condition set based on the complete T-spherical fuzzy relation set and the entropy weight method is as follows: The entropy weight method first calculates the information entropy of each indicator in the weather condition set. ,in ;when season Based on the definition of information entropy, the index weights of the weather condition set can be further obtained. ; Based on this, regional weights can be further obtained, and the calculation steps are similar to those for the indicator weights of the weather condition set mentioned above; first, the information entropy of each region is calculated. ,in Next, the regional weights can be obtained. .

7. The decision-making method for constructing a forest fire early warning system according to claim 3, characterized in that: The multi-granularity T-spherical fuzzy membership degree is defined as follows: in For multi-granularity T-spherical fuzzy membership, To complete the T-spherical fuzzy relation set, the first The first in each region The first indicator and the first The relational value of a date, For the first Standard set under each indicator; Based on this definition and the relative loss function, the threshold for all regions in the relation set under different regions is calculated. The threshold calculation formula is as follows: ， ， in The risk aversion coefficient represents the degree of risk aversion. The larger the threshold, the lower the risk aversion. The calculated threshold matrix is ​​as follows: and .

8. The decision-making method for constructing a forest fire early warning system according to claim 3, characterized in that: The method for obtaining the final threshold by integrating the thresholds of all regions using T-spherical fuzzy regret theory is as follows: Step 1: Use utility functions calculate Calculate their utility values ​​and sort them in ascending order; Step 2: Calculation The euphoria-regret value, the euphoria-regret function is: ,in ,when When the value of joy is expressed, Time represents the regret value; Step 3: Calculation The perceived utility value, the perceived utility function is ; Step 4: Calculate the first... Overall perceived utility value at the threshold in each region ; Step 5: Obtain the maximum value among all regions by comparing the overall perceived utility values, and select the threshold value of the region corresponding to the maximum value as the final threshold value.

9. The decision-making method for constructing a forest fire early warning system according to claim 3, characterized in that: The method for generating the fire probability ranking results for all dates in the date set is as follows: ， ， ， in , , These represent the positive domain, negative domain, and boundary domain, respectively. The scoring function is expressed as , ,in , The comparison rules based on the scoring function are as follows: ， ， 。 10. A device for constructing a forest fire early warning system, characterized in that: The device includes: The acquisition module retrieves a weather condition set, a date set corresponding to the weather condition set, and a set of relationships for different regions corresponding to the weather condition set and the date set. The first generation module is used to generate an incomplete T-spherical fuzzy relation set based on the weather condition set, the date set, and the relation sets under different regions; The transformation module is used to construct a completion algorithm to transform the incomplete T-spherical fuzzy relation set into a complete T-spherical fuzzy relation set; The second generation module is used to generate a fire early warning standard set based on the complete T-spherical fuzzy relation set; The construction module is used to construct a forest fire early warning system based on the weather condition set, the date set, the relation set under different regions, the incomplete T-spherical fuzzy relation set, and the fire early warning standard set; The decision-making device includes: The processing module is used to input the weather condition set, the date set corresponding to the weather condition set, and the relation set under different regions corresponding to the weather condition set and the date set into the forest fire early warning system for data initialization processing, and generate an incomplete T-spherical fuzzy relation set after processing; The third generation module is used to calculate the similarity class of all dates in the date set using the incomplete T-spherical fuzzing processing method based on similarity, and to generate the conditional probability and adjustable conditional probability of all dates in the date set according to the definition of the similarity class. The first calculation module is used to calculate the index weight set of the weather condition set based on the complete T-spherical fuzzy relation set and the entropy weight method, and further calculate the regional weight set of the relation set under different regions based on the weight of the weather condition set index. The second calculation module is used to define the multi-granularity T-spherical fuzzy membership degree, and calculate the threshold of all regions in the relation set under the different regions according to this definition and the relative loss function; The integration module is used to integrate the thresholds of all regions using T-spherical fuzzy regret theory to obtain the final threshold; the fourth generation module is used to generate the fire occurrence probability ranking results for all dates in the date set based on the decision rules, scoring function, adjustable multi-granularity T-spherical fuzzy conditional probability and the final threshold.

Images