Photovoltaic station geological disaster risk evaluation method and system

Abstract

The invention discloses a method for evaluating the risk of geological disasters in a photovoltaic station, and the method achieves the dynamic adjustment of a combination weight through iterative optimization calculation, carries out the geological disaster risk evaluation and partitioning of a photovoltaic station region through a clustering method under the optimal combination weight, and obtains a geological disaster risk partitioning map. Through dynamic adjustment, the prediction precision of the weight is improved, the dynamic adaptability of the weight is better, the method can adapt to the requirements of multi-source data fusion and dynamic weight optimization of the photovoltaic station, and compared with a traditional method, the method significantly improves the prediction precision and accuracy of the geological disaster risk partition, and improves the prediction efficiency of the geological disaster risk partition. Technical support is provided for photovoltaic station full life cycle danger management, and prediction precision and accuracy are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of geological disaster risk assessment, and relates to a method and system for assessing the geological disaster risk of a photovoltaic station. Background Art

[0002] As the global energy mix shifts toward a low-carbon economy, photovoltaic power generation is becoming a crucial form of clean energy. According to statistics, my country's installed photovoltaic power station capacity has exceeded 400 million kilowatts, and projects are primarily located in areas with complex terrain and variable geological conditions. These areas face the threat of geological hazards such as landslides and mudslides during development, which not only impact power plant operations but also pose ecological risks. Traditional geological hazard assessment methods, which rely primarily on expert experience and single models, are ill-suited to the requirements of multi-source data integration and dynamic weight optimization for photovoltaic sites.

[0003] Existing research indicates that the core of geological hazard assessment models lies in the scientific allocation of indicator weights. The analytic hierarchy process (AHP) uses expert experience weights to generate a judgment matrix to determine subjective weights, but this method suffers from issues such as reliance on empirical judgment and a high degree of subjectivity. The entropy weight method objectively assigns weights based on data, but it neglects the guiding role of domain knowledge. In recent years, some researchers have attempted to incorporate machine learning algorithms into weight optimization, but the parameter tuning process often employs grid search or random sampling, resulting in low computational efficiency. Furthermore, existing models often employ a linear weighted combination of subjective and objective weights, failing to consider the dynamic adaptability of weight allocation, resulting in limited prediction accuracy in complex scenarios. Summary of the Invention

[0004] The purpose of the present invention is to solve the problem that the dynamic adaptability of weight distribution is not considered in the existing technology, the prediction accuracy is limited in complex scenarios, resulting in low prediction accuracy and precision, and it is difficult to adapt to the needs of multi-source data fusion and dynamic weight optimization of photovoltaic stations. A method and system for geological hazard assessment of photovoltaic stations are provided.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A method for assessing geological hazards of photovoltaic stations comprises the following steps:

[0007] Construct subjective and objective weights for geological hazard risk assessment of photovoltaic stations;

[0008] A combined weight optimization model is constructed based on the subjective weight and the objective weight. The initial AUC value of the combined weight is calculated based on the combined weight optimization model. The initial data set D is constructed according to the initial AUC value.

[0009] An objective function is constructed based on the initial AUC value and the initial data set D. The initial AUC value and the initial data set D are iteratively calculated according to the objective function to obtain an updated AUC value and data set D. The optimal combination weight is obtained based on the updated AUC value and data set D. Layer overlay calculation is performed based on the optimal combination weight to obtain the calculation results. The calculation results are clustered and analyzed to obtain the geological hazard assessment results of the photovoltaic station.

[0010] A further improvement of the present invention is:

[0011] The subjective weights for the geological hazard assessment of photovoltaic stations are constructed, including:

[0012] Establish a geological disaster risk indicator system for photovoltaic sites;

[0013] Based on the geological hazard risk index system of photovoltaic stations, the importance coefficient matrix X of geological hazard risk index of photovoltaic stations is generated. M ;

[0014] Based on the importance coefficient matrix X of geological disaster risk indicators of photovoltaic stations M Generate the initial judgment matrix A M , calculate the initial judgment matrix A M Repeat this step to iteratively generate multiple groups of evaluation index initial judgment matrix AM and consistency ratio CR. M and the consistency ratio CR to generate the initial point data set [X initial ,Y initial ];

[0015] Through iterative optimization, update the initial point data set [X initial ,Y initial ], get the updated judgment matrix A M and consistency index ratio CR;

[0016] Determine whether the consistency index ratio CR meets the constraint conditions. If so, output the judgment matrix A corresponding to the current consistency index ratio CR M If not, the judgment matrix A M Adjust the elements in until the consistency index ratio CR satisfies the constraint conditions, and output the current corresponding judgment matrix A M ;

[0017] According to the final output judgment matrix A M Get its corresponding subjective weight.

[0018] By iterative optimization, the initial point data set [X initial ,Y initial], get the updated judgment matrix A M And consistency index ratio CR, including:

[0019] Construct the fitting objective function:

[0020]

[0021] Where C * is the penalty coefficient;

[0022] Construct constraints for the function;

[0023] IF=tf(1,2)*tf(1,3)*tf(1,4)*……*tf(18,19)

[0024] Where tf(i,j) represents element a (i,j) The constraint condition in the generation process indicates that the importance coefficient of the i-th indicator is consistent with that of the j-th indicator, that is, I i =I j ;

[0025] Define the acquisition function:

[0026] EI constrained (x)=EI(x)·IF

[0027] EI(x)=(μ(x)-y best -ξ)Φ(Z)+σ(x)φ(Z)

[0028]

[0029] Where Φ and φ are the cumulative distribution function CDF and probability density function PDF of the standard normal distribution, and ξ is the exploration parameter;

[0030] Iterative optimization:

[0031] The initial point data set [X initial ,Y initial ] as the input of the fitting objective function to generate the judgment matrix A M , combined with the acquisition function, calculate the current judgment matrix A M EI value; select the x point with the maximum EI value next , combined with the fitting objective function, calculate the y value corresponding to the maximum point; based on the obtained x next and y values, update the point dataset [X initial ,Y initial ] and EI model, repeat this step until the preset number of iterations is reached, and the latest point data set [X initial ,Y initial ], and based on the latest point dataset [X initial,Y initial ] Get the corresponding consistency index ratio CR and judgment matrix A M .

[0032] The judgment consistency index ratio CR satisfies the constraint condition. If so, the judgment matrix A corresponding to the current consistency index ratio CR is output. M If not, the judgment matrix A M Adjust the elements in until the consistency index ratio CR satisfies the constraint conditions, and output the current corresponding judgment matrix A M ,include:

[0033] Set the consistency index ratio CR constraint to 0.1. When the output consistency index ratio CR is greater than 0.1, the element x in the judgment matrix (i,j) Make adjustments;

[0034] Based on the adjusted judgment matrix, recalculate its corresponding consistency index ratio CR until the consistency index ratio CR is less than or equal to 0.1, and output the current corresponding judgment matrix A M .

[0035] The objective weights for the geological hazard assessment of photovoltaic stations are constructed, including:

[0036] Establish a geological disaster risk indicator system for photovoltaic sites;

[0037] Based on the relationship between the indicators in the geological hazard index system of photovoltaic stations, the original data matrix is constructed;

[0038] Based on the original data moment, the weight of each indicator is calculated, and the objective weight is obtained according to the weight of each indicator.

[0039] The construction of a combined weight optimization model based on subjective weights and objective weights includes:

[0040] Introducing parameter ν∈(0,1) to convert the subjective weight W AHP and objective weight W Entropy Fusion via linear combination:

[0041] W comb =ν·W AHP +(1-ν)·W Entropy

[0042] The subjective weight W AHP and objective weight W Entropy As the input of the model, 10 points are selected within the parameter ν∈(0,1), their corresponding AUC values are calculated, and the initial data set D is constructed:

[0043]

[0044] The objective function is constructed according to the initial AUC value and the initial data set D, including:

[0045] Objective(ν)=AUC(W comb )

[0046] Among them, W comb represents the combined weight optimization model;

[0047] Build the expected improvement model:

[0048] EI(ν)=(μ(ν)-AUC best -ξ)Φ(Z)+σ(ν)φ(Z)

[0049]

[0050] Where Φ and φ are the cumulative distribution function CDF and probability density function PDF of the standard normal distribution; ξ is the exploration parameter;

[0051] Iterative optimization:

[0052] According to the expected improvement model, calculate the EI value of the candidate point ν, and select the ν with the maximum EI value next , and calculate its corresponding AUC value based on the obtained ν next And AUC value, combined with the objective function to update the data set D and the expected improvement model, repeat this step, when the model iterates to the preset number of iterations, the output updated AUC value and data set D, and the corresponding optimal combination weight is obtained

[0053] A photovoltaic station geological disaster risk assessment system, comprising:

[0054] The subjective and objective weight acquisition module is used to construct the subjective and objective weights for geological hazard assessment of photovoltaic stations;

[0055] A weight combination module is used to construct a combined weight optimization model based on the subjective weight and the objective weight, calculate the initial AUC value of the combined weight based on the combined weight optimization model, and construct an initial data set D according to the initial AUC value;

[0056] The result acquisition module is used to construct an objective function based on the initial AUC value and the initial data set D, iteratively calculate the initial AUC value and the initial data set D according to the objective function to obtain an updated AUC value and data set D, obtain the optimal combination weight based on the updated AUC value and data set D, perform layer overlay calculation based on the optimal combination weight, obtain the calculation results, perform cluster analysis on the calculation results, and obtain the geological hazard assessment results of the photovoltaic station.

[0057] A terminal device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of any one of the methods of the present invention when executing the computer program.

[0058] A computer-readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the steps of any method described in the present invention.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] The present invention discloses a method for assessing the geological hazard risk of photovoltaic stations. Through iterative optimization calculation, dynamic adjustment of combination weights is achieved. Under the optimal combination weights, a clustering method is used to evaluate and zone the geological hazard risk of photovoltaic stations, and a geological hazard risk zoning map is obtained. During iterative optimization, the present invention improves the prediction accuracy of weights through dynamic adjustment, and the dynamic adaptability of weights is better, which can meet the needs of multi-source data fusion and dynamic weight optimization of photovoltaic stations. Compared with traditional methods, this method significantly improves the prediction accuracy and accuracy of geological hazard risk zoning, provides technical support for risk management throughout the life cycle of photovoltaic stations, and improves prediction accuracy and precision. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0062] Figure 1 is a flow chart of the method of the present invention; DETAILED DESCRIPTION

[0063] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings herein can be arranged and designed in various different configurations.

[0064] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention as claimed, but rather merely represents selected embodiments of the present invention. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without creative effort shall fall within the scope of protection of the present invention.

[0065] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings.

[0066] In the description of the embodiments of the present invention, it should be noted that if the terms "upper," "lower," "horizontal," "inner," etc. appear, the orientation or positional relationship indicated is based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship in which the inventive product is typically placed when in use. These terms are merely for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or component referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limitations on the present invention. In addition, the terms "first," "second," etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0067] In addition, if the term "horizontal" appears, it does not mean that the component must be absolutely horizontal, but can be slightly tilted. For example, "horizontal" only means that its direction is more horizontal than "vertical", and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0068] In the description of the embodiments of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "disposed," "installed," "connected," and "connected" should be understood in a broad sense. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to internal connections between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0069] The present invention is described in further detail below with reference to the accompanying drawings:

[0070] See also Figure 1, an embodiment of the present invention discloses a method for evaluating the geological hazard risk of photovoltaic stations. By constructing a geological hazard evaluation system containing 19 multi-dimensional indicators (covering factors such as lithology type, slope, aspect, and vegetation coverage), a judgment matrix that passes the consistency test is automatically generated using the Bayesian optimization algorithm. The entropy weight method is combined to extract data-driven objective weights, and finally the Bayesian optimization algorithm is used to achieve dynamic adjustment of the combined weights through parameter adaptive optimization. The K-means clustering method is used under the optimal combined weights to evaluate and partition the geological hazard risk of the photovoltaic station area, and a geological hazard risk zoning map is obtained. Compared with traditional methods, this algorithm significantly improves the prediction accuracy and accuracy of geological hazard risk zoning, and provides technical support for risk management throughout the life cycle of photovoltaic stations, including the following steps:

[0071] Step 1: Construction of geological disaster risk index system for photovoltaic stations

[0072] Specifically, in step one, the construction of a geological disaster risk index system for photovoltaic sites includes the following:

[0073] Based on the collected data such as lithology type, cover thickness, distance from faults, distance from folds, slope structure, slope gradient, slope aspect, slope height, plane curvature, profile curvature, gully density, rainfall, distance from roads, cut and fill area, photovoltaic panel density, vegetation coverage, land use type, soil erosion resistance, and surface deformation rate, the target layer (geological disaster risk), criterion layer (i.e., stratum lithology conditions, topographic conditions, geological structure conditions, meteorological and hydrological conditions, and human conditions) and indicator layer are identified, and a geological disaster risk indicator system for photovoltaic stations is established, as well as the spatial development characteristics of geological disasters in various indicators. The indicator data are clustered and all data are converted into categorical data.

[0074] Step 2: Automatic generation of the AHP judgment matrix based on the Bayesian optimization algorithm

[0075] Specifically, in step 2, the analytic hierarchy process judgment matrix based on the Bayesian optimization algorithm is automatically generated, including the following contents:

[0076] 2.1 Generation of indicator importance coefficient matrix based on expert experience

[0077] First, based on the preliminary judgment of the relative importance of the above 19 indicators in the geological hazard risk assessment of photovoltaic stations by field experts, importance coefficients are assigned to all data indicators. The value range of the importance coefficient is (0,1). The closer the importance coefficient is to 1, the more important the indicator is for the geological hazard risk assessment of photovoltaic stations. Conversely, the closer it is to 0, the less important the indicator is for the geological hazard risk assessment of photovoltaic stations.

[0078] For different indicators, the importance coefficient only serves as a basis for judging relative importance. That is, between two different indicators, the one with a higher importance coefficient indicates a higher relative importance. However, the difference in importance coefficients does not represent the specific difference in importance. If the importance coefficients of two indicators are the same, then they are equally important for the geological hazard risk assessment of photovoltaic stations.

[0079] Then, according to the importance coefficient assignment results of all the above indicators, the importance coefficient matrix X of the geological disaster risk indicators of photovoltaic stations is generated. M :

[0080] X M =[I1,I2,...,I 19 ] (0.1)

[0081] Where, I i Represents the importance coefficient of the i-th indicator, I1 to I 19 They correspond to lithology type, cover thickness, distance from fault, distance from fold, slope structure, slope gradient, slope aspect, slope height, plan curvature, profile curvature, gully density, rainfall, distance from road, cut-and-fill area, photovoltaic panel density, vegetation coverage, land use type, soil erosion resistance and surface deformation rate.

[0082] 2.2 Initial judgment matrix generation

[0083] According to the five basic rules, a 19×19 evaluation index initial judgment matrix A is randomly generated. M The five rules are:

[0084] (1) The initial judgment matrix of the evaluation index only needs to generate the upper half of the elements, and the values of the lower half of the matrix elements are based on Obtained by formula calculation.

[0085] (2) In the initial judgment matrix of the evaluation index, the diagonal elements are all 1.

[0086] (3) In the non-diagonal elements of the upper triangular matrix, if the importance coefficients of the i-th indicator and the j-th indicator are consistent, that is, I i =I j , then a (i,j) =1.

[0087] (4) In the non-diagonal elements of the upper triangular matrix, if the importance coefficient of the i-th indicator is greater than the importance coefficient of the j-th indicator, that is, I i >I j , then a number m will be randomly selected from the importance set M = {2,3,4,5,6,7,8,9}, let a(i,j) =m.

[0088] (5) In the non-diagonal elements of the upper triangular matrix, if the importance coefficient of the i-th indicator is less than the importance coefficient of the j-th indicator, that is, I i <I j , then from the importance set Randomly select a number n from (i,j) =n.

[0089] According to the above five generation rules and the index importance coefficient matrix, the randomly generated 19×19 evaluation index initial judgment matrix A M for:

[0090]

[0091] 2.3 Calculation of the initial judgment matrix CR value

[0092] Evaluation index initial judgment matrix A M The calculation formula of the characteristic root is:

[0093]

[0094] By solving the characteristic root equation, we can find the maximum eigenvalue λ from the eigenvalues λ. max .

[0095] The maximum eigenvalue λ max Substitute into the following formula to solve the consistency index CI of the initial judgment matrix of the evaluation index:

[0096]

[0097] Then, the consistency ratio CR is calculated according to the following formula:

[0098]

[0099] 2.4 Bayesian optimization algorithm to calculate the initial point data set generation

[0100] Repeat the above steps 2.2 and 2.3 100 times to obtain a total of 100 sets of evaluation index initial judgment matrix A M And consistency ratio CR. As shown in the following formula, considering that for different judgment matrices A M For example, since the values of the lower half of the matrix elements are based on The formula is calculated and the diagonal elements are all 1. Therefore, only the non-diagonal elements in the upper triangular matrix are meaningful for the construction of the matrix. Therefore, the initial judgment matrix A of 100 groups of evaluation indicators can be M Simplify and expand into a 1-dimensional matrix A' M :

[0101] A' M ={[a (1,2) ,a (1,3) ,...,a (1,19) ],[a (2,3) ,a (2,4) ,...,a (2,19) ],...,[a (17,18) ,a (17,19) ],[a (18,19) ]} (1.7)

[0102] Where A' M Contains C(19,2)=171 elements a (i,j) . Where element a (i,j) The value is based on the different indicators I in the importance coefficient matrix i and I j The relative size relationship is determined by the relationship between the i-th index and the j-th index. If the importance coefficient of the i-th index is consistent with that of the j-th index, that is, I i =I j , then a (i,j) =1; if the importance coefficient of the i-th indicator is greater than the importance coefficient of the j-th indicator, that is, I i >I j , then a number m will be randomly selected from the importance set M = {2,3,4,5,6,7,8,9}, let a (i,j) =m; if the importance coefficient of the i-th indicator is less than the importance coefficient of the j-th indicator, that is, I i <I j , then from the importance set Randomly select a number n from (i,j) =n.

[0103] The 100 sets of evaluation index initial judgment matrices and consistency ratios are combined to generate the initial point data set [X initial ,Y initial ]:

[0104]

[0105] 2.5 Automatic Generation of Bayesian Optimization Judgment Matrix

[0106] The dataset [X initial ,Y initial] as the initial calculation point of the judgment matrix of the Bayesian optimization algorithm to optimize the hierarchical analysis method, and according to the importance coefficient matrix obtained in step 2.1, the fitting objective function of the surrogate model in the Bayesian optimization algorithm is established (the objective function is y = CR + penalty coefficient C*(CR>0.1, if CR>0.1, take 1, otherwise, take 0)) and its constraints (according to C(19,2)=171, 171 pairs of elements a are generated (i,j) The constraint equations are then generated, and the acquisition function of the fusion constraint probability is collected. The next evaluation point is selected within the feasible domain through sampling. The model is then updated at the evaluation point. During the iteration process, the candidate points are always ensured to meet the boundary condition constraints until the preset number of iterations is reached, and the optimal solution within the feasible domain is finally output. The specific calculation process is as follows:

[0107] (1) Define the search space: The initial point dataset [X initial ,Y initial ] is the initial calculation point of this Bayesian optimization algorithm. Each 1-dimensional matrix A' M There are C(19,2)=171 elements a in total (i,j) , element a (i,j) The value is based on the different indicators I in the importance coefficient matrix i and I j The relative size relationship is determined by the relationship between the i-th index and the j-th index. If the importance coefficient of the i-th index is consistent with that of the j-th index, that is, I i =I j , then a (i,j) =1; if the importance coefficient of the i-th indicator is greater than the importance coefficient of the j-th indicator, that is, I i >I j , then a number m will be randomly selected from the importance set M = {2,3,4,5,6,7,8,9}, let a (i,j) =m; if the importance coefficient of the i-th indicator is less than the importance coefficient of the j-th indicator, that is, I i <I j , then from the importance set Randomly select a number n from (i,j) =n. This generates 171 elements a (i,j) Determine a set of 1-dimensional matrices A' M , by A' M Generate A M (i.e. x). Let the 1-dimensional matrix A' M All possible corresponding A M The combination of is set M, which is the value range of x in the initial point data set.

[0108] x∈M (0.7)

[0109] (2) Constructing the objective function: Based on the importance coefficient matrix obtained in step 2.1, the fitting objective function of the surrogate model in the Bayesian optimization algorithm is established:

[0110]

[0111] Where C * is the penalty coefficient. If CR>0.1, it is 1, otherwise, it is 0.

[0112] The constraint equation is:

[0113] IF=tf(1,2)*tf(1,3)*tf(1,4)*……*tf(18,19)(0.9)

[0114] Where tf(i,j) represents element a (i,j) The constraint condition in the generation process is that the importance coefficient of the i-th indicator is consistent with that of the j-th indicator, that is, I i =I j , then a (i,j) =1; if the importance coefficient of the i-th indicator is greater than the importance coefficient of the j-th indicator, that is, I i >I j , then a number m will be randomly selected from the importance set M = {2,3,4,5,6,7,8,9}, let a (i,j) =m; if the importance coefficient of the i-th indicator is less than the importance coefficient of the j-th indicator, that is, I i <I j , then from the importance set Randomly select a number n from (i,j) =n. There are 171 IFs like a (i,j) constraints.

[0115] (3) Define the acquisition function: Use the constraint EI to balance exploration and exploitation:

[0116] EI constrained (x)=EI(x)·IF (0.10)

[0117] EI(x)=(μ(x)-y best -ξ)Φ(Z)+σ(x)φ(Z) (0.11)

[0118]

[0119] Where Φ and φ are the cumulative distribution function (CDF) and probability density function (PDF) of the standard normal distribution, and ξ is the exploration parameter (default is 0.01).

[0120] (4) Iterative optimization:

[0121] ① Based on the current high-speed model, generate a judgment matrix A under the constraints of the conditional constraint equation IF M (i.e. x), calculate its EI value;

[0122] ②Then select the x point with the maximum EI value next , calculate its y value;

[0123] ③Use the obtained y value and x next Update point dataset [X initial ,Y initial ] and constrained EI models;

[0124] ④ Repeat the above steps ① to ③ until the preset number of iterations (100 times) is reached and the calculation stops;

[0125] ⑤ Output the optimal CR value and the corresponding judgment matrix A among the latest ones M (i.e. x).

[0126] Step 3: Determine subjective weights based on the fine-tuning of the AHP method based on the automatically generated judgment matrix

[0127] Specifically, in step 3, the subjective weight determination of the AHP method based on the fine-tuning of the automatically generated judgment matrix includes the following:

[0128] When the optimal CR value of the above output is greater than 0.1, according to personal experience, the judgment matrix A corresponding to the CR value at this time is M Make fine adjustments.

[0129]

[0130] For the element x that needs to be adjusted (i,j) The specific adjustment rules are:

[0131] (1) In the judgment matrix, only the upper half of the elements need to be adjusted, and the values of the lower half of the matrix elements are based on Obtained by formula calculation.

[0132] (2) In the judgment matrix, the diagonal elements are all 1.

[0133] (3) In the non-diagonal elements of the upper triangular matrix, if the importance coefficients of the i-th indicator and the j-th indicator are consistent, that is, I i =I j , then x (i,j) =1, no further adjustment is allowed.

[0134] (4) In the non-diagonal elements of the upper triangular matrix, if the importance coefficient of the i-th indicator is greater than the importance coefficient of the j-th indicator, that is, I i >I j, then a number m' different from the previous one will be randomly selected from the importance set M = {2,3,4,5,6,7,8,9}, and let x (i,j) =m'.

[0135] (5) In the non-diagonal elements of the upper triangular matrix, if the importance coefficient of the i-th indicator is less than the importance coefficient of the j-th indicator, that is, I i <I j , then from the importance set Randomly select a number n' different from the previous one, let x (i,j) =n'.

[0136] After fine-tuning, the judgment matrix A' M Calculate the CR value according to the steps in 3.3 above. If CR>0.1, continue to judge the matrix A' M Perform the above "fine-tuning-calculation-verification" process, otherwise, output the fine-tuned judgment matrix A at this time M And its corresponding subjective weight W AHP .

[0137] Step 4: Determine the objective weight using the entropy weight method

[0138] Specifically, in step 4, the objective weight of the entropy weight method is determined, which includes the following:

[0139] 4.1 Constructing the original data matrix

[0140] Based on preliminary work and field surveys, n geological disaster sample points were found. The causes of each geological disaster sample point and the relationship between them and the 19 data indicators were analyzed, and the data matrix X was constructed based on this:

[0141]

[0142] Where x (i,j) Represents the j-th data indicator value of the i-th sample data.

[0143] 4.2 Data Standardization

[0144] Standardization is carried out according to the indicator type, which is divided into positive data indicators and negative data indicators.

[0145] (1) Positive data indicators (the larger the value, the higher the risk, such as slope):

[0146]

[0147] (2) Negative data indicators (the smaller the value, the higher the risk, such as vegetation coverage):

[0148]

[0149] In the formula, max(x j ) represents the maximum value of the jth data index in all sample data; min(x j ) represents the minimum value of the j-th data indicator among all sample data.

[0150] In order to avoid ln(0) errors in subsequent calculations, it is necessary to perform a translation correction on the standardized data, that is, the overall translation ε = 10 -6 :

[0151] p' (i,j) =p (i,j) +10 -6 (0.17)

[0152] 4.3 Calculation of indicator weights

[0153] For the standardized data, calculate the proportion of the jth data indicator in the i-th geological disaster sample point:

[0154]

[0155] Where q (i,j) Need to meet

[0156] 4.4 Calculating Information Entropy

[0157] Calculate the information entropy e of the jth data indicator j :

[0158]

[0159] 4.5 Calculating the coefficient of variation

[0160] Calculate the difference coefficient d of the corresponding indicator according to the entropy value j :

[0161] d j =1-e j (0.20)

[0162] 4.6 Determining objective weights

[0163] The normalized result of the data indicator difference coefficient is the final weight of the corresponding data indicator:

[0164]

[0165] Where w j Needs to be satisfied

[0166] The final objective weight is:

[0167] W Entropy =[wj ](0.22)

[0168] Step 5: Combination weight optimization based on Bayesian optimization algorithm

[0169] Specifically, in step 5, the combination weight optimization is performed based on the Bayesian optimization algorithm, which includes the following:

[0170] 5.1 Combination Weight Optimization Model

[0171] Introducing parameter ν∈(0,1) to convert the subjective weight W AHP and objective weight W Entropy Fusion via linear combination:

[0172] W comb =ν·W AHP +(1-ν)·W Entropy (0.23)

[0173] Input subjective weight W AHP , objective weight W Entropy Then, 10 points are uniformly selected within the parameter ν∈(0,1), their corresponding AUC values are calculated, and the initial data set D is constructed:

[0174]

[0175] 5.2 AUC value calculation

[0176] The comprehensive scores of n geological disaster sample points and m non-geological disaster sample points (m≥3n) are calculated respectively. The kth sample point (containing 19 data index values y k1 ,y k2 ,…,y k19 )’s comprehensive rating is:

[0177]

[0178] Get the positive example score S pos ={S pos,1 ,S pos,2 ,S pos,3 ,…,S pos,i ,…,S pos,n-i ,S pos,n}; Negative example score S neg ={S neg,1 ,S neg,2 ,S neg,3 ,…,S neg,j ,…,S neg,m Substitute the positive and negative scores into the AUC value calculation formula (based on the Wilcoxon-Mann-Whitney statistic):

[0179]

[0180] 5.3 Bayesian optimization algorithm for model optimization steps

[0181] The dataset D obtained in step 5.1 is used as the initial calculation point for the Bayesian optimization algorithm to optimize the AHP judgment matrix. The objective function is established based on the relationship between the objective function AUC and ν in steps 5.1 and 5.2 above. The next evaluation point is selected by sampling within the parameter ν∈(0,1). The model is then updated at the evaluation point. During the iteration process, it is always ensured that the candidate point meets the boundary condition constraints until the preset number of iterations is reached, and the optimal solution within the feasible region is finally output. The specific calculation process is as follows:

[0182] (1) Define the search space: Use the initial point dataset D obtained in step 5.1 as the initial calculation point of this Bayesian optimization algorithm, and its parameter ν takes the value of (0,1).

[0183] (2) Establishing the objective function: Generate the functional relationship between the objective function AUC and ν according to formulas 1.25, 1.27, and 1.28:

[0184] Objective(ν)=AUC(W comb )(0.27)

[0185] (3) Using Expected Improvement (EI) to balance exploration and exploitation:

[0186] EI(ν)=(μ(ν)-AUC best -ξ)Φ(Z)+σ(ν)φ(Z)(0.28)

[0187]

[0188] Where Φ and φ are the cumulative distribution function (CDF) and probability density function (PDF) of the standard normal distribution, and ξ is the exploration parameter (default is 0.01).

[0189] (4) Iterative optimization:

[0190] ① Calculate the EI value of the candidate point ν based on the current high-speed model;

[0191] ②Then select the ν at the point with the maximum EI value next , calculate its AUC value;

[0192] ③Use the obtained AUC value and ν next Update the dataset D and the expected improved model;

[0193] ④ Repeat the above steps ① to ③ until the preset number of iterations (100 times) is reached and the calculation stops;

[0194] ⑤Output the optimal AUC value at this time and the corresponding optimal combination weight

[0195] Step 6: PV site geological disaster risk assessment and zoning based on optimal combined weights

[0196] Specifically, in step 6, the PV site geological disaster risk assessment and zoning based on the optimal combined weights include the following:

[0197] 6.1 Calculation of risk score

[0198] The optimal combination weight corresponding to the optimal AUC value obtained above for:

[0199]

[0200] For a certain evaluation unit i in the photovoltaic station area, calculate its geological disaster risk score S i :

[0201]

[0202] Where, Represents the combined weight of the j-th data indicator of evaluation unit i.

[0203] The hazard score S of all evaluation units in the photovoltaic station area is i Mapped to the interval [0,1] for uniform grading:

[0204]

[0205] In the formula, max(S) and min(S) refer to the minimum and maximum scores of all evaluation units.

[0206] 6.2 Danger Partitioning (K-means Clustering)

[0207] Set the number of clusters K=4.

[0208] Initialize cluster centers (K-means++ optimization): Randomly select the first center point μ1, then perform cluster optimization, select the center point farthest from the existing center point as the second center point; perform cluster optimization again, select the center point farthest from the existing center point as the third center point; perform cluster optimization again, select the center point farthest from the existing center point as the fourth center point. At this point, the number of centers equals the number of clusters K, and the cluster optimization ends, outputting the four center point values.

[0209] Iterative optimization: assign each region i to the nearest cluster center μ k :

[0210] Cluster(i)=argmin|S i -μ k | (1.36)

[0211] Then recalculate the center of each cluster:

[0212]

[0213] The newly generated cluster center μ k Substitute Equation 1.36 into the equation and repeat steps 1.36 and 1.37 until the preset number of iterations (100) is reached. Output the final three cluster center values.

[0214] Arrange the final four cluster center values in ascending order. μ1 corresponds to the minimum center value and is set as the low-risk area; μ2 corresponds to the center value and is set as the medium-risk area; μ3 corresponds to the center value and is set as the high-risk area; and μ4 corresponds to the maximum center value and is set as the extremely high-risk area.

[0215] 6.3 Partitioning Results Visualization

[0216] Based on the final four cluster center values, the PV station area is divided into four zones: low-risk, medium-risk, high-risk, and extremely high-risk. Each of the four zones is assigned a color: green for low-risk, yellow for medium-risk, orange for high-risk, and red for extremely high-risk. The corresponding areas are then colored according to the specified colors, and a legend is automatically generated. The resulting visualization of the geological hazard zones (i.e., the geological hazard zone zoning map) is then output.

[0217] A photovoltaic station geological disaster risk assessment system, comprising:

[0218] The subjective and objective weight acquisition module is used to construct the subjective and objective weights for geological hazard assessment of photovoltaic stations;

[0219] A weight combination module is used to construct a combined weight optimization model based on the subjective weight and the objective weight, calculate the initial AUC value of the combined weight based on the combined weight optimization model, and construct an initial data set D according to the initial AUC value;

[0220] The result acquisition module is used to construct an objective function based on the initial AUC value and the initial data set D, iteratively calculate the initial AUC value and the initial data set D according to the objective function to obtain an updated AUC value and data set D, obtain the optimal combination weight based on the updated AUC value and data set D, perform layer overlay calculation based on the optimal combination weight, obtain the calculation results, perform cluster analysis on the calculation results, and obtain the geological hazard assessment results of the photovoltaic station.

[0221] A schematic diagram of a terminal device provided in one embodiment of the present invention. The terminal device in this embodiment includes: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of each of the aforementioned method embodiments are implemented. Alternatively, when the processor executes the computer program, the functions of each module / unit in each of the aforementioned device embodiments are implemented.

[0222] The computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to accomplish the present invention.

[0223] The terminal device may be a computing device such as a desktop computer, a notebook computer, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0224] The processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0225] The memory may be used to store the computer programs and / or modules, and the processor implements various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory.

[0226] If the module / unit integrated in the terminal device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the process in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, it can implement the steps of each of the above-mentioned method embodiments. The computer program includes computer program code, which can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0227] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for assessing geological hazards in photovoltaic stations, characterized in that: The following steps are involved: Construct subjective and objective weights for geological hazard risk assessment of photovoltaic stations; A combined weight optimization model is constructed based on the subjective weight and the objective weight. The initial AUC value of the combined weight is calculated based on the combined weight optimization model. The initial data set D is constructed according to the initial AUC value. An objective function is constructed based on the initial AUC value and the initial data set D. The initial AUC value and the initial data set D are iteratively calculated according to the objective function to obtain an updated AUC value and data set D. The optimal combination weight is obtained based on the updated AUC value and data set D. Layer overlay calculation is performed based on the optimal combination weight to obtain the calculation results. The calculation results are clustered and analyzed to obtain the geological hazard assessment results of the photovoltaic station.

2. A method for assessing geological hazards of photovoltaic stations according to claim 1, characterized in that: The subjective weights for the geological hazard assessment of photovoltaic stations are constructed, including: Establish a geological disaster risk indicator system for photovoltaic sites; Based on the geological hazard risk index system of photovoltaic stations, the importance coefficient matrix X of geological hazard risk index of photovoltaic stations is generated. M ; Based on the importance coefficient matrix X of geological disaster risk indicators of photovoltaic stations M Generate the initial judgment matrix A M , calculate the initial judgment matrix A M Repeat this step to iteratively generate multiple groups of evaluation index initial judgment matrix AM and consistency ratio CR. M and the consistency ratio CR to generate the initial point data set [X initial ,Y initial ]; Through iterative optimization, update the initial point data set [X initial ,Y initial ], get the updated judgment matrix A M and consistency index ratio CR; Determine whether the consistency index ratio CR meets the constraint conditions. If so, output the judgment matrix A corresponding to the current consistency index ratio CR M If not, the judgment matrix A m Adjust the elements in until the consistency index ratio CR satisfies the constraint conditions, and output the current corresponding judgment matrix A M ; According to the final output judgment matrix A M Get its corresponding subjective weight.

3. A method for assessing geological hazards of photovoltaic stations according to claim 2, characterized in that: By iterative optimization, the initial point data set [X initial ,Y initial ], get the updated judgment matrix A m And consistency index ratio CR, including: Construct the fitting objective function: Where C * is the penalty coefficient; Construct constraints for the function; IF=tf(1,2)*tf(1,3)*tf(1,4)*……*tf(18,19) Where tf(i,j) represents element a (i,j) The constraint condition in the generation process indicates that the importance coefficient of the i-th indicator is consistent with that of the j-th indicator, that is, I i =I j ; Define the acquisition function: NO constrained (x)=NO(x) IF EI(x)=(μ(x)-y best -ξ)Φ(Z)+σ(x)φ(Z) Where Φ and φ are the cumulative distribution function CDF and probability density function PDF of the standard normal distribution, and ξ is the exploration parameter; Iterative optimization: The initial point data set [X initial ,Y initial ] as the input of the fitting objective function to generate the judgment matrix A M , combined with the acquisition function, calculate the current judgment matrix A M EI value; select the x point with the maximum EI value next , combined with the fitting objective function, calculate the y value corresponding to the maximum point; based on the obtained x next and y values, update the point dataset [X initial ,Y initial ] and EI model, repeat this step until the preset number of iterations is reached, and the latest point data set [X initial ,Y initial ], and based on the latest point dataset [X initial ,Y initial ] Get the corresponding consistency index ratio CR and judgment matrix A M .

4. A method for assessing geological hazards of photovoltaic stations according to claim 3, characterized in that: The judgment consistency index ratio CR satisfies the constraint condition. If so, the judgment matrix A corresponding to the current consistency index ratio CR is output. M If not, the judgment matrix A m Adjust the elements in until the consistency index ratio CR satisfies the constraint conditions, and output the current corresponding judgment matrix A M ,include: Set the consistency index ratio CR constraint to 0.

1. When the output consistency index ratio CR is greater than 0.1, the element x in the judgment matrix (i,j) Make adjustments; Based on the adjusted judgment matrix, recalculate its corresponding consistency index ratio CR until the consistency index ratio CR is less than or equal to 0.1, and output the current corresponding judgment matrix A M .

5. A method for assessing geological hazards of photovoltaic stations according to claim 1, characterized in that: The objective weights for the geological hazard assessment of photovoltaic stations are constructed, including: Establish a geological disaster risk indicator system for photovoltaic sites; Based on the relationship between the indicators in the geological hazard index system of photovoltaic stations, the original data matrix is constructed; Based on the original data moment, the weight of each indicator is calculated, and the objective weight is obtained according to the weight of each indicator.

6. A method for assessing geological hazards of photovoltaic stations according to claim 1, characterized in that: The construction of a combined weight optimization model based on subjective weights and objective weights includes: Introducing parameter ν∈(0,1) to convert the subjective weight W AHP and objective weight W Entropy Fusion via linear combination: W comb =ν·W AHP +(1-ν)·W Entropy The subjective weight W AHP and objective weight W Entropy As the input of the model, 10 points are selected within the parameter ν∈(0,1), their corresponding AUC values are calculated, and the initial data set D is constructed:

7. A method for assessing geological hazards of photovoltaic stations according to claim 6, characterized in that: The objective function is constructed according to the initial AUC value and the initial data set D, including: Objective(ν)=AUC(W comb ) Among them, W comb represents the combined weight optimization model; Build the expected improvement model: EI(ν)=(μ(ν)-AUC best -ξ)Φ(Z)+σ(ν)φ(Z) Where Φ and φ are the cumulative distribution function CDF and probability density function PDF of the standard normal distribution; ξ is the exploration parameter; Iterative optimization: According to the expected improvement model, calculate the EI value of the candidate point ν, and select the ν with the maximum EI value next , and calculate its corresponding AUC value based on the obtained ν next And AUC value, combined with the objective function to update the data set D and the expected improvement model, repeat this step, when the model iterates to the preset number of iterations, the output updated AUC value and data set D, and the corresponding optimal combination weight is obtained 8. A photovoltaic station geological disaster risk assessment system, characterized by: include: The subjective and objective weight acquisition module is used to construct the subjective and objective weights for geological hazard assessment of photovoltaic stations; A weight combination module is used to construct a combined weight optimization model based on the subjective weight and the objective weight, calculate the initial AUC value of the combined weight based on the combined weight optimization model, and construct an initial data set D according to the initial AUC value; The result acquisition module is used to construct an objective function based on the initial AUC value and the initial data set D, iteratively calculate the initial AUC value and the initial data set D according to the objective function to obtain an updated AUC value and data set D, obtain the optimal combination weight based on the updated AUC value and data set D, perform layer overlay calculation based on the optimal combination weight, obtain the calculation results, perform cluster analysis on the calculation results, and obtain the geological hazard assessment results of the photovoltaic station.

9. A terminal device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images