Sound barrier collaborative design method and system based on the not-in-my-backyard effect

Abstract

The application provides a sound barrier collaborative design method and system based on adjacent avoidance effect, which comprises obtaining design parameters of each alternative scheme and its acoustic, safety and adjacent avoidance effect indexes to form an initial decision matrix; the adjacent avoidance effect is quantified by establishing a sunshine blocking and visual field occupation model and combining expert scoring for nonlinear weighting; the combination weight is determined by fusing the analytic hierarchy process and the entropy weight method to construct a weighted and normalized decision matrix; for the particularity of the adjacent avoidance effect, the minimum value of the probability density of the evaluation value of all schemes is taken as the positive ideal solution component, and the maximum value is taken as the negative ideal solution component, the relative closeness is calculated by combining the Mahalanobis distance for preliminary screening to obtain an optimized scheme set; then, the derived schemes are generated by Monte Carlo sampling based on the joint distribution of the kernel density estimation of the design parameters of the set, the derived schemes are combined with the initial schemes, and the decision sorting is re-executed to output a collaborative optimization design scheme sequence.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of collaborative design, and particularly relates to a sound barrier collaborative design method and system based on NIMBY effect. BACKGROUND

[0002] Sound barriers can alleviate the impact of traffic noise on residents along the line, and in the design, balance needs to be sought among multiple or even conflicting targets. In the design, focus is usually placed on core technical indicators such as acoustic performance and structural safety. In order to select the best from multiple alternative schemes, a multi-attribute decision-making method MCDM can be used for evaluation, for example, analytic hierarchy process (AHP) or technique for order preference by similarity to ideal solution (TOPSIS) is used to score and rank the preset scheme set. However, the negative impact of sound barriers as large structures on the surrounding environment and the psychology of residents, i.e. NIMBY effect, such as visual compression and sunlight blocking, is often ignored. If social acceptance is ignored, a technically feasible scheme may be difficult to implement due to opposition from residents, affecting the social benefits and implementation efficiency of the project. For complex indicators such as NIMBY effect that contain subjective feelings, there is a lack of a systematic and objective quantitative evaluation model, and existing processing methods mostly remain at qualitative description or single simplified calculation, and cannot accurately obtain comprehensive impact. In the allocation of decision weights, subjective weighting methods such as AHP that simply rely on expert scoring or objective weighting methods such as entropy weight method that are completely based on data distribution cannot truly reflect the comprehensive importance of each indicator. The definition of positive and negative ideal solutions in the standard TOPSIS method is not completely applicable to non-monotonic indicators such as NIMBY effect that may have a moderate rather than extreme optimal interval; and the Euclidean distance used does not take into account the correlation between indicators, which may lead to distorted ranking results. The existing decision-making process can only perform one-time screening on a limited number of initial alternative schemes, and lacks a mechanism for re-optimization of the preferred scheme area, limiting the possibility of finding a globally better solution. SUMMARY

[0003] To solve the above problems, in a first aspect, the application provides a sound barrier collaborative design method based on NIMBY effect, comprising the following steps:

[0004] A plurality of sound barrier alternative schemes defined by respective design parameter vectors are obtained, and an initial decision matrix including evaluation indicators of acoustic performance, structural safety and NIMBY effect is constructed; for each alternative scheme, a three-dimensional ray tracing model of sunlight blocking degree and a spherical projection model of visual field invasion rate are established, and the model output values are weighted with landscape coordination expert scores to obtain the NIMBY effect evaluation value of each alternative scheme;

[0005] A combination weighting method is used to fuse the subjective weight determined by the analytic hierarchy process and the objective weight determined by the entropy weight method to obtain the combined weight of each evaluation indicator; the initial decision matrix is normalized, and the combined weight is used to construct a weighted and normalized decision matrix;

[0006] determining positive ideal solution and negative ideal solution; calculating the distance between each alternative and the positive and negative ideal solution, and calculating the relative closeness according to the distance to preliminarily sort and screen out the preferred solution set;

[0007] extracting the design parameter vector of each solution in the preferred solution set, constructing a multi-dimensional joint probability distribution model based on the kernel density estimation method, generating a group of derivative alternative solutions through Monte Carlo sampling; calculating the evaluation index values of the group of derivative alternative solutions, merging the evaluation index values with the initial alternative solutions, and performing the steps of constructing a weighted decision matrix and sorting again on the merged solution set to obtain the collaborative design solution sorting.

[0008] In a second aspect, the application provides a sound barrier collaborative design system based on neighborhood avoidance effect, comprising the following modules:

[0009] a matrix construction module, which acquires a plurality of sound barrier alternative solutions defined by respective design parameter vectors, constructs an initial decision matrix including evaluation indexes of acoustic performance, structural safety and neighborhood avoidance effect, and obtains the neighborhood avoidance effect evaluation value of each alternative solution by weighting the model output value and the landscape coordination expert score through the establishment of a three-dimensional ray tracing model of sunshine blocking degree and a spherical projection model of visual intrusion rate;

[0010] a weight fusion module, which fuses the subjective weight determined by the analytic hierarchy process and the objective weight determined by the entropy weight method to obtain the combined weight of each evaluation index; normalizes the initial decision matrix and constructs a weighted and normalized decision matrix using the combined weight;

[0011] a solution set screening module, which determines positive ideal solution and negative ideal solution; calculates the distance between each alternative and the positive and negative ideal solution, and calculates the relative closeness according to the distance to preliminarily sort and screen out the preferred solution set;

[0012] a collaborative design solution sorting module, which extracts the design parameter vector of each solution in the preferred solution set, constructs a multi-dimensional joint probability distribution model based on the kernel density estimation method, generates a group of derivative alternative solutions through Monte Carlo sampling; calculates the evaluation index values of the group of derivative alternative solutions, merges the evaluation index values with the initial alternative solutions, and performs the steps of constructing a weighted decision matrix and sorting again on the merged solution set to obtain the collaborative design solution sorting.

[0013] The application brings the key social and psychological factor of the noise barrier's NIMBY effect into the collaborative design framework by constructing a systematic quantitative model, so that the evaluation of the design scheme is more comprehensive, and the social acceptance of the project is improved; the combination weighting method combining subjective and objective methods is adopted to ensure the balance and rationality of the evaluation index weight; for the non-monotonic index such as NIMBY effect, the minimum value of its probability density function is taken as the positive ideal solution, which overcomes the limitations of the traditional TOPSIS method in dealing with moderate optimal target, and through the kernel density estimation and Monte Carlo sampling of the optimal scheme set, potential better designs outside the initial scheme can be generated and explored, realizing the improvement from selecting the best from limited schemes to optimizing the excellent design space, and increasing the probability of obtaining the globally optimal collaborative design scheme. BRIEF DESCRIPTION OF DRAWINGS

[0014] Figure 1 is an index diagram of alternative schemes;

[0015] Figure 2 is a diagram of the weighted fusion of NIMBY effect evaluation values;

[0016] Figure 3 is a diagram of index combination weighting;

[0017] Figure 4 is a diagram of the distance between the scheme and the ideal solution;

[0018] Figure 5 is a diagram of generating derivative schemes based on kernel function estimation;

[0019] Figure 6 is a diagram of ranking derivative schemes and initial schemes. DETAILED DESCRIPTION

[0020] To make the purpose, technical scheme and advantages of the present application clearer, the technical scheme of the present application will be described clearly and completely below in combination with specific embodiments of the present application and corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0021] The terms used in the embodiment part of the present application are only used to explain the specific embodiments of the present application, and are not intended to limit the present application.

[0022] Specific embodiment one, the embodiment proposes a noise barrier collaborative design method based on NIMBY effect, as shown in Figure 1 The method comprises the following steps:

[0023] S1, obtain a plurality of sound barrier alternatives defined by respective design parameter vectors, and construct an initial decision matrix including evaluation indexes of acoustic performance, structural safety, and neighborhood avoidance effect; for each alternative, a three-dimensional ray tracing model of sunshine blocking degree and a spherical projection model of visual field invasion rate are established, and the model output values are weighted with expert scores of landscape coordination to obtain an evaluation value of neighborhood avoidance effect of each alternative;

[0024] A design parameter vector of the sound barrier is defined, for example, including variables such as barrier height, length, material type, transparency ratio, etc. An array of alternative solutions is generated within the value range of each parameter, for example, 50, by using a design of experiment method such as Latin hypercube sampling. For each solution, its evaluation index value is obtained by professional software simulation or calculation: the sound insertion loss is simulated and calculated by using SoundPLAN or Cadna / A software as the acoustic performance index; the stress and displacement under wind load are analyzed by using finite element software such as ANSYS or ABAQUS as the structural safety index; the evaluation value of neighborhood avoidance effect is calculated by subsequent steps; each alternative solution is taken as a row, and each evaluation index is taken as a column, to construct an initial decision matrix, for example, 50 rows and 3 columns, as shown in Figure 1 .

[0025] The three-dimensional ray tracing model can accurately calculate how long and how large an area of shadow the sound barrier will cause to the surrounding buildings in a day; the spherical projection model simulates the real feeling of the human eye and calculates how large a proportion of the residents' visual field this huge object will occupy. Through the three-dimensional ray tracing model and the spherical projection model, vague complaints are converted into specific data. In an embodiment, the three-dimensional ray tracing model outputs the sunshine blocking degree, more specifically, for the sunshine blocking degree, a three-dimensional geometric model of the sound barrier and the surrounding buildings is established by using Rhino and Grasshopper software, combined with the solar path algorithm, the light path from the sun to the sensitive points such as the windows of the residential buildings is traced on the key dates such as the winter solstice, and the percentage of sunshine time loss caused by the barrier blocking is calculated. The spherical projection model outputs the visual field invasion rate, more specifically, for the visual field invasion rate, a virtual wide-angle camera is set at a typical observation point, the barrier entity is projected onto a unit sphere, and the ratio of the projection area to the effective visual field sphere area of the observer is calculated. For landscape coordination, a plurality of experts in the fields of architecture and environment are invited to score based on the fusion degree of the color, material, and form of the barrier with the surrounding environment on a scale of 1 to 10; after normalization of the three index values, a nonlinear power function weighting formula such as is used to fuse the three into a single neighborhood avoidance effect evaluation value, as shown in Figure 2 .

[0026] S2, the subjective weight determined by the analytic hierarchy process and the objective weight determined by the entropy weight are fused by the combination weighting method to obtain the combination weight of each evaluation index; the initial decision matrix is normalized, and the combination weight is used to construct a weighted normalized decision matrix;

[0027] The analytic hierarchy process is adopted, decision experts are invited to compare three indexes of acoustic performance, structural safety and neighborhood avoidance effect two by two, a judgment matrix is constructed, a subjective weight vector is obtained by calculating the maximum eigenvalue and the corresponding eigenvector; based on the data in the initial decision matrix, the information entropy of each index is calculated by using the entropy weight method, the smaller the information entropy, the more information and the greater the difference the index provides, the higher the objective weight, and an objective weight vector is obtained; the two weights are fused by a linear weighting formula, for example, the combination weight is equal to 0.5 times the subjective weight plus 0.5 times the objective weight, to obtain the combination weight, as shown in Figure 3 .

[0028] The data in each column of the initial decision matrix is processed, for the benefit type index such as acoustic performance, the formula of value minus minimum value divided by range is used for normalization; for the indexes such as structural safety and neighborhood avoidance effect, the formula of maximum value minus value divided by range is used for normalization, so that all index values are converted to between 0 and 1; each column element in the normalized matrix is multiplied by the combination weight value corresponding to the column index to obtain a weighted normalized decision matrix.

[0029] S3, determine the positive ideal solution and the negative ideal solution; calculate the distance between each alternative scheme and the positive and negative ideal solutions, and preliminarily sort according to the relative closeness degree calculated by the distance to screen out the preferred scheme set;

[0030] The positive ideal solution is a virtual best scheme composed of the optimal values of each index in all schemes, and the negative ideal solution is a virtual worst scheme composed of the worst values; for monotonic indexes such as acoustic performance, the positive ideal solution component is the normalized maximum value 1, and the negative ideal solution component is the minimum value 0; for the neighborhood avoidance effect index, the neighborhood avoidance effect evaluation values of all alternative schemes are collected, in an optional embodiment, the kernel density estimation algorithm is used to fit the probability density distribution curve, and the evaluation value corresponding to the minimum value of the probability density distribution curve is regarded as the most acceptable ideal state of non-extreme, as the positive ideal solution component thereof; the negative ideal solution component takes the maximum value in all evaluation values, representing the worst case.

[0031] A covariance matrix of the weighted normalized decision matrix is calculated, which represents the correlation between the evaluation indexes; for each alternative, the distance to the positive ideal solution vector and the negative ideal solution vector is calculated using, for example, the Euclidean distance or Mahalanobis distance formula; the score of each scheme is calculated by applying the standard formula of the TOPSIS method, i.e. the relative closeness is equal to the distance to the negative ideal solution divided by the sum of the distance to the positive ideal solution and the distance to the negative ideal solution; all alternative schemes are ranked in descending order of the relative closeness score, and the top-ranked schemes, such as the top ten percent of the schemes, are selected to form the preferred scheme set.

[0032] S4, the design parameter vectors of each scheme in the preferred scheme set are extracted, a multi-dimensional joint probability distribution model is constructed based on the kernel density estimation method, and a set of derived alternative schemes is generated by Monte Carlo sampling; the values of each evaluation index of the set of derived alternative schemes are calculated, the values of each evaluation index are combined with the initial alternative schemes, and the steps of constructing a weighted decision matrix and ranking are performed again on the combined scheme set to obtain the ranking of the collaborative design schemes.

[0033] The design parameter vectors corresponding to each scheme in the preferred scheme set are extracted, such as the combination of original design variables such as barrier height and length; these excellent parameter vectors are input into the multi-dimensional kernel density estimation algorithm as sample data to establish a probability model that can describe the joint distribution characteristics of excellent design parameters; based on this probability model, a large number of random samples are generated using the Monte Carlo method, for example, 500 new design parameter vectors, which are similar to the preferred schemes in design but have certain random perturbations, constitute the derived alternative schemes.

[0034] The calculation process in the first and second steps is repeated for the 500 derived alternative schemes to obtain their respective acoustic performance, structural safety, and NIM index values; the 500 derived schemes and their index values are combined with the initial 50 alternative schemes to form an expanded scheme set containing 550 schemes; the processes of the fourth to sixth steps are completely repeated for this expanded scheme set, i.e. re-normalization, weighting, determination of new positive and negative ideal solutions, calculation of Mahalanobis distance and relative closeness, to obtain the final ranking result.

[0035] In an optional embodiment, the weighting of the model output value and the landscape coordination expert score to obtain the NIM evaluation value of each alternative scheme includes:

[0036] For any alternative scheme, a three-dimensional geometric model containing the sound barrier, surrounding buildings, and solar trajectory is established, and the average solar shading degree and the average visual intrusion rate of the three-dimensional geometric model at a plurality of preset observation points and time periods are calculated;

[0037] The landscape coordination expert score is converted into a penalty term;

[0038] The average sunshine obstruction degree, the average visual field invasion rate and the landscape coordination penalty term are combined by a preset weighting function to obtain the NIM evaluation value of the alternative scheme.

[0039] Exemplarily, a sound barrier scheme to be evaluated is designed to be 4.5 meters high and 300 meters long, and a digital twin model is constructed in a three-dimensional modeling software such as Rhinoceros. A resident building of 12 floors around the sound barrier is also accurately modeled. The software automatically generates the annual sun trajectory according to the local geographic latitude, for example, 39 degrees north latitude. Selecting the three months of the most critical lighting in winter, the simulation period is from 9 am to 3 pm every day, and 15 observation points are set at the window positions of the 3rd, 6th and 9th floors on the side of the resident building facing the sound barrier. Through ray tracing simulation, it is calculated that the average sunshine obstruction degree caused by the sound barrier scheme is 18%.

[0040] In calculating the average visual field invasion rate, the above-mentioned 15 observation points are still used. From each observation point, a standard observation cone with a horizontal angle of 120 degrees and a vertical angle of 90 degrees is defined to simulate the normal visual field range of the human eye. The proportion of the solid angle occupied by the sound barrier in each cone to the total solid angle of the cone is calculated, and the results of the 15 points are averaged to obtain the average visual field invasion rate of the sound barrier scheme, which is 0.22. Invite three landscape design experts to score the fusion degree of the color, texture and surrounding environment of the sound barrier, with a score range of 1 to 10 points, and the average score of the three experts is 7 points. According to the preset formula, the landscape coordination penalty term is calculated as 0.3. Through the weighting function, for example, the NIM evaluation value of the sound barrier scheme is 0.216, which is equal to 0.5 times the average sunshine obstruction degree plus 0.3 times the average visual field invasion rate plus 0.2 times the landscape coordination penalty term.

[0041] In an alternative embodiment, the design parameter vector comprises:

[0042] The height, length, material type, color scheme and transverse position relative to the road of the sound barrier.

[0043] In order to facilitate the optimization calculation of the algorithm, the complete design scheme of a sound barrier is abstracted into a numerical vector. Each dimension of the numerical vector represents a key design parameter. For example, a five-dimensional design parameter vector can completely describe a scheme. The first dimension is the height, with a value range of 3.0 meters to 6.0 meters; the second dimension is the length, with a value range of 200 meters to 800 meters; the third dimension is the transverse distance relative to the road curb, with a value range of 0.5 meters to 1.5 meters.

[0044] Encoding is also applied to non-numeric parameters. For example, the fourth dimension is material type, with number 1 representing polycarbonate plate, number 2 representing acrylic plate, and number 3 representing metal microporous plate. The fifth dimension is color scheme, with number 1 representing gray system, number 2 representing green system, and number 3 representing blue system. A specific design parameter vector, such as 4.5, 500, 1.0, 2, 1, represents a design scheme, i.e., the height of the sound barrier is 4.5 meters, the length is 500 meters, the distance from the curb is 1.0 meter, the acrylic plate material is used, and the gray system is painted. The vector representation method enables the computer to efficiently process and iterate thousands of different design schemes.

[0045] In an optional embodiment, the combination weighting method is used to fuse the subjective weight determined by the analytic hierarchy process and the objective weight determined by the entropy weight method to obtain the combined weight of each evaluation index, comprising:

[0046] A pairwise comparison judgment matrix of each evaluation index is constructed, and after consistency check, the subjective weight vector is calculated ; based on the initial decision matrix, the information entropy of each index is calculated, and the objective weight vector is calculated according to the information entropy ; the combined weight is calculated by using the linear combination formula , wherein α is a preset subjective weight preference coefficient with a value between 0 and 1.

[0047] Experts in the field are invited to compare the importance of the three evaluation indexes of acoustic performance, structural safety and neighborhood avoidance effect pairwise, and a judgment matrix is constructed. For example, the expert considers that the acoustic performance is more important than the structural safety, and gives an evaluation value of 5. After all comparisons are completed, the maximum eigenvalue and eigenvector of the matrix are calculated, and consistency check is performed to ensure that the logic of the judgment is not self-contradictory. The normalized eigenvector obtained after the check passes is the subjective weight, for example, acoustic performance 0.5, structural safety 0.2, and neighborhood avoidance effect 0.3.

[0048] Based on the decision matrix composed of the specific performance data of 50 initial alternative schemes on four indexes, the objective weight is calculated. If the evaluation values of the structural safety index are concentrated in a very small range, it indicates that the economic cost index has low discrimination, the information entropy is large, and a lower weight should be given. After calculation by the information entropy method, a set of objective weights may be obtained, for example, acoustic performance 0.6, structural safety 0.1, and neighborhood avoidance effect 0.3. Set a subjective weight preference coefficient α as 0.7, which represents more emphasis on the expert's experience judgment, and calculate the final combined weight by the linear combination formula, which not only combines the expert's prior knowledge, but also considers the distribution characteristics of the data itself.

[0049] In an optional embodiment, the determination of the positive ideal solution and the negative ideal solution comprises:

[0050] For the acoustic performance index, the maximum value of the acoustic performance index is taken as the positive ideal solution component, and the minimum value is taken as the negative ideal solution component.

[0051] For the structural safety index, the minimum value of the structural safety index is taken as the positive ideal solution component, and the maximum value is taken as the negative ideal solution component.

[0052] For the NIMBY index, the minimum value of the evaluation values of all the alternative schemes is taken as the positive ideal solution component, and the maximum value of the evaluation values is taken as the negative ideal solution component.

[0053] Suppose there are 100 alternative schemes. For the acoustic performance index, which is a benefit type index, the larger the value is, the better. The maximum value of the noise reduction effect in the 100 schemes is 25 db, and the minimum value is 15 db. Therefore, the acoustic performance component of the positive ideal solution is 25, and the negative ideal solution is 15. For the structural safety index, the smaller the value is, the better.

[0054] The NIMBY index value is also the smaller the better, but directly taking the minimum value may result in an ideal solution that is an extreme case difficult to achieve. Suppose the NIMBY evaluation values of the 100 schemes are distributed between 0.1 and 0.8, but the evaluation values of most excellent schemes are concentrated around 0.2. In an embodiment, for the NIMBY index, the minimum value of the evaluation values of all the alternative schemes is taken as the positive ideal solution component, for example, the minimum value is 0.21. Therefore, 0.21 is taken as the positive ideal solution component. In an alternative embodiment, kernel density estimation is performed on the 100 evaluation values to obtain a probability density function curve, and the minimum value appears at the position of 0.21. This minimum value represents the most concentrated area of excellent schemes, and is an ideal state that is more representative and achievable. Therefore, 0.21 is taken as the positive ideal solution component of the NIMBY index, and the maximum value 0.8 of all the evaluation values is taken as the negative ideal solution component. The positive ideal solution thus determined combines the theoretical optimum and the practical feasibility.

[0055] In an optional embodiment, the calculation of the distance between each alternative scheme and the positive and negative ideal solutions comprises:

[0056] The covariance matrix S between the evaluation indexes is calculated based on the weighted normalized decision matrix, and the Mahalanobis distance between each alternative scheme and the positive and negative ideal solutions is calculated according to the formula

[0057] ​In the multi-dimensional evaluation space, the evaluation indexes are not independent of each other. For example, increasing the height of a sound barrier usually improves the acoustic performance, but also increases the NIMBY effect, which shows that there is a correlation between the indexes. The traditional Euclidean distance cannot describe this correlation. Therefore, based on the weighted and normalized performance data of all candidate solutions, a covariance matrix S is calculated. The elements on the diagonal of the covariance matrix are the variances of each index itself, and the non-diagonal elements are the covariances between different index pairs, which represent the linear relationship between them.

[0058] When calculating the distance between a candidate solution x and the ideal solution y, the Mahalanobis distance formula is preferably used. The inverse matrix S of the covariance matrix in the formula -1 is a kind of transformation of the data space, which eliminates the correlation between the indexes and normalizes the variances of the indexes. It is equivalent to calculating the distance in a new standardized coordinate system. Even if a solution has a large difference with the ideal solution in an index with large variance, as long as this difference is within the trend of coordinated change with other indexes, the Mahalanobis distance will not be greatly enlarged. The calculated distance can more truly reflect the comprehensive gap of a candidate solution after considering the inherent correlation between all indexes, as shown in Figure 4 .

[0059] In an optional embodiment, the kernel density estimation method constructs a multi-dimensional joint probability distribution model, and generates a set of derived candidate solutions through Monte Carlo sampling, including:

[0060] The design parameter vectors of all solutions in the preferred solution set are extracted to form a multi-dimensional sample set. The bandwidth of the kernel function is determined by the bandwidth selection method, and the kernel density estimation of the multi-dimensional sample set is performed to construct the multi-dimensional joint probability density function of the design parameter vector. Based on the constructed joint probability density function, a new set of design parameter vectors is generated through Monte Carlo random sampling as a set of derived candidate solutions.

[0061] After one round of optimization, the top 50 preferred solutions are selected. The five-dimensional design parameter vectors of the 50 solutions, i.e. height, length, material, color and transverse position, are extracted to form a 50x5 sample matrix. In order to construct a model that can describe the distribution of these excellent solutions in the design space, kernel density estimation is used. The optimal bandwidth is automatically calculated by methods such as cross-validation to ensure the fitting accuracy of the model. Thus, a five-dimensional joint probability density function is obtained, which has a higher function value in the region of the excellent solution set and a lower function value in the region with poor performance, as shown in Figure 5 .

[0062] The constructed probability density function indicates the regions in the design parameter space where superior solutions are more likely to occur. To explore these high-potential regions more deeply, the Monte Carlo method is used to randomly sample from this probability density function. For example, suppose 2000 new derivative solutions are generated. The sampling process will favor samples from regions with high probability density, and the generated new design parameter vectors, such as height between 4.2 and 4.8 meters and length between 450 and 550 meters, are more likely to combine into a high-performance solution than completely randomly generated parameters. These 2000 derived new solutions will serve as input for the next round of optimization evaluation, enabling the search of the design space, as shown in Figure 6

[0063] In a second embodiment, a sound barrier collaborative design system based on the neighborhood avoidance effect is proposed, which includes the following modules:

[0064] A matrix construction module obtains a plurality of sound barrier candidate solutions defined by respective design parameter vectors, constructs an initial decision matrix including evaluation indexes of acoustic performance, structural safety, and neighborhood avoidance effect; for each candidate solution, a three-dimensional ray tracing model of sun-shading degree and a spherical projection model of visual intrusion rate are established, and the model output values are weighted with landscape coordination expert scores to obtain the neighborhood avoidance effect evaluation value of each candidate solution;

[0065] A weight fusion module fuses the subjective weight determined by the analytic hierarchy process and the objective weight determined by the entropy weight method to obtain the combined weight of each evaluation index; the initial decision matrix is normalized, and the combined weight is used to construct a weighted and normalized decision matrix;

[0066] A scheme set screening module determines the positive ideal solution and the negative ideal solution; calculates the distance between each candidate solution and the positive and negative ideal solutions, and according to the distance, calculates the relative closeness degree for preliminary sorting, and screens out an optimal scheme set;

[0067] A collaborative design scheme sorting module extracts the design parameter vectors of each scheme in the optimal scheme set, constructs a multi-dimensional joint probability distribution model based on the kernel density estimation method, generates a group of derivative candidate solutions through Monte Carlo sampling; calculates the values of each evaluation index of the group of derivative candidate solutions, merges the values of each evaluation index with the initial candidate solutions, and executes the steps of constructing a weighted decision matrix and sorting again on the merged scheme set to obtain the collaborative design scheme sorting.

[0068] ​It should be noted that, for the aforementioned method embodiments, the sequences of the described actions are not necessarily required to implement the present application, and certain actions can be performed in other sequences, or even at the same time, in accordance with the present application. Furthermore, certain actions can not be required to implement the present application. Additionally, the described embodiments are not necessarily the only possible implementation of the present application.

[0069] In the above embodiments, the description of each embodiment focuses on different aspects, and the parts not described in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0070] The preferred embodiments of the present application disclosed above are only used to clarify the present application. The alternative embodiments do not describe all the details and limit the present application to the specific embodiments described. Obviously, according to the content of the embodiments of the present application, many modifications and changes can be made. The embodiments are selected and described in detail in order to better explain the principles and practical applications of the embodiments of the present application, so that those skilled in the art can well understand and use the present application. The present application is limited by the claims and their full scope and equivalents.

Claims

1. A collaborative design method for sound barriers based on the NIMBY (Not In My Backyard) effect, characterized in that, Includes the following steps: Multiple alternative sound barrier schemes defined by their respective design parameter vectors are obtained, and an initial decision matrix is ​​constructed that includes evaluation indicators such as acoustic performance, structural safety, and NIMBY effect. For each alternative scheme, a three-dimensional ray tracing model of solar shading and a spherical projection model of field of view encroachment are established. The model output values ​​are weighted with the landscape harmony expert scores to obtain the NIMBY effect evaluation value of each alternative scheme. A combined weighting method is used to integrate the subjective weights determined by the analytic hierarchy process and the objective weights determined by the entropy weight method to obtain the combined weights of each evaluation index; the initial decision matrix is ​​normalized, and a weighted normalized decision matrix is ​​constructed using the combined weights; Determine the positive and negative ideal solutions, calculate the distance between each alternative solution and the positive and negative ideal solutions, and perform preliminary sorting based on the relative proximity calculated according to the distances shown, and select the preferred solution set; Extract the design parameter vectors of each scheme in the preferred scheme set, construct a multidimensional joint probability distribution model based on the kernel density estimation method, and generate a set of derivative alternative schemes through Monte Carlo sampling; Calculate the evaluation index values ​​of the set of derived alternative solutions, merge the evaluation index values ​​with the initial alternative solutions, and perform the steps of constructing a weighted decision matrix and ranking the merged solution set again to obtain the ranking of collaborative design solutions.

2. The method according to claim 1, characterized in that, The step of weighting the model output values ​​with landscape harmony expert scores to obtain the NIMBY (Not In My Backyard) effect evaluation values ​​for each alternative scheme includes: For any alternative scheme, a three-dimensional geometric model including the sound barrier, surrounding buildings and solar trajectory is established, and the average solar shading and average field of view occupancy of the three-dimensional geometric model are calculated at multiple preset observation points and time periods. Transform landscape harmony expert scores into penalty items; By using a preset weighting function, the average solar shading, average field of view encroachment, and landscape harmony penalty are combined to obtain the NIMBY effect evaluation value of the alternative scheme.

3. The method according to claim 1, characterized in that, The design parameter vector includes: The height, length, material type, color scheme, and lateral position of the sound barrier relative to the road.

4. The method according to claim 1, characterized in that, The method of combining subjective weights determined by the analytic hierarchy process (AHP) and objective weights determined by the entropy weight method to obtain the combined weights of each evaluation index includes: Construct pairwise comparison judgment matrices for each evaluation indicator, and calculate the subjective weight vector after passing the consistency test. ; Based on the initial decision matrix, the information entropy of each indicator is calculated, and the objective weight vector is obtained based on the information entropy. ; Using linear combination formula The combined weights are calculated, where α is a preset subjective weight preference coefficient with a value between 0 and 1.

5. The method according to claim 1, characterized in that, The determination of the positive ideal solution and the negative ideal solution includes: For acoustic performance indicators, the maximum value of the acoustic performance indicator is taken as the positive ideal solution component, and the minimum value is taken as the negative ideal solution component. For structural safety indices, the minimum value of the structural safety index is taken as the positive ideal solution component, and the maximum value is taken as the negative ideal solution component. For the NIMBY (Not In My Backyard) effect index, the minimum value of the evaluation of all alternative solutions is taken as the positive ideal solution component, and the maximum value is taken as the negative ideal solution component.

6. The method according to claim 1, characterized in that, The calculation of the distance between each alternative solution and the positive and negative ideal solutions includes: The covariance matrix S among the evaluation indicators is calculated based on the weighted normalized decision matrix, and the inverse matrix of the covariance matrix is ​​used. According to the formula Calculate the distance between each alternative solution and the positive ideal solution and the negative ideal solution respectively.

7. The method according to claim 1, characterized in that, The multidimensional joint probability distribution model constructed based on the kernel density estimation method generates a set of derivative alternatives through Monte Carlo sampling, including: Extract the design parameter vectors of all schemes in the preferred scheme set to form a multi-dimensional sample set; The bandwidth of the kernel function is determined by a bandwidth selection method, and the kernel density is estimated on the multidimensional sample set to construct a multidimensional joint probability density function of the design parameter vector. Based on the constructed joint probability density function, random sampling is performed using the Monte Carlo method to generate a new set of design parameter vectors as a set of derivative alternatives.

8. A collaborative design system for sound barriers based on the NIMBY (Not In My Backyard) effect, characterized in that, Includes the following modules: The matrix construction module obtains multiple sound barrier alternatives defined by their respective design parameter vectors, and constructs an initial decision matrix that includes evaluation indicators such as acoustic performance, structural safety, and NIMBY effect. For each alternative scheme, a three-dimensional ray tracing model of solar shading and a spherical projection model of field of view encroachment are established. The model output values ​​are weighted with the landscape harmony expert scores to obtain the NIMBY effect evaluation value of each alternative scheme. The weight fusion module uses a combined weighting method to fuse the subjective weights determined by the analytic hierarchy process and the objective weights determined by the entropy weight method to obtain the combined weights of each evaluation index; the initial decision matrix is ​​normalized, and a weighted normalized decision matrix is ​​constructed using the combined weights; The solution set filtering module determines the positive and negative ideal solutions; Calculate the distance between each alternative solution and the positive and negative ideal solutions, and perform preliminary sorting based on the relative proximity calculated according to the distances shown, and select the preferred solution set; The collaborative design scheme ranking module extracts the design parameter vectors of each scheme in the preferred scheme set, constructs a multidimensional joint probability distribution model based on the kernel density estimation method, and generates a set of derivative alternative schemes through Monte Carlo sampling. Calculate the evaluation index values ​​of the set of derived alternative solutions, merge the evaluation index values ​​with the initial alternative solutions, and perform the steps of constructing a weighted decision matrix and ranking the merged solution set again to obtain the ranking of collaborative design solutions.

9. The system according to claim 8, characterized in that, The step of weighting the model output values ​​with landscape harmony expert scores to obtain the NIMBY (Not In My Backyard) effect evaluation values ​​for each alternative scheme includes: For any alternative scheme, a three-dimensional geometric model including the sound barrier, surrounding buildings, and solar trajectory is established. The average solar shading and average view encroachment of the three-dimensional geometric model are calculated at multiple preset observation points and time periods. The landscape harmony expert score is converted into a penalty item. The average solar shading, average view encroachment, and landscape harmony penalty item are combined through a preset weighting function to obtain the NIMBY effect evaluation value of the alternative scheme.

10. The system according to claim 8, characterized in that, The design parameter vector includes: The height, length, material type, color scheme, and lateral position of the sound barrier relative to the road.

Images