Green residential building design scheme feasibility evaluation method based on AI

Abstract

The invention discloses a green residential building design scheme feasibility evaluation method based on AI, and relates to the field of data evaluation, and the method comprises the following steps: 1, pre-defining one or more design targets of each region in a current to-be-designed residential target, and setting a value range allowed to be received for each design target; 2, obtaining a to-be-evaluated design scheme, and extracting corresponding data from the design scheme based on a design target; 3, comparing the extracted data with a preset numerical range, defining the design targets in the preset range as hit targets, and defining the design targets not in the preset range as missed targets; automatic analysis of the design scheme is realized, low efficiency and subjective deviation of traditional manual evaluation are avoided, a designer can quickly identify problems by utilizing a preset numerical value range and gain and benefit reduction trend quantification, human errors are reduced, evaluation precision and efficiency are improved, and thus the design period is shortened and the trial and error cost is reduced.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of data evaluation, and particularly relates to an AI-based green residential building design scheme feasibility evaluation method. BACKGROUND

[0002] In response to climate change, the global building industry is rapidly transforming towards low-carbon buildings and sustainable development. Governments have introduced stricter building energy efficiency regulations and green building certification standards, making green residences popular. Green building design is no longer a simple structure plus appearance, but needs to consider multiple disciplines such as building physics, energy systems, material science, environmental meteorology and human comfort.

[0003] AI technology, especially in the fields of machine learning, computer vision and predictive optimization, provides a new technical means for dealing with the above complex and multi-variable systemic problems. AI can learn rules from massive data, perform rapid iteration simulation and intelligent optimization, thereby greatly improving evaluation efficiency and accuracy.

[0004] Traditional methods usually evaluate and analyze each performance indicator independently, and lack a unified platform to evaluate the comprehensive impact of a design decision on multiple objectives. For example, reducing the window area to reduce energy consumption may severely compromise lighting and visual comfort. However, this trade-off relationship is difficult to quantify and evaluate, and the evaluation process is fragmented, lacking a quantitative and unified scoring system to objectively measure the overall green completion of the design scheme. SUMMARY

[0005] (I) Technical problems solved In view of the above-mentioned shortcomings of the prior art, the application provides an AI-based green residential building design scheme feasibility evaluation method, which can effectively solve the problems of the prior art.

[0006] (II) Technical scheme To achieve the above object, the application is implemented by the following technical scheme: The application discloses an AI-based green residential building design scheme feasibility evaluation method, comprising the following steps: Step 1: predefine one or more design targets for each region in the current target residential design, and set a value range allowed to be received for each design target; Step 2: obtain the design scheme to be evaluated, and extract corresponding data from the design scheme based on the design target; Step 3: compare the extracted data with the preset value range, define the design target within the preset range as a hit target, and define the design target not within the preset range as a miss target; Step 4: Evaluate the gain trend of the hit target, which is used to characterize the degree of positive contribution of the hit target to the overall design scheme; evaluate the loss trend of the miss target, which is calculated based on the deviation of the actual value of the miss target from the preset range boundary; Step 5: Determine the influence coefficient of each region in the overall residence based on the functional attributes of the residential regions and the inter-regional correlation; Step 6: Build a regional evaluation model, input the gain trend or loss trend data of a certain region, combine the influence coefficient of the region, and the distribution state of the hit target and the miss target in the current residential region, and the regional evaluation model analyzes the design completion degree of the region; Step 7: Based on the design completion degree of all regions, the design completion degree of the overall design scheme is calculated by weighting with a preset weight, and compared with the preset standard design completion degree. If it reaches or exceeds the preset standard, it is determined to be qualified.

[0007] Further, the design targets in step 1 include environmental parameters, spatial layout parameters, and equipment energy efficiency parameters; the numerical range of the design target is set based on green building evaluation standards, regional climate characteristics, and user individualized needs.

[0008] Further, the environmental parameters include daylighting coefficient, ventilation frequency, indoor thermal comfort; the spatial layout parameters include space area, space net height, static and dynamic partition index; the equipment energy efficiency parameters include lamp light efficiency, air conditioner energy efficiency ratio, and hot water equipment thermal efficiency.

[0009] Further, the calculation process of the gain trend in step 4 is: obtain the degree that the actual value of the target exceeds the lower limit value of its preset numerical range, and the ratio between the exceeding degree and the overall width of the target preset numerical range is taken as the basic gain value of the target. Multiply the basic gain value by the weight factor set for the target in advance to obtain the gain trend data of the hit target.

[0010] Further, the calculation process of the loss trend in step 4 is: read the nearest preset range boundary value deviated from its actual value, calculate the absolute distance of the actual value deviated from the boundary value, and the ratio between the absolute distance and the overall width of the target preset numerical range is taken as the basic deviation degree of the target. Multiply the basic deviation degree by the weight factor set for the target in advance, and the result is the loss trend value of the miss target.

[0011] Further, the calculation logic of the influence coefficient in step 5 includes: Step 51: Define the inter-regional interaction intensity value between each two regions in the residence; Step 52: Establish a regional influence transmission network, taking each region as a node and the interaction intensity value as an edge weight, and calculate the reachability index of each region affecting all other regions through direct or indirect paths, which is proportional to the sum of the reciprocal of the shortest paths from the target region to other nodes in the network, and the functional positioning of the region in the residence is determined; Step 53: Multiply the reachability index by the base importance coefficient to obtain the initial influence value of each region, extract the data dispersion characteristics of the design target of each region, increase the weight of the region whose target value fluctuation amplitude is greater than the threshold value, aggregate the weights of the region group whose correlation coefficient between targets is greater than the preset threshold value, and scale the corrected value according to the total value proportion of the whole house region, and finally output the global influence coefficient of each region.

[0012] Further, the evaluation process of the interaction intensity value between regions in step 51 is as follows: Calculate the spatial distance weight between directly connected regions, taking the length of adjacent walls or the area of passageways as the measurement basis; Statistical analysis of the sharing rate of device pipeline systems between regions, with the proportion of the number of cross-region connections of ventilation pipes and circuit pipelines to the total pipeline as the value; Evaluate the behavior flow frequency generated by functional complementarity based on user activity trajectory simulation data to determine the average daily human flow interaction frequency between regions; Normalize and weight the cumulative spatial distance weight, device pipeline system sharing rate, and behavior flow frequency to obtain the interaction intensity value between regions.

[0013] Further, the inter-regional association relationship in step 5 includes functional flow line association, spatial line of sight association, energy supply and demand association, and physical environment interference association.

[0014] Further, the region evaluation model in step 6 is a regression model based on a neural network algorithm, which includes region gain, loss data, influence coefficient, hit or miss target distribution proportion as input, and region design completion degree as output, and its calculation formula is: ; In the formula, represents the design completion degree of the current region , represents the total number of hit targets in the region , represents the gain influence parameter of the first hit target in the region , represents the global influence coefficient of the region , represents the total number of miss targets in the region , represents the gain influence parameter of the first hit target in the region . ​​​​​The first missed target reduction trend index, The reduction correction factor of the representative area .

[0015] Further, the modification rule of the design completion degree of the overall design scheme in step 7 comprises: When the influence coefficient of a certain area on the overall residence exceeds the preset first threshold value, a reinforcement penalty weight is applied to the reduction trend data generated by the missed target in the area; If there are multiple consecutive achieved hit targets in a certain area, a stepwise enhancement process is started for the gain influence data of the subsequent hit targets in the area; When the reduction trend data of a certain area continues to deteriorate and the influence coefficient is higher than the preset second threshold value, the gain overflow compensation function of the adjacent area is forcibly activated, and the surplus of the adjacent area exceeding the baseline gain value is proportionally transferred to the deteriorating area.

[0016] (Three) beneficial effects Compared with the known prior art, the technical scheme provided by the present application has the following beneficial effects: 1. Through AI-driven quantitative evaluation, automatic analysis of the design scheme is realized, avoiding the inefficiency and subjective bias of traditional manual evaluation. By using preset numerical range and gain and reduction trend quantization, designers can quickly identify problems, reduce human errors, improve evaluation accuracy and efficiency, thereby shortening the design cycle and reducing trial and error costs, while enhancing the scientificity and systematicness of residential space optimization.

[0017] 2. By introducing the area influence coefficient and the evaluation model, the correlation between the areas of the residence can be considered comprehensively, and the strategy can be dynamically adjusted, providing more scientific overall design completion degree analysis to help designers optimize the scheme and avoid global failure caused by local optimization, thereby improving the overall reliability of green building design.

[0018] 3. By integrating environmental parameters and equipment energy efficiency parameter settings with the area evaluation model, qualification determination and optimization compensation functions are performed in the early design stage of the residence, thereby reducing building energy consumption, improving user comfort, and reducing later modification expenses, and providing a fine quantitative evaluation mechanism. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the technical scheme in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creating any creative labor.

[0020] Figure 1 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0022] The present invention will be further described below with reference to embodiments.

[0023] This embodiment presents an AI-based feasibility assessment method for green residential building design schemes, such as... Figure 1 As shown, it includes the following steps: Step 1: Predefine one or more design objectives for each area of ​​the residential building to be designed, and set an acceptable numerical range for each design objective; the areas include, but are not limited to, bedrooms, living rooms, kitchens, bathrooms, and balconies; the design objectives include environmental parameters, spatial layout parameters, and equipment energy efficiency parameters; the environmental parameters include daylight factor, ventilation rate, and indoor thermal comfort; the spatial layout parameters include space area, net space height, and active / quiet zoning index; the equipment energy efficiency parameters include luminous efficacy, air conditioning energy efficiency ratio, and hot water equipment thermal efficiency; the numerical range of the design objectives is set based on at least one of the following: green building evaluation standards, regional climate characteristics, and user personalized needs.

[0024] Step 2: Obtain the design scheme to be evaluated, and extract the corresponding data from the design scheme based on the design objectives.

[0025] Step 3: Compare the extracted data with the preset value range, define the design targets within the preset range as hit targets, and define the design targets outside the preset range as miss targets.

[0026] Step 4: Evaluate the gain trend of hit targets, which characterizes the positive contribution of hit targets to the overall design scheme; evaluate the loss trend of missed targets, which is calculated based on the deviation of the actual value of the missed targets from the preset range boundary. The gain trend is calculated as follows: The degree to which the actual value of the target exceeds the lower limit of its preset value range is obtained. The ratio between this excess degree and the overall width of the target's preset value range is used as the target's base gain value. This base gain value is multiplied by a weighting factor pre-set for the target to obtain the gain trend data of the hit target. The larger the gain trend value, the further it deviates from the standard, and the higher its negative impact.

[0027] Step 5: Based on the functional attributes of residential areas and the inter-area relationships, determine the influence coefficient of each area in the overall residential area; the inter-area relationships include functional circulation relationships, spatial sight relationship relationships, energy supply and demand relationship relationships, and physical environmental interference relationship relationships.

[0028] Step 6: Construct a regional assessment model. Input the gain trend or loss trend data of a certain region, combine the influence coefficient of the region, and the distribution of hit and miss targets in the current residential area. The regional assessment model analyzes and determines the design completion level of the region.

[0029] Step 7: Based on the design completion rate of all areas, calculate the overall design completion rate of the design scheme using preset weights, and compare it with the preset standard design completion rate. If it reaches or exceeds the preset standard, it is judged as qualified. The rules for revising the overall design completion status include: When the impact coefficient of a certain area on the overall housing exceeds a preset first threshold, an enhanced penalty weight is applied to the negative trend data of the non-targets in that area. The value range of the enhanced penalty weight is set to 1.2 to 1.5 times the base weight. If multiple consecutive targets are hit in a certain area, a tiered enhancement process is initiated for the gain impact data of subsequent targets hit in that area. When the number of consecutive targets hit reaches the first-level threshold, the gain impact data is increased by 10%; when the number of consecutive targets hit exceeds the second-level threshold, the increase in the gain impact data expands to 25%. When the loss trend data of a certain region continues to deteriorate and the impact coefficient is higher than the preset second threshold, the gain overflow compensation function of the adjacent region is forcibly activated, and the surplus of the adjacent region exceeding the benchmark gain value is transferred to the deteriorating region proportionally. By introducing a high-impact region penalty mechanism, the impact of missing targets can be dynamically adjusted, thereby more accurately reflecting the actual performance of the design scheme. The continuous hit gain accumulation mechanism enables multiple target achievement to produce more positive effects. The cross-regional compensation trigger mechanism realizes the collaborative optimization between regions, enhancing the flexibility and adaptability of the overall design.

[0030] At other levels, this embodiment provides a calculation logic for the influence coefficient, including: Step 51: Define the inter-area interaction intensity value between every two areas in the residence; the evaluation process for the inter-area interaction intensity value is as follows: calculate the spatial distance weight of direct connections between areas, using the length of adjacent walls or the area of ​​passageways as the measurement benchmark; calculate the sharing rate of equipment and pipeline systems between areas, assigning values ​​based on the proportion of ventilation ducts and electrical wiring connections across areas to the total pipelines; evaluate the frequency of behavioral flow lines generated by functional complementarity, determining the average daily number of human interactions between areas based on user activity trajectory simulation data; normalize and weight the spatial distance weight, the sharing rate of equipment and pipeline systems, and the frequency of behavioral flow lines to obtain the inter-area interaction intensity value; Step 52: Establish a regional influence transmission network, with each region as a node and the interaction strength value as the edge weight. Calculate the accessibility index of each region to all other regions through direct or indirect paths. This index is proportional to the sum of the reciprocals of the shortest paths from the target region to other nodes in the network. Based on the functional positioning of the region in the residential area, preset the basic importance coefficient: core functional area > transition area > service area. Step 53: Multiply the accessibility index by the basic importance coefficient to obtain the initial impact value of each region, extract the data dispersion characteristics of the design target of each region, increase the weight of regions whose target value fluctuation is greater than the threshold, perform weight aggregation on the region group whose target correlation coefficient is greater than the preset threshold, scale the corrected value according to the total value of the whole residential area, and finally output the global impact coefficient of each region.

[0031] By comprehensively considering the interaction intensity of multiple dimensions such as spatial distance between areas, pipeline sharing, and user behavior flow, a more accurate regional impact coefficient model was constructed. In addition, the impact value was dynamically corrected by using the entropy weight method, which can effectively reflect the fluctuation characteristics of the design target data, thereby improving the flexibility and accuracy of the assessment, enhancing the system's understanding of the importance and complementarity of regional functions, and realizing optimization based on actual usage, thus improving the feasibility assessment effect of green housing design.

[0032] This embodiment provides a regional assessment model, which is a regression model built based on a neural network algorithm. Its inputs include regional gain, loss data, influence coefficient, and the proportion of targets hit or missed. The output is the regional design completion rate, calculated using the following formula: ; In the formula, Represents the current region Design completion level Representative area The total number of targets hit. Representative area No. The gain effect parameter for each hit target. Representative area The global impact coefficient, Representative area The total number of targets missed. Representative area No. The loss trend index of missing targets, Representative area The system incorporates a loss correction factor; by dynamically predefining design objectives and their numerical ranges, it can accurately identify hit and miss objectives, thereby quantifying the gains and losses, improving the accuracy of the assessment, and introducing a comprehensive calculation of regional influence coefficients and design completion. This not only considers the performance of a single objective but also reflects the interaction between regions, enhancing the scientific nature of the overall assessment. The dynamic correction mechanism and cross-regional compensation algorithm ensure that the system can be flexibly adjusted in different design schemes, optimizing resource allocation and achieving more efficient green building design.

[0033] This is a regression model built based on a neural network algorithm. Its inputs include regional gain, loss data, influence coefficient, and the distribution ratio of hit or miss targets. The output is the regional design completion rate, and its calculation formula is as follows: ; In the formula, Represents the current region Design completion level Representative area The total number of targets hit. Representative area No. The gain effect parameter for each hit target. Representative area The global impact coefficient, Representative area The total number of targets missed. Representative area No. The loss trend index of missing targets, Representative area The system incorporates a loss correction factor; by dynamically predefining design objectives and their numerical ranges, it can accurately identify hit and miss objectives, thereby quantifying the gains and losses, improving the accuracy of the assessment, and introducing a comprehensive calculation of regional influence coefficients and design completion. This not only considers the performance of a single objective but also reflects the interaction between regions, enhancing the scientific nature of the overall assessment. The dynamic correction mechanism and cross-regional compensation algorithm ensure that the system can be flexibly adjusted in different design schemes, optimizing resource allocation and achieving more efficient green building design.

[0034] In summary, this invention provides a clear basis for AI evaluation by refining and pre-defining the specific design objectives and numerical ranges of each region, thereby achieving dynamic and interconnected evaluation processes. It not only identifies whether the scheme data hits the preset objectives, but also deeply analyzes the actual gain intensity of the hit objectives and the loss trend of the unhit objectives, and evaluates the local and global impacts in combination with the functional weights of each region.

[0035] This invention achieves comprehensive and quantifiable evaluation results. The constructed regional evaluation model can integrate multiple factors such as gain, loss, regional importance, and target distribution to output a completion score for a specific region. Finally, the scores are weighted and summarized into an overall design completion value that can be intuitively compared, improving the efficiency of design optimization. This method can quickly identify the shortcomings of the solution, clarify the optimization direction, provide designers with data-driven decision support, effectively shorten the design cycle, and reduce rework costs.

[0036] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A feasibility assessment method for green residential building design schemes based on AI, characterized in that, Includes the following steps: Step 1: Predefine one or more design targets for each area in the current residential target to be designed, and set the allowed numerical range for each design target; Step 2: Obtain the design scheme to be evaluated, and extract the corresponding data from the design scheme based on the design objectives; Step 3: Compare the extracted data with the preset value range, define the design targets within the preset range as hit targets, and define the design targets outside the preset range as miss targets; Step 4: Evaluate the gain trend of the hit target. The gain trend is used to characterize the positive contribution of the hit target to the overall design scheme. Assess the loss trend of missed targets, which is calculated based on the degree of deviation between the actual value of the missed target and the preset range boundary; Step 5: Based on the functional attributes of residential areas and the inter-area relationships, determine the influence coefficient of each area in the overall residential area; Step 6: Construct a regional assessment model. Input the gain trend or loss trend data of a certain region, combine the influence coefficient of the region, and the distribution of hit and miss targets in the current residential area. The regional assessment model analyzes and obtains the design completion degree of the region. Step 7: Based on the design completion rate of all areas, calculate the overall design completion rate of the design scheme using preset weights, and compare it with the preset standard design completion rate. If it reaches or exceeds the preset standard, it is judged as qualified.

2. The feasibility assessment method for an AI-based green residential building design scheme according to claim 1, characterized in that, The design objectives in step 1 include environmental parameters, spatial layout parameters, and equipment energy efficiency parameters; the numerical range of the design objectives is set based on green building evaluation standards, regional climate characteristics, and user-specific needs.

3. The feasibility assessment method for an AI-based green residential building design scheme according to claim 2, characterized in that, The environmental parameters include daylight factor, ventilation rate, and indoor thermal comfort; the spatial layout parameters include space area, net height, and dynamic / static zoning index; and the equipment energy efficiency parameters include luminous efficacy, air conditioning energy efficiency ratio, and hot water equipment thermal efficiency.

4. The feasibility assessment method for an AI-based green residential building design scheme according to claim 1, characterized in that, The calculation process of the gain trend in step 4 is as follows: obtain the degree to which the actual value of the target exceeds the lower limit of its preset value range, and use the ratio between the degree of exceeding and the overall width of the preset value range of the target as the basic gain value of the target. Multiply the basic gain value by the weighting factor set in advance for the target to obtain the gain trend data of the hit target.

5. The feasibility assessment method for an AI-based green residential building design scheme according to claim 1, characterized in that, The calculation process of the loss trend in step 4 is as follows: read the nearest preset range boundary value that the actual value deviates from, calculate the absolute distance of the actual value from the boundary value, and use the ratio between the absolute distance and the overall width of the target preset value range as the target basic deviation degree. Multiplying the baseline deviation by the pre-set weighting factor for the target yields the loss trend value for the missed target.

6. The feasibility assessment method for an AI-based green residential building design scheme according to claim 1, characterized in that, The calculation logic for the influence coefficient in step 5 includes: Step 51: Define the inter-zone interaction strength value between every two zones in the residence; Step 52: Establish a regional influence transmission network, with each region as a node and the interaction strength value as the edge weight. Calculate the accessibility index of each region to all other regions through direct or indirect paths. This index is proportional to the sum of the reciprocals of the shortest paths from the target region to other nodes in the network, based on the functional positioning of the region in the residential area. Step 53: Multiply the accessibility index by the basic importance coefficient to obtain the initial impact value of each region, extract the data dispersion characteristics of the design target of each region, increase the weight of regions whose target value fluctuation is greater than the threshold, perform weight aggregation on the region group whose target correlation coefficient is greater than the preset threshold, scale the corrected value according to the total value of the whole residential area, and finally output the global impact coefficient of each region.

7. The feasibility assessment method for an AI-based green residential building design scheme according to claim 6, characterized in that, The evaluation process for the inter-regional interaction strength value in step 51 is as follows: Calculate the spatial distance weight of direct connections between regions, using the length of adjacent walls or the area of ​​passageways as the measurement benchmark; The sharing rate of equipment and pipeline systems between statistical areas is assigned as the proportion of the number of cross-regional connections of ventilation ducts and electrical pipelines to the total number of pipelines. Assess the frequency of behavioral flows resulting from functional complementarity, and determine the average daily number of human interactions between areas based on user activity trajectory simulation data; The spatial distance weight, the sharing rate of the equipment pipeline system, and the frequency of behavioral flow are normalized and weighted to obtain the interaction intensity value between regions.

8. The feasibility assessment method for an AI-based green residential building design scheme according to claim 1, characterized in that, The inter-regional relationships in step 5 include functional flow line relationships, spatial line-of-sight relationships, energy supply and demand relationships, and physical environmental interference relationships.

9. The feasibility assessment method for an AI-based green residential building design scheme according to claim 1, characterized in that, The regional assessment model in step 6 is a regression model built based on a neural network algorithm. Its inputs include regional gain, loss data, influence coefficient, and the distribution ratio of hit or miss targets. The output is the regional design completion degree, and its calculation formula is as follows: ； In the formula, Represents the current region Design completion level Representative area The total number of targets hit. Representative area No. The gain effect parameter for each hit target. Representative area The global impact coefficient, Representative area The total number of targets missed. Representative area No. The loss trend index of missing targets, Representative area The loss correction factor.

10. The feasibility assessment method for an AI-based green residential building design scheme according to claim 1, characterized in that, The rules for correcting the overall design completion status in step 7 include: When the impact coefficient of a certain area on the overall housing exceeds the preset first threshold, a stronger penalty weight is applied to the negative trend data of the non-targets generated in that area. If a region has multiple consecutive hit targets, then a step-wise enhancement process will be initiated for the gain impact data of subsequent hit targets in that region. When the loss trend data of a certain region continues to deteriorate and the impact coefficient is higher than the preset second threshold, the gain overflow compensation function of the adjacent region is forcibly activated, and the surplus of the adjacent region exceeding the benchmark gain value is transferred to the deteriorating region proportionally.

Images