Data-driven construction project industrialization company site selection optimization method and device

Abstract

The invention provides a data-driven construction project industrial company site selection optimization method and device. The method comprises the following steps: S1, selecting and determining main influence factors influencing construction project industrial company site selection; s2, determining the importance degree of each main influence factor to a target by using a 1-9 scaling method, and calculating the subjective weight of each main influence factor by using an analytic hierarchy process; s3, calculating objective weights of the main influence factors by using an entropy evaluation method according to related existing data, and determining subjective and objective comprehensive weights of the main influence factors in combination with the step S2; and S4, comprehensively evaluating and sorting the site selection schemes according to the combination of the subjective and objective comprehensive weight and an ideal approximation method, and preferentially selecting site selection for the construction project industrial company. According to the evaluation system adopted by the invention, subjective empowerment is calculated through an analytic hierarchy process, and objective weight is obtained through a data-driven entropy evaluation method, so that the personal intention of a leader is effectively weakened, and the subjective influence of expert preference on evaluation indexes is reduced.

Description

Technical Field

[0001] This invention relates to the field of intelligent logistics supply chain production and inventory location selection, specifically to a data-driven method and apparatus for optimizing the location of industrialized construction engineering companies. Background Technology

[0002] Under the requirements of new quality productivity, the development of construction engineering has entered a new era. The traditional construction industry suffers from problems such as low efficiency, resource waste, and environmental pollution, while industrialized production models can significantly improve efficiency, reduce costs, and decrease carbon emissions. Therefore, establishing Construction Engineering Industrialization Companies (CEICs) is a trend and a current in promoting the industrialization of construction engineering. Site selection is a key link in promoting the industrialization process of construction engineering. The background technology of site selection involves multiple dimensions of factors, including policy guidance, technological development, market demand, resource distribution, environmental sustainability, and economic costs. Through scientific site selection strategies, both the efficiency of construction engineering industrialization can be improved, and the industry can be driven towards a green, intelligent, and sustainable transformation. In the future, with technological advancements and policy deepening, company site selection must place greater emphasis on regional development planning, supply chain collaboration, social benefits, sustainable development, and iterative updates.

[0003] Current site selection methods for industrialized construction companies have the following problems.

[0004] I. Reliance on traditional experience and lack of multi-objective decision-making

[0005] Currently, many site selection plans still rely on expert experience or historical corporate data, primarily driven by experience, and sometimes even guided by leadership preferences or wishes. Interdisciplinary integration is weak, cost-benefit analysis is one-sided, and analysis is mainly based on engineering and economics, failing to consider issues from interdisciplinary perspectives such as management science and sociology. The lack of systematic quantitative analysis tools leads to strong subjectivity and poor objectivity, making it difficult to make reasonable choices in complex environments. Furthermore, there is often a singular focus on economic indicators, insufficient consideration of multiple objectives, and neglect of the synergistic optimization of multi-dimensional goals.

[0006] II. The evaluation indicators are not comprehensive, and sustainable development is not prominent.

[0007] Because multi-objective decision-making is impossible, key indicators are often missing and sensitive indicators are avoided when selecting evaluation metrics. Currently, most construction industrialization companies are invested in by state-owned enterprises, forming part of state-owned capital and endowed with certain social functions (such as emergency relief, flood control, and fire prevention). However, these social benefit indicators were not considered in traditional site selection. As state-owned capital, sustainable development and iteration indicators are fundamental to its continued development and the preservation and appreciation of state-owned capital. Therefore, potential for site renovation and technological upgrading are the most important evaluation indicators. Furthermore, past evaluation indicators focused on local economic indicators or economic development, rarely considering the impact of industrial clusters and social development trends on industrialization companies. The technology spillover and regional economic driving effects brought about by industrial clusters are often overlooked, leading to an underestimation of the comprehensive value of sites with these effects.

[0008] Third, the analysis methods are unscientific and the model has poor reliability.

[0009] Current analytical methods are mostly based on financial management and accounting, focusing on economic indicators such as break-even point, internal rate of return, and payback period. They fail to incorporate evaluation methods from management science and operations research. Policy-related implicit costs (such as regional transportation costs, carbon emission costs, and environmental impact costs) are ignored, and social costs (such as additional communication costs due to community opposition and legal dispute costs) are not included. Some qualitative indicators (such as social acceptance and policy stability) and indicators that are difficult to quantify often rely on subjective scoring. All of these factors contribute to unscientific analytical methods that fail to accurately reflect the comprehensive situation of site selection for industrialized construction companies, thus reducing the reliability of the models.

[0010] IV. Primarily based on static models, with minimal involvement of dynamic data.

[0011] Existing models rely on overly simplistic assumptions, primarily static models, and are based on fixed parameters (such as fixed market demand and fixed costs). They fail to dynamically reflect the impact of policy changes, technological iterations, or market fluctuations, and do not account for future automation upgrades or supply chain changes. Furthermore, they lack sufficient assessment of resilience to dynamic factors such as real-time traffic flow, regional policy adjustments, and natural disasters, and lack flexible site selection strategies. In multi-criteria decision-making, parameter weights depend on expert scoring, which is subjective and lacks objective quantitative evidence, leading to significant discrepancies in site selection conclusions among different scorers.

[0012] Under the constraint of limited investment resources, the site selection of construction industrialization companies, which are state-owned capital, within a fixed region is essentially a ranking and optimization problem. The Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) provides an intuitive and simple method that ranks alternative solutions by calculating the Euclidean distance between them and the ideal and negative ideal solutions. It is a very suitable optimization method for the site selection of construction industrialization companies.

[0013] Regarding the calculation of the evaluation weights for each alternative scheme, although each evaluation method has its own advantages, it also has its shortcomings. Summary of the Invention

[0014] To address the above technical problems, this invention provides a data-driven method and apparatus for optimizing site selection in industrialized construction engineering companies.

[0015] The technical solution of this invention is: a data-driven optimization method for site selection of industrialized construction engineering companies, comprising:

[0016] A data-driven site selection optimization method for industrialized construction companies includes:

[0017] Step S1: Select and determine the main factors influencing the site selection of the construction industrialization company;

[0018] Step S2: Use the 1-9 scale method to determine the importance of each major influencing factor to the target, and use the analytic hierarchy process (AHP) to calculate the subjective weight of each major influencing factor;

[0019] Step S3: Based on the relevant existing data, calculate the objective weight of each major influencing factor using the entropy method, and combine it with Step S2 to determine the combined subjective and objective weight of each major influencing factor;

[0020] Step S4: Based on a combination of subjective and objective weighting and the ideal approximation method, the site selection schemes are comprehensively evaluated and ranked, and the best site is selected for the construction engineering industrialization company.

[0021] A data-driven site selection optimization device for industrialized construction engineering companies, comprising:

[0022] The influencing factors determination module is used to select and determine the main influencing factors affecting the site selection of construction engineering industrialization companies;

[0023] The subjective weight calculation module is used to determine the importance of each major influencing factor to the target using the 1-9 scale method, and to calculate the subjective weight of each major influencing factor using the analytic hierarchy process.

[0024] The subjective and objective comprehensive weight calculation module is used to calculate the objective weight of each major influencing factor based on relevant existing data and predicted data using the entropy method, and to determine the subjective and objective comprehensive weight of each major influencing factor in conjunction with step S2.

[0025] The ranking module is used to comprehensively evaluate and rank the site selection schemes based on a combination of subjective and objective weights and the ideal approximation method, and select the best site for the construction engineering industrialization company.

[0026] A computing device includes: at least one processor and a memory storing program instructions; when the program instructions are read and executed by the processor, the computing device causes the computing device to perform a data-driven site selection optimization method for industrialized construction companies.

[0027] A readable storage medium storing program instructions, which, when read and executed by a computing device, cause the computing device to perform a data-driven site selection optimization method for industrialized construction companies.

[0028] Compared with the prior art, the present invention has the following advantages:

[0029] The influencing factors identified in this invention are comprehensive, not only focusing on general economic indicators and short-term corporate conditions, but also taking into full account the characteristics of construction engineering industrialization companies, sustainable development, factors affecting updates and iterations, and social benefits.

[0030] The evaluation system adopted in this invention fully incorporates the intentions of leaders and the opinions of experts, calculates subjective weighting through the analytic hierarchy process, and obtains objective weights through the entropy method driven by the data itself. By using a subjective-objective fusion model, it effectively weakens the subjective influence of leaders' personal intentions and experts' preferences on evaluation indicators, strengthens the role of data objectivity in decision-making, reduces the weight of subjective decisions, and increases the weight of the objective needs of the enterprise's own development and real-time situation, thus providing a more scientific quantitative basis for site selection decisions.

[0031] This invention constructs a comprehensive quantitative evaluation model that integrates multi-dimensional indicators from management and operations research, and adopts a system multi-objective decision-making algorithm to achieve quantitative evaluation and ranking of site selection schemes. Attached Figure Description

[0032] Figure 1 This is a flowchart of a data-driven site selection optimization method for industrialized construction companies, as described in an embodiment of the present invention.

[0033] Figure 2 This is a diagram of a hierarchical structure model constructed using the analytic hierarchy process in an embodiment of the present invention.

[0034] Figure 3This is a flowchart illustrating the calculation of subjective weights of influencing factors using the analytic hierarchy process in an embodiment of the present invention.

[0035] Figure 4 This is a flowchart illustrating the process of calculating the objective weights of each influencing factor using the entropy method and comprehensively evaluating and ranking the site selection schemes using the ideal approximation method in this embodiment of the invention. Detailed Implementation

[0036] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0037] This invention provides a data-driven optimization method for the site selection of industrialized construction companies, thereby improving the scientific rigor of site selection. To enhance the robustness and reliability of the evaluation results, this invention employs a combination evaluation method (CEM).

[0038] Combinatorial evaluation methods integrate the results or processes of two or more individual comprehensive evaluation techniques to achieve a more robust and comprehensive assessment. This approach fully recognizes the inherent advantages and limitations of each evaluation method; by organically combining them, the strengths of each method can be fully utilized while compensating for each other's weaknesses.

[0039] In the combined evaluation method employed in this invention, the Analytic Hierarchy Process (AHP) is used to calculate the subjective weights of the scheme indicators, and the entropy method is used to calculate the objective weights of the scheme indicators. Wherein:

[0040] The Analytic Hierarchy Process (AHP) provides a structured and systematic approach to handling complex decision-making problems involving multiple evaluation criteria. Its main advantage lies in its ability to systematically incorporate subjective expert opinions and preferences into the decision-making process. This is particularly applicable to decision-making indicators for site selection in construction industrialization companies that involve evaluation indicators that are difficult to quantify or strategic issues.

[0041] The analytic hierarchy process (AHP) uses a clear hierarchical structure to help break down complex problems into more manageable components, while pairwise comparisons effectively guide experts to express their judgments on indicators in a structured way.

[0042] The main advantage of the entropy method lies in its objectivity. The weights of the indicators it ultimately determines are entirely determined by the distribution characteristics of the original data itself, without relying on the subjective judgment of experts. This is crucial for effectively mitigating the subjective influence of leaders' personal intentions and reducing expert preferences on evaluation indicators.

[0043] Entropy methods are often applied in objective, data-driven scenarios. When dealing with large datasets containing numerous evaluation metrics, subjective weighting methods can be difficult to implement and prone to inconsistencies, while entropy methods can effectively address this issue. Furthermore, entropy methods can also play a crucial role when there is a lack of clear theoretical basis or expert consensus to determine the weights of each metric.

[0044] The entropy method, a data-driven approach, can effectively reduce human bias and thus improve the credibility of evaluation results, making it particularly suitable for scenarios that require objective weighting.

[0045] The main advantage of the comprehensive evaluation method composed of the analytic hierarchy process and the entropy method is that it can introduce different perspectives and methods, and the evaluation results no longer rely excessively on the specific assumptions and limitations of a single technology.

[0046] The combined evaluation method integrates objective and subjective weighting methods, comprehensively considering the evaluation problem from different dimensions. It can achieve a balance between data-driven insights and expert experience, thereby reducing the bias that may exist in a single method and ultimately improving the credibility and acceptability of the evaluation or ranking results.

[0047] Figure 1 This is a flowchart illustrating a data-driven site selection optimization method for industrialized construction companies, as described in an embodiment of the present invention. Figure 1 As shown, the method includes the following steps:

[0048] Step S1: Select and determine the main factors influencing the site selection of the construction industrialization company;

[0049] Step S2: Use the 1-9 scale method to determine the importance of each major influencing factor to the target, and use the analytic hierarchy process (AHP) to calculate the subjective weight of each major influencing factor;

[0050] Step S3: Based on the relevant existing data, calculate the objective weight of each major influencing factor using the entropy method, and combine it with Step S2 to determine the combined subjective and objective weight of each major influencing factor;

[0051] Step S4: Based on a combination of subjective and objective weighting and the ideal approximation method, the site selection schemes are comprehensively evaluated and ranked, and the best site is selected for the construction engineering industrialization company.

[0052] Step S1 includes: based on expert interviews and literature review related to the site selection of construction industrialization companies, and taking into full account the company's cost-benefit indicators, social indicators, sustainable development and iterative updates indicators, selecting and determining the main influencing factors affecting the site selection of construction industrialization companies.

[0053] Based on expert interviews and site selection literature, in addition to considering economic costs and benefits, the impact on the company throughout its entire life cycle is considered from the perspectives of social benefits, sustainable development and upgrading of the enterprise, and the main influencing factors affecting the site selection of construction engineering industrialization companies are selected and determined.

[0054] Table 1 shows the main influencing factors on the site selection of construction industrialization companies as determined in the embodiments of the present invention, and the evaluation index system constructed therefrom:

[0055] Table 1. Site Selection Evaluation Index System for Construction Engineering Industrialization Companies

[0056]

[0057] Step S2: Use the 1-9 scale method to determine the importance of each major influencing factor to the objective, and use the analytic hierarchy process (AHP) to calculate the subjective weight of each major influencing factor, specifically including:

[0058] I. Construction of Site Selection Evaluation Index System

[0059] First, construct an evaluation index system: build a framework including the target layer, the criterion layer (first-level indicators), the indicator layer (second-level indicators), and the scheme layer (third-level indicators).

[0060] Target layer: refers to the site selection of construction engineering industrialization companies, which is the highest level of the evaluation index system.

[0061] Criterion layer: refers to the factors that influence the decision-making process for the target, or the intermediate links in achieving the target. It is the intermediate layer of the evaluation indicator system.

[0062] Indicator layer: refers to the sub-factors that influence the decision-making process for the target, or the sub-steps in the intermediate steps to achieve the target. It is the intermediate layer of the evaluation indicator system.

[0063] Solution layer: refers to the specific options and measures that can be selected, and is the lowest level of the evaluation indicator system.

[0064] This constitutes a hierarchical structure model, where elements at higher levels act as criteria and govern related elements at lower levels, such as... Figure 2 As shown.

[0065] II. Determining Subjective Weights using the Analytic Hierarchy Process (AHP)

[0066] After the evaluation index system is constructed, the weights of the primary indicators are generally calculated first, followed by the weights of the secondary indicators. Then, the weights of the primary and secondary indicators are multiplied together to obtain the comprehensive weight. In this embodiment, firstly, judgment matrices are constructed for the primary indicators B1, B2, B3, B4, and B5 to calculate their weight values; then, judgment matrices are constructed for the secondary indicators C1-C3 under B1 to calculate their weight values; for the secondary indicators C4-C8 under B2 to calculate their weight values; for the secondary indicators C9-C12 under B3 to calculate their weight values; for the secondary indicators C13-C15 under B4 to calculate their weight values; and for the secondary indicators C16-C18 under B5 to calculate their weight values. Finally, the weights of the primary and secondary indicators are multiplied together to obtain the comprehensive weight. Figure 3 As shown, the specific calculation steps are as follows:

[0067] S2-1, Determine the judgment matrix of the criterion layer.

[0068] For the target layer, pairwise comparisons are performed on each criterion in the criterion layer, and the 1-9 scale method is used to determine their importance to the target. The comparison results are then filled into the judgment matrix. middle.

[0069] ,in Represents element Relative to element The degree of importance.

[0070] The judgment matrix has the following properties:

[0071] ;

[0072] ;

[0073] .

[0074] Table 2 is a table of 1-9 scale methods used in the embodiments of the present invention.

[0075] Table 2.1-9 Scale Method Table

[0076]

[0077] S2-2, Calculate the eigenvectors of the judgment matrix.

[0078] First, the eigenvectors must be normalized: normalize each component of the eigenvector according to the column vector so that the sum of the components is 1.

[0079] Then, sum the row vectors and normalize them again to obtain the eigenvectors. .

[0080] because It is an approximate eigenvector. Approximate to , It is a judgment matrix eigenvalues, It is the largest eigenvalue.

[0081] For the judgment matrix eigenvectors , largest eigenvalue Satisfy the equation Therefore, each component satisfies .

[0082] The formula for calculating the largest eigenvalue is: by averaging over all components, a eigenvalue can be obtained. Estimate .

[0083] The elements are the ranking weights of factors at the same level relative to factors at the next higher level. This process is called hierarchical single ranking.

[0084] Whether a hierarchical single order can be confirmed requires a consistency check. A consistency check refers to... Determine the acceptable range of inconsistencies.

[0085] Using the eigenvector corresponding to the largest eigenvalue as the weight vector for the degree of influence of the compared factor on a higher-level factor, the greater the inconsistency, the greater the judgment error. Therefore, it can be used... Measured by numerical value The degree of inconsistency, This is the dimension of the matrix. Therefore:

[0086] S2-3, Calculate the consistency ratio

[0087] The core of the analytic hierarchy process (AHP) lies in constructing a judgment matrix to compare the relative importance of different factors. However, people's judgments often exhibit a degree of inconsistency. For example, if... Compare important, Compare It is important, so logically... It should be more than More importantly. But people may make mistakes in their actual judgments. Better This inconsistency arises in important situations.

[0088] Because the consistency matrix has the following properties:

[0089] If a set of judgment matrices is consistent, then its largest eigenvalue is exactly equal to the order of the matrix. If they are inconsistent, the largest eigenvalue will be greater than... Therefore, by calculating the largest eigenvalue, the consistency ratio (CR) of the judgment matrix can be evaluated, thereby verifying whether the judgment matrix is ​​acceptable.

[0090] The calculation steps are as follows:

[0091] S2-3-1, Calculate the consistency index The formula for the consistency index is: .

[0092] in, It determines the order (i.e., the dimension) of a matrix.

[0093] S2-3-2, Finding the random consistency index The random consistency index is an average consistency index calculated from a large number of random matrices.

[0094] Appendix 3 shows the random consistency in the embodiments of the present invention. Lookup table;

[0095] Table 3 Random Consistency lookup table

[0096]

[0097] S2-3-3, Calculate the consistency ratio The formula for the consistency ratio is: .if If the matrix has satisfactory consistency, it passes the consistency test.

[0098] Otherwise, the judgment matrix needs to be adjusted until the consistency requirement is met.

[0099] Obtain the criterion layer judgment matrix of The factors are ranked hierarchically as follows: .

[0100] S2-4, Determine the judgment matrix for the indicator layer.

[0101] The hierarchical order of the indicator layers is calculated according to the above steps. .

[0102] but The overall inter-layer sorting is as follows:

[0103] Inter-layer total sorting calculation table

[0104]

[0105] S2-5, Consistency check of hierarchical overall sorting

[0106] set up layer For the upper level ( Factors in the layer The hierarchical single-order consistency index is The random consistency index is The consistency ratio of the overall hierarchical ranking is:

[0107] ,

[0108] when At that time, the overall hierarchical ranking is considered to have passed the consistency test.

[0109] The overall hierarchical sorting has satisfactory consistency; otherwise, the element values ​​of the judgment matrices with high consistency ratios need to be readjusted until the consistency requirements are met.

[0110] S2-6, Determine the subjective weights of each influencing factor based on the overall hierarchical ranking of the indicator layer.

[0111]

[0112] , j is the j-th influencing factor in criterion i, and n is the total number of influencing factors.

[0113] Step S3: Based on the relevant existing data, calculate the objective weight of each major influencing factor using the entropy method, and combine it with Step S2 to determine the comprehensive weight of each major influencing factor.

[0114] I. Data Preparation

[0115] 1. Construction cost indicators

[0116] Land costs are based on recent auction prices of similar land, existing data, and relevant policies.

[0117] Fixed asset investment is determined with reference to the pricing rules of the bill of quantities for construction projects issued by ministries and provincial governments, and in conjunction with market prices.

[0118] Warehousing costs take into account the fixed storage costs for emergency relief and the storage costs of raw materials, finished components, semi-finished components, and consumables. They also consider the ease of transporting raw materials and the accessibility of transportation within the business's coverage area. For example, if the location has access to water and rail transport, the inventory costs of raw materials and consumables are lower. Higher road accessibility within the business's coverage area results in lower inventory costs for finished components and semi-finished products. Conversely, lower accessibility leads to higher inventory costs.

[0119] 2. Operating costs

[0120] Raw material costs take into account market information prices, local mineral resource exploitation potential, and the improvement of water transport capacity in medium- and long-term plans.

[0121] Labor costs take into account existing market labor prices, the urbanization situation of population outflow, and the return of population.

[0122] Transportation costs take into account transportation prices within the business's coverage area, as well as the medium- and long-term planning and construction of the transportation network (especially the reconstruction and expansion of national and provincial highways and county and township roads).

[0123] Ecological protection costs include carbon emission costs, waste disposal costs, cold recycling and secondary processing costs, air and water pollution costs, noise-related treatment costs, and related communication and coordination costs with surrounding residents.

[0124] Management costs take into account the cost changes following updates and iterations of management technologies and tools.

[0125] 3. Economic benefits

[0126] This section of data is based on existing data and economic and technical indicators.

[0127] 4. Social benefits

[0128] With increased efforts to protect intellectual property rights, the recognition of high-tech enterprises, as a form of tax and other related policy subsidy, is part of the long-term benefits.

[0129] When natural disasters strike, the materials, personnel, and equipment stored by construction industrialization companies can be deployed to the scene as a rescue force, reducing the losses caused by natural disasters and thus gaining a good reputation and brand effect, which is part of the social benefits.

[0130] The industrial cluster effect can drive construction investment and total investment scale, which is an important benefit that enterprises need to consider in their development.

[0131] 5. Sustainable development capability

[0132] The sustainable development capability of construction industrialization companies is crucial. Many existing industrialization companies suffer from outdated production capacity, shutdowns, or even demolitions due to a lack of upgradable sites or inability to upgrade their technology. Therefore, whether there is reserved land for upgrading and renovation around existing sites, and whether equipment has ports for technological upgrades, are important factors to consider for sustainable development capability.

[0133] Policy support in different regions (such as whether the project is introduced as an investment attraction project and whether tax subsidies are available) are also key factors that construction industrialization companies consider.

[0134] II. Calculation of Objective Weights using the Entropy Method

[0135] Entropy method is an objective weighting technique based on information theory. Its core idea is to determine the weight of each evaluation index by analyzing the amount of information contained in each index.

[0136] In this approach, entropy is considered as a measure of the uncertainty or randomness of a system.

[0137] If the entropy value of an evaluation indicator is high, it means that the numerical differences between the alternatives on that indicator are small, and the information provided by that indicator is less discriminative, so it is usually given a lower weight. Conversely, an indicator with a low entropy value indicates that the numerical differences between the alternatives on that indicator are large, and the information is more discriminative, so it will receive a higher weight.

[0138] The objective weights of the scheme are calculated using the entropy method. When assigning weights to combinations, the entropy method generally involves directly calculating the weights of the secondary indicators.

[0139] like Figure 4 As shown, the implementation process of the entropy method typically includes the following steps:

[0140] S3-1, Construct the original matrix: Assume there are m main influencing factors and n alternative solutions. Use the values ​​of each alternative solution on each main influencing factor to form the original matrix. ,

[0141] in, It is the value of the i-th alternative in the j-th major influencing factor.

[0142] S3-2, Normalization: Normalize the original matrix to eliminate the influence of different indicator dimensions and value ranges, and ensure that all indicators are on a comparable scale.

[0143] For positive primary influencing factors (such as returns), then .

[0144] For negative primary influencing factors (such as cost), then .

[0145] The standardized decision matrix is ​​obtained. .

[0146] S3-3, Calculate the relative contribution of the main influencing factors. For the normalized data, calculate the proportion of each alternative's value under each main influencing factor to the sum of the values ​​of all alternatives under that main influencing factor. This proportion reflects the relative contribution of each alternative to that main influencing factor:

[0147] ,

[0148] in, It is the first The first scheme is in the The values ​​under the main influencing factors, It is the number of solutions. It is the number of major influencing factors.

[0149] S3-4, calculate the entropy value of each major influencing factor. The entropy formula quantitatively describes the degree of uncertainty of the indicator based on the above proportions. A lower entropy value means that the alternatives perform differently on the indicator, indicating that the indicator value has a large degree of variation, provides more information, has a strong distinguishing ability, plays a greater role in the comprehensive evaluation, and should have a larger weight. Conversely, a higher entropy value means that the alternatives perform similarly on the indicator, indicating that the indicator value has a small degree of variation, provides less information, has a weaker distinguishing ability, plays a smaller role in the comprehensive evaluation, and should have a smaller weight.

[0150] Calculate the first Entropy values ​​of the main influencing factors:

[0151] .

[0152] S3-5, Calculate the difference coefficient (redundancy) of entropy values:

[0153] Difference coefficient (redundancy) .

[0154] A larger difference coefficient indicates that the indicator plays a greater role in the alternative options, and that the indicator is better.

[0155] S3-6, Calculate the objective weights of each major influencing factor:

[0156] , .

[0157] III. Calculating the weight of the comprehensive index

[0158] Since both subjective and objective weights have been obtained, the results of the two weights are combined using the multiplication method to calculate the final combined weight, also known as the comprehensive weight.

[0159] As described above, in step S2, the subjective weights were calculated using the analytic hierarchy process (AHP). .

[0160] Based on the aforementioned subjective and objective weights, the weight of the comprehensive index is calculated using a combination of subjective and objective weighting methods:

[0161] ,

[0162] The result is .

[0163] Step S4: Based on a combination of subjective and objective weighting and the ideal approximation method, comprehensively evaluate the site selection scheme.

[0164] Evaluate and rank them.

[0165] The ideal approximation method is a commonly used multi-criteria decision analysis method. Its core idea is to rank alternatives based on their proximity to the ideal solution and their distance from the negative ideal solution. The most popular alternative should have the shortest distance to the ideal solution and the greatest distance from the negative ideal solution in the multi-dimensional space.

[0166] A positive ideal solution is a hypothetical optimal solution that achieves the best value across all considered evaluation metrics; while a negative ideal solution is a hypothetical worst solution composed of the worst values ​​across all evaluation metrics.

[0167] The application of the ideal approximation method typically involves the following steps:

[0168] S4-1, Construct a decision matrix: each row represents an alternative solution, and each column represents a major influencing factor. The elements in the matrix represent the values ​​of each alternative solution on different major influencing factors.

[0169] S4-2 normalizes the original data in the decision matrix to eliminate the incomparability between evaluation indicators of different dimensions.

[0170] S4-3, Calculate the weighted decision matrix: Multiply each column of the normalized decision matrix by the combined subjective and objective weights of the corresponding main influencing factors to obtain the weighted normalized decision matrix:

[0171] .

[0172] S4-4, Calculate the positive and negative ideal solutions: Determine the positive and negative ideal solutions from the weighted normalized decision matrix. For benefit-type indicators (i.e., the larger the indicator value, the better), the positive ideal solution takes the maximum value for that indicator, while the negative ideal solution takes the minimum value; for cost-type indicators (i.e., the smaller the indicator value, the better), the opposite is true, the positive ideal solution takes the minimum value, while the negative ideal solution takes the maximum value.

[0173] Calculate the positive ideal solution. The positive ideal solution is derived from the weighted normalized decision matrix. The maximum element of each column vector (the minimum element of the negative index column) constitutes the structure. .

[0174] Calculating the negative ideal solution: The negative ideal solution is obtained by weighted normalized decision matrix. It is composed of the smallest element of each column vector (the largest element of the negative index column). .

[0175] S4-5, Calculate Euclidean distance: For each alternative, calculate its Euclidean distance to the ideal solution and the negative ideal solution.

[0176] Calculate the Euclidean distance between each alternative solution and the positive ideal solution. .

[0177] Calculate the Euclidean distance between each alternative solution and the negative ideal solution. .

[0178] S4-6 Calculate the relative proximity coefficient for each alternative solution. The relative proximity coefficient is the distance between the alternative solution and the negative ideal solution, divided by the sum of the distance between the alternative solution and the positive ideal solution and the distance between the alternative solution and the negative ideal solution. The value of this coefficient is between 0 and 1, and the higher the value, the closer the alternative solution is to the ideal solution.

[0179] Proximity coefficient .

[0180] S4-7, Ranking the Options. Sort the candidate options from highest to lowest based on their proximity coefficient. The closer the coefficient is to 1, the closer the alternative is to the optimal level. Conversely, the closer the coefficient is to 0, the closer the alternative is to the worst-case level. The alternative with the highest relative closeness coefficient is considered the best choice.

[0181] This invention also provides a data-driven site selection optimization device for industrialized construction companies, comprising:

[0182] The influencing factors determination module is used to select and determine the main influencing factors affecting the site selection of construction engineering industrialization companies;

[0183] The subjective weight calculation module is used to determine the importance of each major influencing factor to the target using the 1-9 scale method, and to calculate the subjective weight of each major influencing factor using the analytic hierarchy process.

[0184] The subjective and objective comprehensive weight calculation module is used to calculate the objective weight of each major influencing factor based on relevant existing data using the entropy method, and to determine the subjective and objective comprehensive weight of each major influencing factor in conjunction with step S2.

[0185] The ranking module is used to comprehensively evaluate and rank the site selection schemes based on a combination of subjective and objective weights and the ideal approximation method, and select the best site for the construction engineering industrialization company.

[0186] Those skilled in the art, through the description of the above embodiments, can clearly understand that these embodiments can be implemented either solely by software or in conjunction with necessary general-purpose hardware platforms, such as:

[0187] SPSS itself does not have a dedicated module for Analytic Hierarchy Process (AHP). Therefore, implementation typically requires manual pairwise comparisons, eigenvector calculations, and consistency checks, or these steps can be implemented using scripts. R provides a package called Analytic Hierarchy Process (AHP), which offers comprehensive functionality for building AHP models, performing pairwise comparisons, calculating weights, and checking consistency. Python also provides libraries such as PyAHP, which can be used to build hierarchical structures and perform pairwise comparisons to derive priority weights in the Analytic Hierarchy Process (AHP).

[0188] SPSS doesn't have a dedicated entropy function, but users can utilize its matrix operations and mathematical functions to perform steps such as normalization, scaling, entropy calculation, and weight determination. R can implement entropy calculations through custom functions or by using packages like DecisionSupport, which typically provide functions for calculating entropy. Python, with its powerful numerical computing libraries (such as NumPy and Pandas), is well-suited for implementing the mathematical operations involved in entropy calculations. Users can write concise code snippets to perform data normalization, scaling, entropy calculation, and final weight determination.

[0189] In SPSS, matrix operations and mathematical functions can be used to perform normalization, weighting, determination of ideal and negative ideal solutions, distance calculation, and solving for proximity coefficients. In R, the dedicated TOPSIS package provides complete TOPSIS analysis functions. In Python, numerical computing libraries such as NumPy and Pandas can be used to implement all computational steps of the ideal approximation algorithm through array operations.

[0190] Based on this, the above technical solution can be embodied as a software product. This software can be stored in a non-volatile storage medium. The instructions contained therein can cause computer devices (including personal computers, servers, or network devices, etc.) to execute the methods described in the embodiments of this invention.

[0191] Therefore, the present invention also provides a readable storage medium storing program instructions, which, when read and executed by a computing device, cause the computing device to perform a data-driven optimization method for site selection of an industrialized construction company; and a computing device comprising: at least one processor and a memory storing program instructions; wherein, when the program instructions are read and executed by the processor, the computing device performs a data-driven optimization method for site selection of an industrialized construction company.

[0192] Based on the above ranking method, this invention uses different calculation methods to empirically rank the site selection schemes of construction engineering industrialization companies (taking the top 10 schemes).

[0193] The ranking of the schemes calculated using the cost-benefit method is 10-14-6-11-5-4-12-2-16-7; the ranking of the schemes calculated using the subjective level method is 6-5-4-11-16-7-10-14-12-2; the ranking of the schemes calculated using the objective entropy method is 6-5-12-11-4-14-16-7-10-3; and the ranking of the schemes calculated using a combination of subjective and objective weighting and the ideal approximation method is 17-6-16-5-4-7-8-18-3-12.

[0194] Appendix Tables 4 to 8 are sorting tables of various calculation schemes calculated by different calculation methods in the embodiments of the present invention.

[0195] Table 4 Ranking of Cost-Benefit Calculation Schemes

[0196]

[0197] Table 5 Ranking of Calculation Schemes Using Subjective Hierarchy Method

[0198]

[0199] Table 6 Ranking of Objective Entropy Weight Method Calculation Schemes

[0200]

[0201] Table 7. Ranking of Schemes Calculated by Combining Subjective and Objective Weights with the Ideal Approximation Method

[0202]

[0203] Table 8. Summary of the ranking of calculation schemes using different calculation methods

[0204]

[0205] By comparing the rankings of schemes calculated using different methods, we can see that:

[0206] 1. Comparing the ranking of schemes calculated by the cost-benefit method with the ranking of schemes calculated by the subjective hierarchy method, the top 10 schemes are exactly the same, and the rankings of schemes 6, 5, 4, and 11 are relatively concentrated and consistent, reflecting the subjectivity of leaders or experts.

[0207] 2. The ranking of the top 10 schemes calculated by the cost-benefit method differs significantly from that calculated by the objective entropy method. This indicates that the objective scheme ranking is not only based on cost and benefit considerations, but also fully reflects the objectivity of the scheme data.

[0208] 3. The ranking of schemes calculated by the cost-benefit method is completely different from that calculated by combining subjective and objective comprehensive weights with the ideal approximation method, indicating that cost-benefit is not the sole basis for the latter ranking.

[0209] 4. Comparing the ranking of schemes calculated by the subjective hierarchy method with that calculated by the objective entropy value method, the two rankings are significantly different, indicating that the schemes reflect the different advantages of subjective weight and objective weight, and that they play their respective roles.

[0210] 5. The ranking of schemes calculated by combining subjective and objective comprehensive weights with the ideal approximation method shows that the scheme ranking integrates the advantages of the subjective hierarchy method and the objective entropy value method. It not only adopts the intentions of leaders and the opinions of experts, but also effectively weakens the subjective influence of leaders' personal intentions and experts' preferences on the evaluation indicators, increases the influence of objective facts on the ranking, reduces the weight of subjective decision-making, and increases the weight of the objective needs of the enterprise's own development and real-time situation. It achieves a balance between data-driven insights and expert experience, reduces the bias of single methods, and improves the credibility and acceptability of evaluation or ranking results, providing a rigorous reference for scientific site selection decisions.

[0211] The site selection optimization method provided by this invention starts from the perspective of ranking alternative schemes. It uses existing and predicted data, and assigns subjective and objective weights to the evaluation indicators through the analytic hierarchy process and the entropy method, respectively. Then, it combines the two weight results using the multiplication synthesis method to calculate the comprehensive weight. Finally, it uses the ideal approximation method to rank the schemes, so as to make an accurate and rapid response for the site selection of construction engineering industrialization companies and maximize the benefits of site selection.

[0212] It should be noted that, for the sake of convenience and brevity, the evaluation function is described in terms of modules. In practical applications, different individual modules can be used as needed; that is, each functional module can be used independently or in combination to achieve all or part of the above functions.

[0213] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A data-driven method for optimizing site selection in industrialized construction engineering companies, characterized in that, include: Step S1: Select and determine the main factors influencing the site selection of the construction industrialization company; Step S2: Use the 1-9 scale method to determine the importance of each major influencing factor to the target, and use the analytic hierarchy process (AHP) to calculate the subjective weight of each major influencing factor; Step S3: Based on the relevant existing data, calculate the objective weight of each major influencing factor using the entropy method, and combine it with Step S2 to determine the combined subjective and objective weight of each major influencing factor; Step S4: Based on a combination of subjective and objective weighting and the ideal approximation method, the site selection schemes are comprehensively evaluated and ranked, and the best site is selected for the construction engineering industrialization company.

2. The data-driven site selection optimization method for industrialized construction companies according to claim 1, characterized in that, Step S2 includes: constructing a site selection evaluation index system. First, construct an evaluation index system: build a framework including the target layer, criterion layer, index layer, and solution layer; Target layer: refers to the site selection of construction engineering industrialization companies, which is the highest level of the evaluation index system; Criterion layer: refers to the factors that influence the decision-making process for the target, or the intermediate links in achieving the target; it is the intermediate layer of the evaluation indicator system. Indicator layer: refers to the sub-factors that influence the decision-making process for the target, or the sub-steps in the intermediate steps to achieve the target. It is the intermediate layer of the evaluation indicator system. Solution layer: refers to the specific options and measures that can be selected, and is the lowest level of the evaluation indicator system.

3. The data-driven site selection optimization method for industrialized construction companies according to claim 2, characterized in that, Step S2 includes: using the analytic hierarchy process (AHP) to determine subjective weights, including: S2-1, Determine the judgment matrix of the criterion layer. For the target layer, pairwise comparisons are performed on each criterion in the criterion layer. The 1-9 scale method is used to determine their importance to the target, and the comparison results are filled into the judgment matrix. middle; ,in Represents element Relative to element The degree of importance; ； S2-2, Calculate the eigenvectors of the judgment matrix. First, normalize the eigenvectors: normalize each component of the eigenvectors according to the column vectors so that the sum of the components is 1; Then, sum the row vectors and normalize them again to obtain the eigenvectors. ; because It is an approximate eigenvector. Approximate to , It is a judgment matrix eigenvalues, It is the largest eigenvalue; For the judgment matrix eigenvectors , largest eigenvalue Satisfy the equation Therefore, each component satisfies ; The formula for calculating the largest eigenvalue is: by averaging over all components, a eigenvalue is obtained. Estimate ; S2-3, Calculate the consistency ratio ,include: S2-3-1, Calculate the consistency index The formula for the consistency index is: ; in, It determines the order of the matrix; S2-3-2, Finding the random consistency index The random consistency index is an average consistency index calculated from a large number of random matrices. S2-3-3, Calculate the consistency ratio The formula for the consistency ratio is: ,if If the consistency is satisfactory, the judgment matrix has satisfactory consistency and passes the consistency test; otherwise, the judgment matrix needs to be adjusted until the consistency requirement is met. S2-4, Determine the judgment matrix for the indicator layer. The hierarchical order of the indicator layers is calculated according to the above steps. ; but The overall inter-layer sorting is as follows: ; S2-5, Consistency check of the overall hierarchical order: set up layer For the upper level Layer factors The hierarchical single-order consistency index is The random consistency index is The consistency ratio of the overall hierarchical ranking is: ， when At that time, the overall hierarchical ranking is considered to have passed the consistency test; S2-6, Determine the subjective weights of each influencing factor based on the overall hierarchical ranking of the indicator layers: ； , j is the j-th influencing factor in criterion i, and n is the total number of influencing factors.

4. The data-driven site selection optimization method for industrialized construction companies according to claim 1, characterized in that, Step S3 includes: Existing data includes: construction cost indicators, operating costs, economic benefits, social benefits, and sustainable development capabilities.

5. The data-driven site selection optimization method for industrialized construction companies according to claim 4, characterized in that, Step S3 includes: S3-1, Constructing the original matrix: Assuming there are m main influencing factors and n alternative solutions, construct the original matrix using the values ​​of each alternative solution for each main influencing factor: , in, It is the value of the i-th alternative in the j-th main influencing factor; S3-2, Normalization: Normalize the original matrix: For positive primary influencing factors (such as returns), then ; For negative primary influencing factors (such as cost), then ; The standardized decision matrix is ​​obtained. ; S3-3, For the normalized data, calculate the proportion of each alternative's value under each major influencing factor to the sum of the values ​​of all alternatives under that major influencing factor. This proportion reflects the relative contribution of each alternative to that major influencing factor: ， in, It is the first The first scheme is in the The values ​​under the main influencing factors, It is the number of solutions. It is the number of major influencing factors; S3-4, Calculate the entropy value of each major influencing factor: Calculate the first Entropy values ​​of the main influencing factors: ； S3-5, Calculate the difference coefficient of entropy values: Difference coefficient (redundancy) ; S3-6, Calculate the objective weights of each major influencing factor: ， 。 6. The data-driven site selection optimization method for industrialized construction companies according to claim 5, characterized in that, Step S3 includes: calculating the overall weight using a combination of subjective and objective weighting methods. ， The result is .

7. The data-driven site selection optimization method for industrialized construction companies according to claim 5, characterized in that, Step S4 includes: S4-1, Construct a decision matrix: where each row represents an alternative solution, each column represents a major influencing factor, and the elements in the matrix represent the values ​​of each alternative solution on different major influencing factors; S4-2, normalize the original data in the decision matrix to eliminate the incomparability between evaluation indicators of different dimensions; S4-3, Calculate the weighted decision matrix: Multiply each column of the normalized decision matrix by the combined subjective and objective weights of the corresponding main influencing factors to obtain the weighted normalized decision matrix: ； S4-4, Calculate the positive and negative ideal solutions: Determine the positive and negative ideal solutions from the weighted normalized decision matrix; for benefit-type indicators, the positive ideal solution takes the maximum value for the indicator, while the negative ideal solution takes the minimum value; for cost-type indicators, the opposite is true, the positive ideal solution takes the minimum value, while the negative ideal solution takes the maximum value. Calculating the positive ideal solution: The positive ideal solution is obtained by weighted normalized decision matrix. The maximum element of each column vector (the minimum element of the negative index column) constitutes the structure. ; Calculating the negative ideal solution: The negative ideal solution is obtained by weighted normalized decision matrix. It is composed of the smallest element of each column vector (the largest element of the negative index column). ; S4-5, Calculate the Euclidean distance: For each alternative, calculate its Euclidean distance to the ideal solution and the negative ideal solution. Calculate the Euclidean distance between each alternative solution and the positive ideal solution. ; Calculate the Euclidean distance between each alternative solution and the negative ideal solution. ; S4-6, Calculate the relative proximity coefficient for each alternative: ； S4-7, Scheme Ranking: Sort the candidate schemes from largest to smallest according to their proximity coefficient. The closer the coefficient is to 1, the closer the alternative is to the optimal level; conversely, the closer the coefficient is to 0, the closer the alternative is to the worst level; the alternative with the highest relative proximity coefficient is considered the best choice.

8. A data-driven site selection optimization device for industrialized construction engineering companies, characterized in that, include: The influencing factors determination module is used to select and determine the main influencing factors affecting the site selection of construction engineering industrialization companies; The subjective weight calculation module is used to determine the importance of each major influencing factor to the target using the 1-9 scale method, and to calculate the subjective weight of each major influencing factor using the analytic hierarchy process. The subjective and objective comprehensive weight calculation module is used to calculate the objective weight of each major influencing factor based on relevant existing data using the entropy method, and to determine the subjective and objective comprehensive weight of each major influencing factor in conjunction with step S2. The ranking module is used to comprehensively evaluate and rank the site selection schemes based on a combination of subjective and objective weights and the ideal approximation method, and select the best site for the construction engineering industrialization company.

9. A computing device, characterized in that, include: At least one processor and a memory storing program instructions; When the program instructions are read and executed by the processor, the computing device performs the data-driven site selection optimization method for industrialized construction companies as described in any one of claims 1-7.

10. A readable storage medium storing program instructions, characterized in that, When the program instructions are read and executed by the computing device, the computing device performs the data-driven site selection optimization method for industrialized construction companies as described in any one of claims 1-7.

Images