A Real-Time Optimization System and Method for Architectural Design Based on Multi-Source Heterogeneous Data

Abstract

This invention discloses a real-time optimization processing system and method for architectural design based on multi-source heterogeneous data, belonging to the field of architectural design optimization technology. The real-time optimization processing system for architectural design based on multi-source heterogeneous data includes: a data acquisition module, a design optimization module, a candidate acquisition module, an execution analysis module, a comprehensive evaluation module, a feedback optimization module, and a completed repair module. This invention constructs an architectural design optimization model based on convolutional neural networks, performs deep feature extraction on the multi-source heterogeneous data acquired during the construction phase of the building object, and combines structural templates and functional templates for matching and splicing, thereby generating an initial set of schemes that meet environmental conditions, regulatory requirements, and user functional needs, significantly improving the efficiency of architectural design generation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of architectural design optimization, and particularly relates to a real-time optimization processing system and method for architectural design based on multi-source heterogeneous data. BACKGROUND

[0002] With the increasing complexity and customization requirements of construction projects, architectural design is facing challenges from multiple aspects. On the one hand, during the construction phase of a building, a large amount of heterogeneous data needs to be considered comprehensively, such as environmental conditions, building codes, construction schedules, and user functional requirements. On the other hand, after a building is completed and put into use, its functional state and structural performance will deviate over time, and it is necessary to maintain its safety and applicability through dynamic adjustment and local repair.

[0003] Traditional architectural design processes mostly use methods based on empirical rules or static BIM parameter driving, which are usually one-time finalized and difficult to respond to real-time construction feedback or performance degradation in the operation phase.

[0004] Taking actual engineering scenarios as examples, in the process of urban renewal, complex site development, or large-scale building group construction, user demand changes often lead to spatial function reconstruction; the construction schedule deviates from the original plan; the local area has been constructed, but the overall scheme needs to be adjusted; the completed building has problems of structural aging, high energy consumption, or functional degradation. In the above scenarios, if there is a lack of real-time optimization mechanism and feedback loop based on multi-source data, it will lead to design lag, uncoordinated schemes, and even resource waste and functional failure. SUMMARY

[0005] The present application proposes a real-time optimization processing method and system for architectural design based on multi-source heterogeneous data, which constructs a design optimization model with a convolutional neural network as the core to address the above problems, and integrates it with functional evaluation, structural strength analysis, execution matching, real-time feedback mechanism, and repair paths in the completed stage, providing intelligent, flexible, and controllable optimization support for architectural projects throughout their life cycle.

[0006] The real-time optimization processing method for architectural design based on multi-source heterogeneous data comprises:

[0007] Obtaining multi-source heterogeneous data of the architectural object, including environmental data, code data, construction schedule data, and user demand data in the construction phase, or structural surveying data, state monitoring data, and renovation demand data in the completed phase;

[0008] For the architectural object in the construction phase, input its multi-source heterogeneous data into the design optimization model based on the convolutional neural network, and obtain an initial scheme set through template matching, template splicing, and functional evaluation;

[0009] Obtain the structural strength evaluation value of each initial scheme and the function evaluation value in the design optimization model, and obtain the frame evaluation value after weighted calculation, select the initial scheme with the top three frame evaluation values in the initial scheme set as the candidate scheme, and the candidate scheme also includes three initial schemes with the structural strength evaluation value exceeding the preset threshold and the top three function evaluation values;

[0010] Divide the building strength value of the candidate scheme into three grades, the building strength value is composed of the structural strength evaluation value and the material strength evaluation value, select the construction technology and structural material corresponding to the material strength evaluation value according to the building strength value of the three grades, and obtain the execution evaluation value of the building strength value of each grade of the candidate scheme under the corresponding construction technology and structural material;

[0011] According to the user preference, the building strength value, the function evaluation value and the execution evaluation value of the candidate scheme are weighted and calculated to obtain the comprehensive evaluation value, and the final building design scheme is obtained;

[0012] According to the construction progress data and the on-site abnormal feedback, the building design scheme is triggered to be reacquired to realize real-time optimization of the building design scheme.

[0013] As a preferred technical solution of the present application, the design optimization model based on the convolutional neural network comprises:

[0014] The input layer is used for receiving multi-source heterogeneous data, and the multi-source heterogeneous data is converted into a multi-source heterogeneous feature tensor after embedding coding and standardization processing;

[0015] The convolution matching layer is used for performing feature extraction and matching operation on the multi-source heterogeneous feature tensor through a convolution kernel, outputting each matched structure template and the matching feature map between the sub-regions of the structure template and the function templates, and being used for representing the structure template and the corresponding preselected function template set, wherein the convolution kernel is composed of pre-trained structure templates and function templates;

[0016] The feature splicing layer is used for combining and splicing each sub-region of the structure template and the corresponding preselected function template set according to the matching feature map, generating the to-be-evaluated schemes covering all the sub-regions one by one, and forming a to-be-evaluated scheme set, and inputting to the function evaluation layer;

[0017] The function evaluation layer is used for performing independent function evaluation, adjacent function evaluation and global function evaluation on each to-be-evaluated scheme respectively, and performing weighted calculation on the evaluation values of the three types of evaluation to obtain the function evaluation value;

[0018] The output layer is used for outputting the to-be-evaluated scheme with the function evaluation value meeting the preset threshold as the initial scheme set.

[0019] As a preferred technical solution of the present application, the function evaluation layer comprises:

[0020] independent function evaluation unit: for evaluating the function template of each sub-region in the to-be-evaluated scheme, obtaining the corresponding function evaluation value; combining the evaluation values of the same function to construct a function matrix, and comparing with the demand matrix extracted based on the multi-source heterogeneous data; when the value of the corresponding item in the function matrix is greater than or equal to the value of the corresponding item in the demand matrix, it is considered to meet the requirements; the proportion of the number of conforming items to the total number of valid items in the demand matrix is calculated to obtain the quantitative independent function evaluation value;

[0021] adjacent function evaluation unit: for evaluating the function configuration relationship between adjacent sub-regions in the to-be-evaluated scheme, judging the rationality thereof in space layout and function connection, and converting the evaluation result into an adjacent function evaluation value;

[0022] global function evaluation unit: for comprehensively evaluating the energy consumption efficiency, spatial layout coordination and function integrity of the entire to-be-evaluated scheme, and outputting a corresponding global function evaluation value.

[0023] As a preferred technical solution of the present application, the structure template is composed of a plurality of sub-regions with tolerance space range, and the preselected function template is matched and obtained within the tolerance space range of the corresponding sub-region based on the input multi-source heterogeneous feature tensor; in the matching process, the function template is screened according to the spatial position of the sub-region and the corresponding function preference weight; the function preference weight is initially set according to the spatial position of the sub-region, and can be adjusted according to user demand.

[0024] As a preferred technical solution of the present application, the structure template further comprises: in the process of triggering the reacquisition of the building design scheme, the sub-region that has completed construction is set to a fixed state, and the configuration of its corresponding structure and function template is kept unchanged; for the sub-region in the unconstructed state, the function template matching and screening is continued within its original tolerance space range, and the screening process is still based on its spatial position and function preference weight.

[0025] As a preferred technical solution of the present application, the acquisition of the framework evaluation value comprises:

[0026] structure strength evaluation of each initial scheme, obtaining its structure strength evaluation value, which is used to represent the strength performance coefficient of the scheme in structure arrangement, construction rationality and stress path; obtaining the function evaluation value of the initial scheme obtained in the design optimization model, which is used to measure the performance of the scheme in space use, function adaptation and overall performance; the structure strength evaluation value and the function evaluation value are weighted calculated according to the preset weight coefficient to obtain the framework evaluation value of the initial scheme; the weight coefficient is configured according to user preference or building purpose, and can be dynamically adjusted.

[0027] As a preferred technical solution of the present application, the division of the building strength value of the alternative scheme into three levels comprises:

[0028] Three building strength levels are preset, namely a high strength level, a medium strength level and a low strength level, each level corresponding to a fixed building strength target value; for each alternative scheme, at each strength level, the required material strength evaluation value is calculated by dividing the building strength target value of the scheme by the structure strength evaluation value, and under the premise of meeting the material strength evaluation value, the most optimal combination of the selected construction process and structural material is selected, the execution evaluation of each alternative scheme at different strength levels is carried out based on the selected combination of construction process and structural material, and the execution evaluation includes construction period estimation and resource utilization efficiency, and the evaluation process is based on historical data fitting.

[0029] As a preferred technical solution of the present application, the building design real-time optimization processing method based on multi-source heterogeneous data further comprises: for a building object in the construction stage, based on its structure surveying data and state monitoring data, obtaining a building design scheme and a corresponding building strength value, and inputting them into a design optimization model to obtain a corresponding function evaluation value; based on the function evaluation value and the building strength value, identifying a sub-region of the structure template of the building design scheme with function degradation or insufficient strength, and performing local function template re-matching and structure template splicing operations to generate a corresponding repair design scheme.

[0030] The building design real-time optimization processing system based on multi-source heterogeneous data comprises:

[0031] A data acquisition module is configured to acquire multi-source heterogeneous data of a building object;

[0032] A design optimization module is configured to input the multi-source heterogeneous data of the building object into a design optimization model constructed based on a convolutional neural network when the building object is in a construction stage, and output an initial scheme set;

[0033] An alternative acquisition module is configured to acquire structure strength evaluation values and frame evaluation values of each initial scheme, and filter alternative schemes in combination with function evaluation values;

[0034] An execution analysis module is configured to acquire execution evaluation values of building strength values of each level of the alternative scheme under corresponding construction processes and structural materials;

[0035] A comprehensive evaluation module is configured to acquire comprehensive evaluation values of the alternative scheme, and determine a final building design scheme;

[0036] A feedback optimization module is configured to trigger a reacquisition process of the building design scheme based on construction progress data or on-site abnormal feedback;

[0037] A building repair module is configured to generate a corresponding repair design scheme when the building object is in a building stage.

[0038] The present application has the following advantages:

[0039] The present application constructs a building design optimization model based on a convolutional neural network, extracts deep features from multi-source heterogeneous data obtained during the construction stage of the building object, and matches and splices the structure template and the function template, thereby generating an initial scheme set that meets the environmental conditions, specification requirements, and user function requirements, significantly improving the building design generation efficiency.

[0040] The present application further filters out alternative schemes that have both structural safety and functional adaptability by respectively evaluating the structural strength and function of each initial scheme and performing weighted calculation to obtain a framework evaluation value, effectively improving the scientificity and comprehensive performance of scheme selection.

[0041] The present application divides the building strength value into three preset levels, and on this basis, inversely deduces the material strength evaluation value required to meet each strength level, and selects the corresponding construction process and structural material under the premise of meeting the strength requirement, and realizes the selection of alternative schemes considering the execution combined with the execution evaluation result, improving the engineering practicability and resource allocation efficiency of the design result.

[0042] The present application introduces a real-time feedback mechanism, combines construction progress data and on-site abnormal information during construction, triggers the update of the building design scheme, keeps the constructed area unchanged, and only dynamically optimizes and adjusts the unconstructed area, ensuring the flexible response and continuous control of the building design scheme in the actual construction process.

[0043] The present application is suitable for state evaluation and local repair design in the building stage of the building object, obtains the function evaluation value and building strength value of the current building design scheme based on structure surveying data and state monitoring data, and then identifies the sub-area with function degradation or insufficient structural performance, and performs local reconstruction operation of the function template and the structure template, generates a repair design scheme, expands the application ability of the present application in the building operation and maintenance stage, and realizes the whole cycle intelligent optimization from design to operation. BRIEF DESCRIPTION OF DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only schematic drawings of the present application, and those skilled in the art can obtain other drawings according to the provided drawings without any creative effort;

[0045] Figure 1A structural schematic diagram of a building design real-time optimization processing system based on multi-source heterogeneous data used in an embodiment of the present application. DETAILED DESCRIPTION

[0046] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be described in further detail below with reference to the drawings. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0047] Embodiment 1, a building design real-time optimization processing method based on multi-source heterogeneous data, comprises the following steps:

[0048] Step S1: Obtain multi-source heterogeneous data of a building object, including environment data, specification data, construction progress data and user demand data in the construction stage, or structure surveying and mapping data, state monitoring data and reconstruction demand data in the completed stage;

[0049] The multi-source heterogeneous data refers to a data set derived from different collection subjects, having multiple data structures and attribute dimensions, and used to support the building design optimization process, specifically including but not limited to structured data, time series data, image data and expert rule set, and according to the different stages of the building object, it is divided into two categories of construction stage data and completed stage data:

[0050] The multi-source heterogeneous data in the construction stage includes:

[0051] Environment data: refers to natural and surrounding condition data affecting building site selection, structure design and energy consumption performance, used for structure template generation and building orientation and layout preliminary selection, and the sources include geological survey report, real-time or historical data interface of meteorological bureau, and environment monitoring point deployment device.

[0052] Example fields include: annual average wind speed, extreme wind load value (unit: m / s), peak ground acceleration, standard value of soil bearing capacity, and surrounding shelter distribution map (for sunshine analysis).

[0053] Specification data: refers to mandatory and recommended design standards applicable to the geographical location, building purpose and grade of the project, used for rule checking in function evaluation, constructing constraint boundaries of demand matrix, and the sources include building standard database (GB series), local implementation rules and industry guidelines.

[0054] Examples: evacuation distance requirement in building design fireproofing specification; minimum net height limit of kitchen and bathroom in residential building; energy saving index limit value of office building.

[0055] Construction progress data: refers to the progress information reflecting the current construction state of the building project, used for participating in the real-time adjustment mechanism of the scheme, marking the completed area when triggering the reacquisition of the design scheme, and the sources include BIM construction management system, field scheduling platform and construction log records.

[0056] Example fields include: sub-regional construction state (completed / in progress / not started); actual construction start and end time; abnormal working condition log (such as material delay, component deviation).

[0057] User demand data: refers to the function, layout and performance preferences set by the building user or investor, used to convert into preference weight input matched with function templates, and construct the demand matrix reference source of the function evaluation layer, and the sources are the function demand questionnaire filled by the user, the design task book or the results of expert interview.

[0058] Example fields include: function space list (e.g. 4 offices + 1 conference room + public tea room); function space area range; space priority and relative position requirement (such as conference room needs to be close to the lobby); weighted preference for lighting, privacy, expandability (numerical weight).

[0059] Multi-source heterogeneous data in the construction stage includes:

[0060] Structure mapping data: structure state information obtained based on the actual measurement of the completed building, used for calculating the structure strength part of the building strength value, as the input basis for the repair scheme generation, evaluating the component integrity and consistency, and the sources are three-dimensional laser scanning, structure detection instruments (such as radar, infrared temperature measurement), BIM completion model comparison analysis.

[0061] Example fields include: actual beam height, column section size, plate thickness; structure deviation amount (such as wall inclination angle); residual deformation value and historical deformation trend.

[0062] State monitoring data: online monitoring results during building operation, used for function degradation judgment and global evaluation calculation, combined with function evaluation layer results to identify problem areas, sourced from sensors deployed in key structural positions or function spaces.

[0063] Example fields include: load response (stress / strain); temperature and humidity distribution, noise level; real-time energy consumption and equipment efficiency data.

[0064] Renovation demand data: function change, space reconstruction or expansion requirement description proposed by the user, used as the basis for triggering the repair design scheme, directly participating in template re-matching, sourced from user input system, operation log or space utilization analysis.

[0065] For example: hope to increase the open office area, add toilets and other functional changes; the area is used infrequently, and it is recommended to be combined or reduced; feedback on inconvenience (combined with complaint records or survey data).

[0066] Step S2: For the building object in the construction phase, input its multi-source heterogeneous data into the design optimization model based on convolutional neural network, and obtain an initial scheme set through template matching, template splicing and function evaluation;

[0067] The design optimization model is based on a convolutional neural network (CNN), and its core goal is to identify the optimal structure and function combination scheme based on multi-source heterogeneous feature input and output an evaluatable building configuration. The model has end-to-end learning and feature fusion capabilities, clear model hierarchy, and good generalization ability and sample adaptability.

[0068] The model input is multi-source heterogeneous data, and the output is an initial (building design) scheme that meets multi-dimensional performance constraints, forming a closed-loop design, evaluation and screening process. The specific data flow sequence is: multi-source heterogeneous data → feature tensor → convolution matching → template mapping → splicing combination → function evaluation → initial scheme output.

[0069] Key term definition:

[0070] Structure template: the spatial organization framework in building design, which defines the distribution logic and bearing relationship of functional areas, composed of multiple sub-regions with tolerance space. For example, a residential structure template includes a typical pattern of central axis passage + left and right symmetrical functional units.

[0071] Function template: refers to a combination unit that encapsulates spatial function configuration rules, such as office area template (including conference area, office table, storage area), toilet template (including washbasin, toilet, urinal layout), etc. Each template has a specific space occupation, function adaptation range and use demand constraint.

[0072] Tolerance space range: represents the acceptable functional layout floating space of each sub-region in the structure template, supporting flexible matching of different functional templates within this range. Its essence is a set of spatial boundaries and directional restrictions, used to constrain the spatial compatibility of function splicing.

[0073] The design optimization model based on convolutional neural network includes:

[0074] Input layer: used to receive multi-source heterogeneous data, which is converted into multi-source heterogeneous feature tensor after embedding encoding and standardization processing; this layer integrates environment data, specification data and user data into a unified tensor structure through position embedding, attribute embedding and specification vector encoding, ensuring synchronous mapping of space and semantics. After embedding, it is converted into standard floating point format and input into the convolution matching layer.

[0075] Convolution matching layer: feature extraction and matching operation on multi-source heterogeneous feature tensor through convolution kernel, output each matched structure template, and the matching feature map between its sub-regions and function templates, used to represent the structure template and its corresponding pre-selected function template set, the convolution kernel is composed of pre-trained structure template and function template; adopt double channel convolution mechanism, one channel loads structure template parameters, the other channel loads function template parameters. Through local convolution operation on tensor in each spatial region, extract spatial function adaptation relationship, the output matching feature map is used for subsequent splicing judgment.

[0076] Feature splicing layer: used to combine and splice each sub-region of the structure template and its corresponding pre-selected function template set according to the matching feature map, generate one by one the to-be-evaluated scheme covering all sub-regions, and form the to-be-evaluated scheme set, input to the function evaluation layer; take the sub-region of the structure template as the axis, enumerate and splice the matched function templates, and construct a plurality of complete to-be-evaluated building schemes. Boundary consistency detection is performed on the connection relationship (such as channel, shared facilities) between function templates in the splicing process to ensure legal layout.

[0077] Function evaluation layer: independently function evaluation, adjacent function evaluation and global function evaluation are respectively performed on each to-be-evaluated scheme, and the evaluation values of the three types of evaluation are weighted to obtain the function evaluation value; adopt triple path evaluation mechanism to quantify the sub-region matching quality (independence), adjacent function cooperation (adjacency) and overall layout integrity (globality), and comprehensively evaluate through adjustable weighting coefficient, and finally screen out the scheme set meeting the threshold condition.

[0078] Output layer: used to output the to-be-evaluated scheme meeting the preset threshold value of function evaluation value as the initial scheme set.

[0079] The function evaluation layer comprises:

[0080] Independent function evaluation unit: used to evaluate the independence of the function templates of each sub-region in the to-be-evaluated scheme, and obtain the corresponding function evaluation value; the evaluation values of the same type of function are combined to construct a function matrix, which is compared with the demand matrix extracted based on multi-source heterogeneous data; when the value of the corresponding item in the function matrix is greater than or equal to the value of the corresponding item in the demand matrix, it is considered to meet the requirements; the proportion of the number of items meeting the requirements to the total number of effective items in the demand matrix is calculated to obtain the quantitative independent function evaluation value;

[0081] Adjacent function evaluation unit: used for evaluating the adjacent relationship of the function configuration between adjacent sub-regions in the to-be-evaluated scheme, and obtaining the corresponding adjacent function evaluation value; based on the sub-region adjacent relationship graph in the structure template, the spatial connectivity parameters of the pre-selected function template are combined to analyze the spatial logical characteristics such as passability, line-of-sight continuity, acoustic or airflow interference between adjacent sub-regions; in the evaluation process, a function connection matrix is constructed, and it is compared with a set adjacent rationality standard matrix; when the function connection relationship matches the standard, it is determined as a reasonable connection; the proportion of the number of reasonable connection items to the number of all connection items is calculated to quantitatively output the adjacent function evaluation value.

[0082] Global function evaluation unit: used for global performance evaluation of the entire to-be-evaluated scheme, and obtaining the corresponding global function evaluation value; the evaluation contents include but are not limited to: overall energy efficiency of the scheme (based on heat load prediction and ventilation path simulation), spatial layout coordination (based on module level relationship and streamline path rationality), function integrity (whether it contains a function area that meets all items of the function demand matrix); the above indexes are calculated by normalization to form a global evaluation vector, and a global function evaluation value is obtained based on a weighting mechanism, and the weight is configured according to the building purpose, user preference and energy saving policy rules.

[0083] The structure template is a spatial structure prototype used to define the function area division in the building design scheme, which is composed of a plurality of sub-regions with adjustable spatial tolerance; each sub-region includes the following information fields: spatial boundary definition, acceptable function set, initial function preference, adjacent region list and structure rigidity level; the tolerance space range defines the allowed floating range of the sub-region in area, shape, opening direction and connection boundary, which is used to support the dynamic matching demand of diversified function combination under the same structure template; in the function template matching process, based on the semantic demand, site constraint and specification adaptation requirement in the multi-source heterogeneous feature tensor, the function template combination scheme meeting the tolerance condition is automatically selected; the matching strategy introduces the function preference weight to improve the function matching priority difference between the core region and the auxiliary region; the initial value of the function preference weight is based on the default setting of the structure template library, which supports dynamic adjustment according to user instructions or design targets, and realizes a semi-automatic control function template allocation mechanism.

[0084] In the triggering process of reacquiring the building design scheme, the sub-regions that have completed construction are set to a fixed state, and the corresponding structure and function template configurations remain unchanged; for the sub-regions in the unconstructed state, the function template matching and screening are continued within the original tolerance space range of the sub-regions; the screening process is still based on the spatial position and function preference weight.

[0085] During the construction phase, once the construction progress is monitored to be abnormal or a local area is completed, the mechanism of re-acquiring the scheme is automatically triggered. In this mechanism: the structure and function template configuration of the completed area is locked; the unfinished area still performs template matching based on the original tolerance space range and preference weight; the local area is regenerated and spliced with the original completed structure to ensure the design coherence. This mechanism significantly improves the coordination efficiency of the design and construction process, and supports the dynamic deployment mode of design as construction.

[0086] Step S3: Obtain the structural strength evaluation value of each initial scheme and the function evaluation value in the design optimization model, calculate the framework evaluation value after weighting, and select the initial schemes with the top three framework evaluation values in the initial scheme set as the candidate schemes, which also include three initial schemes with structural strength evaluation values exceeding the preset threshold and function evaluation values in the top three;

[0087] The framework evaluation value includes:

[0088] For each initial scheme, the structural strength evaluation value is obtained, which is used to represent the strength performance coefficient of the scheme in structural arrangement, construction rationality and stress path; the function evaluation value obtained in the design optimization model is obtained, which is used to measure the performance of the scheme in space use, function adaptation and overall performance; the structural strength evaluation value and the function evaluation value are weighted according to the preset weight coefficient to obtain the framework evaluation value of the initial scheme; the weight coefficient is configured according to user preference or building purpose, and dynamic adjustment is supported.

[0089] The structural strength evaluation value is a comprehensive strength performance index based on the component arrangement and stress path analysis of each sub-area in the structure template, and the evaluation content includes but is not limited to: component continuity, node stiffness coordination, force transmission path stability and material adaptability; the parameters used for strength evaluation are derived from the structural configuration data output by the design optimization model, the experience coefficients of similar structures in the specification database, and the preset load case simulation results;

[0090] A simplified structure evaluation model of finite element analysis (FEA) is used to generate the structural strength score of each initial scheme, and the score result is standardized to the [0, 1] interval to represent the relative strength level. The function evaluation value is the function evaluation value output by the function evaluation layer in step S2, which has fused independent function, adjacent function and global function, and represents the adaptability of the scheme in building function adaptation, space layout and performance target. The framework evaluation value is obtained by weighting the structural strength evaluation value and the function evaluation value in proportion, and the weight configuration is determined by the following two ways:

[0091] (1) Based on the standard configuration of the building project purpose (such as the ratio of structural strength to function evaluation for residential buildings is 3:2, and for hospitals is 4:1);

[0092] (2) Based on user preference dynamic input, support adjusting weight coefficient through graphical interface, and return to framework evaluation module in real time. The weighting method is linear normalization weighting model, expressed as: F frame =α×F struct +(1-α)×F func , wherein α is the structure strength evaluation weight, F struct is the structure strength evaluation value, F func is the function evaluation value, F frame is the framework evaluation value.

[0093] After framework evaluation of all initial schemes, the top three schemes with the highest framework evaluation values are selected as candidate schemes. In addition, to avoid schemes with weak structure performance but high function scores entering the candidate set, a set of condition filtering strategies is added: if the structure strength evaluation value of a scheme exceeds the preset strength threshold (such as 0.85), and the function evaluation value ranks in the top three, the scheme also enters the candidate scheme set, forming a cross optimization mechanism, enhancing the robustness and diversity of the evaluation system. The structure strength preset threshold is determined according to the importance level of the project and the recommended value of the specification, based on historical project data for empirical fitting, and supports project team self-defined configuration.

[0094] Step S4: dividing the building strength value of the candidate scheme into three grades, the building strength value being composed of the structure strength evaluation value and the material strength evaluation value, selecting the construction technology and structural material corresponding to the material strength evaluation value according to the building strength value of the three grades, and obtaining the execution evaluation value of each grade of the candidate scheme under the corresponding construction technology and structural material;

[0095] Building strength value definition and grading: the building strength value is a composite strength index, used to reflect the overall bearing capacity of the scheme under the design of the structure system and the configuration of the material performance. The value is composed of the following two dimensions: the structure strength evaluation value, the result obtained in the foregoing step S3, representing the load-carrying capacity and layout rationality of the structure design itself; the material strength evaluation value, representing the minimum material performance level required to meet the target building strength, generated in combination with indicators such as material compressive strength, elastic modulus, and shear modulus;

[0096] The division of the building strength value of the candidate scheme into three grades includes:

[0097] Three building strength levels are preset, which are high strength level, medium strength level and low strength level, each level corresponds to a fixed building strength target value, and the levels are revised in combination with actual requirements such as project seismic grade, load grade or service life; for each alternative, the required material strength evaluation value is calculated by dividing the building strength target value of the scheme by the structure strength evaluation value under each strength level, and when the calculated material strength evaluation value is lower than the minimum material strength standard, the combination is automatically removed to prevent non-compliant items from entering the execution evaluation stage.

[0098] Under the premise of meeting the material strength evaluation value, the most optimal combination of execution is selected from the alternative construction process and structural material, and the execution evaluation of each alternative under different strength levels is carried out based on the selected construction process and structural material combination, the execution evaluation includes construction period estimation and resource utilization efficiency, and the evaluation process is based on historical data fitting.

[0099] The execution evaluation includes: construction period estimation, which is evaluated in combination with historical construction period data in the construction progress module and process interference factors; resource utilization efficiency, which analyzes resource indicators such as the proportion of reusable components, material saving rate and energy consumption evaluation in the scheme; after the evaluation indicators are standardized, they are weighted and summed according to the weight to obtain the final execution score, and the user can adjust the weight to adapt to the actual project preference.

[0100] Step S5: The building strength value, functional evaluation value and execution evaluation value of the alternative scheme are weighted and calculated according to the user preference to obtain a comprehensive evaluation value and a final building design scheme;

[0101] The building strength value, functional evaluation value and execution evaluation value are respectively derived from: the building strength value ( R strength ), which is calculated by superimposing the structure strength and the material strength in step S4, representing the overall carrying capacity and safety performance of the scheme; the functional evaluation value ( R function ), which is output by the functional evaluation layer in step S2, including the weighted combination of independent function, adjacent function and global function evaluation; and the execution evaluation value ( R economic ), which is obtained by evaluating the construction process and structural material combination under each strength level in step S4, covering the construction period and resource efficiency dimensions.

[0102] The above three indicators respectively constitute the core performance vector of the candidate building scheme R strength ,R function , R economic}, are all normalized to the interval [0, 1] for quantitative calculation.

[0103] A user preference weight configuration interface is provided to support the design, investment or regulatory parties to express the difference in the degree of attention to different evaluation targets. The weight set is denoted as: W = {w strength ,w function ,w economic } , wherein w strength is the weight of the building strength value, w function is the weight of the functional evaluation value, w economic is the weight of the execution evaluation value, and the sum of the three weights is 1.

[0104] The weights are set in the following ways: default template configuration, providing standardized weight settings according to building use types (such as residential, office, industrial plant, medical building, etc.); manual adjustment mode, users set the proportion of each index through a graphical interface slider, and the calculation result is fed back in real time; automatic learning optimization (optional): based on user historical evaluation data, the optimal weight distribution is iteratively fitted through Bayesian optimization or genetic algorithm.

[0105] Weighted addition is used for multi-index fusion to obtain a comprehensive evaluation value, which represents the overall performance of the current building design scheme in terms of structural safety, functional matching and execution feasibility. The comprehensive evaluation process is supported to be performed under different strength levels.

[0106] Based on the comprehensive evaluation value sorting of all alternative schemes, the scheme with the highest comprehensive evaluation value is selected as the final building design scheme. If there are multiple schemes with close evaluation values (difference less than a set threshold ε), the designer is prompted to make a manual comparison, and the final decision is made from the dimensions of spatial layout, aesthetic style or sustainability. At the same time, the unselected but high-scoring schemes are kept as design version history records, supporting subsequent adjustment or quick replacement in response to sudden changes.

[0107] Step S6: According to the construction progress data and on-site abnormal feedback, the building design scheme is reacquired to realize real-time optimization of the building design scheme.

[0108] On-site abnormal feedback refers to problem data found through monitoring systems or manual reporting during construction, including: abnormal foundation settlement; environmental disturbance (such as climate mutation, noise control not meeting standards); non-compliant materials, equipment failure; on-site safety risks or unforeseen structural change requirements.

[0109] Trigger strategies include but are not limited to the following scenarios: the construction progress of a certain sub-area lags behind the set threshold (e.g. ≥7 days); the actual material used in a certain construction node is inconsistent with the material strength in the design model; the abnormal feedback contains suggestions for changes in key structures or space functions; external systems such as environmental impact assessment and structural monitoring push for correction opinions.

[0110] When any of the trigger conditions are met, the response process is executed. First, the structure template sub-area of the completed construction area is frozen, i.e. set to a fixed state, to prevent it from being re-matched or replaced in subsequent optimization, ensuring that the completed part is not modified. For sub-areas that are still in the unconstructed state, based on the current input data (including the latest construction data and abnormal feedback), perform function template matching; sub-area splicing; function evaluation and strength evaluation; comprehensive evaluation value calculation; dynamic weight adjustment (optional), users can reset the comprehensive evaluation weight based on the priority target of the current construction stage (such as speeding up the progress, reducing material usage, etc.). Form a new building design scheme and output construction update drawings and BIM instruction documents for quick deployment by the construction party.

[0111] After updating the scheme, the data used in this round of reconstruction (such as specific lagging areas, function change reasons) and the reconstruction results (such as changed modules, optimized execution performance improvement) are written back to the building design database, forming a traceable design evolution path for subsequent quality analysis or data training optimization model.

[0112] Step S7: For the completed stage building object, based on its structure mapping data and state monitoring data, obtain the building design scheme and corresponding building strength value, and input it into the design optimization model to obtain the corresponding function evaluation value; based on the function evaluation value and the building strength value, identify the sub-area of the building design scheme whose structure template has function degradation or insufficient strength, and perform local function template re-matching and structure template splicing operations to generate the corresponding repair design scheme.

[0113] Function evaluation: Through the function evaluation layer, compare the existing space distribution with the function demand to determine whether there is insufficient function load (such as a serious lack of classroom capacity in a teaching building), whether the space adjacency configuration is unreasonable (such as the proximity of a surgical area to a pollution source), and whether the energy efficiency deviates from the original setting.

[0114] Strength evaluation: Based on the state monitoring data and the initial design strength target of the building, the current building strength value is recalculated. Once it is found that the actual structure performance of a local area is lower than the safety margin threshold of the original design (e.g. lower than 80% of the design strength), it is marked as an area with insufficient strength.

[0115] The two types of evaluation results are combined to construct a structural template risk marking map, which highlights the areas of functional or strength degradation. For the marked degradation sub-regions, only the following optimization process is performed: the fixed sub-regions in the original structural template framework are retained; the functional degradation sub-regions are re-matched with the functional template set that best meets the current use requirements; the strength deficiency sub-regions are re-calculated to determine their acceptable material strength threshold, and local reinforcement strategies (such as adding steel structures, carbon fiber reinforcement, etc.) are recommended based on the latest construction process; before splicing the new template, the structural compatibility between the new and old components is automatically evaluated to ensure the stability of the new and old connection nodes; the new template is spliced into the original structural template to form a new repair design scheme.

[0116] After the repair design is generated, the updated structural template model and functional layout diagram after repair are output; the construction process specification for the repair area; the material demand prediction table; the functional and strength improvement report before and after repair, quantifying the repair benefits. All repair data are written into the historical version database for subsequent building life cycle management (such as operation and maintenance planning, secondary transformation) or training optimization models.

[0117] Embodiment 2, a building design real-time optimization processing system based on multi-source heterogeneous data, see Figure 1 as shown, comprising the following modules:

[0118] A data acquisition module for acquiring multi-source heterogeneous data of a building object;

[0119] A design optimization module for inputting the multi-source heterogeneous data of the building object into a design optimization model based on a convolutional neural network when the building object is in the construction phase, and outputting an initial scheme set;

[0120] An alternative acquisition module for acquiring structural strength evaluation values and framework evaluation values of each initial scheme, and combining the functional evaluation values to screen alternative schemes;

[0121] An execution analysis module for acquiring execution evaluation values of the building strength values of each level of the alternative schemes under corresponding construction processes and structural materials;

[0122] A comprehensive evaluation module for acquiring comprehensive evaluation values of the alternative schemes and determining a final building design scheme;

[0123] A feedback optimization module for triggering the reacquisition process of the building design scheme based on construction progress data or on-site abnormal feedback;

[0124] A completed repair module for generating a corresponding repair design scheme when the building object is in the completed stage.

[0125] The above detailed description of the specific embodiments of the present application has been given to understand the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above description is only a specific embodiment of the present application and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for real-time optimization processing of architectural design based on multi-source heterogeneous data, characterized in that, The application relates to a method for building object design optimization. The method comprises the following steps: acquiring multi-source heterogeneous data of a building object, including environment data, specification data, construction progress data and user demand data of a construction stage, or structure surveying data, state monitoring data and reconstruction demand data of a completed stage; for a building object in a construction stage, inputting the multi-source heterogeneous data into a design optimization model based on a convolutional neural network, and acquiring an initial scheme set through template matching, template splicing and function evaluation; the design optimization model based on the convolutional neural network comprises: an input layer: used for receiving the multi-source heterogeneous data, and converting the multi-source heterogeneous data into a multi-source heterogeneous feature tensor through embedding coding and standardization processing; a convolution matching layer: used for performing feature extraction and matching operation on the multi-source heterogeneous feature tensor through a convolution kernel, outputting each matched structure template and a matching feature map between a sub-region of the structure template and a preselected function template set corresponding to the structure template, and used for representing the structure template and the preselected function template set corresponding to the structure template, wherein the convolution kernel is composed of pre-trained structure templates and function templates; a feature splicing layer: used for combining and splicing each sub-region of the structure template and the preselected function template set corresponding to the structure template according to the matching feature map, generating a to-be-evaluated scheme covering all the sub-regions one by one, and forming a to-be-evaluated scheme set and inputting the to-be-evaluated scheme set into a function evaluation layer; a function evaluation layer: used for performing independent function evaluation, adjacent function evaluation and global function evaluation on each to-be-evaluated scheme respectively, and performing weighted calculation on evaluation values of the three types of evaluation to obtain a function evaluation value; an output layer: used for outputting a to-be-evaluated scheme with a function evaluation value satisfying a preset threshold as an initial scheme set; the function evaluation layer comprises: an independent function evaluation unit: used for performing independence evaluation on function templates of each sub-region in a to-be-evaluated scheme, obtaining a corresponding function evaluation value, summing up evaluation values of the same type of function to construct a function matrix, and comparing the function matrix with a demand matrix extracted based on the multi-source heterogeneous data; when a value of a corresponding item in the function matrix is greater than or equal to a value of a corresponding item in the demand matrix, the to-be-evaluated scheme is considered to meet the requirements; and a quantitative independent function evaluation value is obtained by calculating a proportion of the number of items meeting the requirements in a total number of effective items in the demand matrix; an adjacent function evaluation unit: used for evaluating a function configuration relationship between adjacent sub-regions in a to-be-evaluated scheme, judging rationality of the function configuration relationship in terms of spatial layout and function connection, and converting an evaluation result into an adjacent function evaluation value; a global function evaluation unit: used for comprehensively evaluating energy consumption efficiency, spatial layout coordination and function integrity of a whole to-be-evaluated scheme, and outputting a corresponding global function evaluation value; obtaining a structure strength evaluation value of each initial scheme and a function evaluation value of the initial scheme in the design optimization model, performing weighted calculation on the structure strength evaluation value and the function evaluation value to obtain a framework evaluation value, and selecting three initial schemes with the top three framework evaluation values in the initial scheme set as candidate schemes, wherein the candidate schemes further include three initial schemes with structure strength evaluation values exceeding a preset threshold and function evaluation values in the top three; The building strength value of the alternative scheme is divided into three levels, the building strength value is composed of a structure strength evaluation value and a material strength evaluation value, a construction process and a structural material corresponding to the material strength evaluation value are selected according to the building strength value of the three levels, and an execution evaluation value of the building strength value of each level of the alternative scheme under the corresponding construction process and structural material is obtained; The building strength value, the function evaluation value and the execution evaluation value of the alternative scheme are weighted and calculated according to the user preference to obtain a comprehensive evaluation value, and a final building design scheme is obtained; According to the construction progress data and the on-site abnormal feedback, the building design scheme is triggered to be reacquired to realize real-time optimization of the building design scheme. 2.The method of claim 1, wherein, The structure template is composed of a plurality of sub-regions with a tolerance space range, and the preselected function template is matched and obtained in the tolerance space range of the corresponding sub-region based on the input multi-source heterogeneous feature tensor; in the matching process, the function template is screened according to the spatial position of the sub-region and the corresponding function preference weight; the function preference weight is initially set according to the spatial position of the sub-region, and can be adjusted according to the user demand. 3.The method of claim 2, wherein, The structure template further comprises: in the process of triggering the reacquisition of the building design scheme, the sub-regions that have completed construction are set to a fixed state, and the corresponding structure and function template configurations remain unchanged; for the sub-regions in the unconstructed state, the function template matching and screening are continued in the original tolerance space range, and the screening process is still based on the spatial position and the function preference weight. 4.The method for building design real-time optimization processing based on multi-source heterogeneous data according to claim 1, wherein, The obtaining framework evaluation value comprises: The structure strength evaluation value of each initial scheme is obtained, which is used to represent the strength performance coefficient of the scheme in the structure arrangement, the construction rationality and the stress path; the function evaluation value obtained by the initial scheme in the design optimization model is obtained, which is used to measure the performance of the scheme in the space use, the function adaptation and the overall performance; the structure strength evaluation value and the function evaluation value are weighted and calculated according to the preset weight coefficient to obtain the framework evaluation value of the initial scheme; the weight coefficient is configured according to the user preference or the building purpose, and can be dynamically adjusted. 5.The method for building design real-time optimization processing based on multi-source heterogeneous data according to claim 1, wherein, The building strength value of the alternative scheme is divided into three levels, the building strength value is composed of a structure strength evaluation value and a material strength evaluation value, a construction process and a structural material corresponding to the material strength evaluation value are selected according to the building strength value of the three levels, and an execution evaluation value of the building strength value of each level of the alternative scheme under the corresponding construction process and structural material is obtained; The building strength value, the function evaluation value and the execution evaluation value of the alternative scheme are weighted and calculated according to the user preference to obtain a comprehensive evaluation value, and a final building design scheme is obtained; 6.The method for building design real-time optimization processing based on multi-source heterogeneous data according to claim 1, characterized in that, According to the construction progress data and the on-site abnormal feedback, the building design scheme is triggered to be reacquired to realize real-time optimization of the building design scheme. The structure template is composed of a plurality of sub-regions with a tolerance space range, and the preselected function template is matched and obtained in the tolerance space range of the corresponding sub-region based on the input multi-source heterogeneous feature tensor; in the matching process, the function template is screened according to the spatial position of the sub-region and the corresponding function preference weight; the function preference weight is initially set according to the spatial position of the sub-region, and can be adjusted according to the user demand. The structure template further comprises: in the process of triggering the reacquisition of the building design scheme, the sub-regions that have completed construction are set to a fixed state, and the corresponding structure and function template configurations remain unchanged; for the sub-regions in the unconstructed state, the function template matching and screening are continued in the original tolerance space range, and the screening process is still based on the spatial position and the function preference weight. The obtaining framework evaluation value comprises: The structure strength evaluation value of each initial scheme is obtained, which is used to represent the strength performance coefficient of the scheme in the structure arrangement, the construction rationality and the stress path; the function evaluation value obtained by the initial scheme in the design optimization model is obtained, which is used to measure the performance of the scheme in the space use, the function adaptation and the overall performance; the structure strength evaluation value and the function evaluation value are weighted and calculated according to the preset weight coefficient to obtain the framework evaluation value of the initial scheme; the weight coefficient is configured according to the user preference or the building purpose, and can be dynamically adjusted. The building strength value of the alternative scheme is divided into three levels, the building strength value is composed of a structure strength evaluation value and a material strength evaluation value, a construction process and a structural material corresponding to the material strength evaluation value are selected according to the building strength value of the three levels, and an execution evaluation value of the building strength value of each level of the alternative scheme under the corresponding construction process and structural material is obtained; The building strength value, the function evaluation value and the execution evaluation value of the alternative scheme are weighted and calculated according to the user preference to obtain a comprehensive evaluation value, and a final building design scheme is obtained; According to the construction progress data and the on-site abnormal feedback, the building design scheme is triggered to be reacquired to realize real-time optimization of the building design scheme. For the building object in the construction stage, based on the structure surveying data and state monitoring data, the building design scheme and the corresponding building strength value are obtained, and input into the design optimization model to obtain the corresponding function evaluation value; Based on the function evaluation value and the building strength value, the sub-regions of the function degradation or the strength deficiency in the structure template of the building design scheme are identified, and the local function template re-matching and structure template splicing operations are performed to generate the corresponding repair design scheme.

7. A building design real-time optimization processing system based on multi-source heterogeneous data, characterized in that, The system applies the building design real-time optimization processing method based on multi-source heterogeneous data in any one of the above claims 1 to 6, comprising: a data acquisition module for acquiring multi-source heterogeneous data of a building object; a design optimization module for inputting the multi-source heterogeneous data of the building object into a design optimization model constructed based on a convolutional neural network when the building object is in the construction stage, and outputting an initial scheme set; an alternative acquisition module for acquiring structure strength evaluation values and frame evaluation values of each initial scheme, and screening alternative schemes in combination with function evaluation values; an execution analysis module for obtaining execution evaluation values of building strength values of each level of the alternative scheme under corresponding construction technology and structure material; a comprehensive evaluation module for obtaining a comprehensive evaluation value of the alternative scheme and determining a final building design scheme; a feedback optimization module for triggering the reacquisition process of the building design scheme based on construction progress data or on-site abnormal feedback; a completed repair module for generating a corresponding repair design scheme when the building object is in the completion stage.

Images