CN-122020325-A - Intelligent evaluation management system and method for historical building repair results

Abstract

The invention relates to the technical field of building repair result evaluation management, in particular to an intelligent evaluation management system and method for historical building repair results, wherein the system comprises a repair data acquisition module for acquiring multi-mode repair data, a digital twin body construction module for generating repair completion digital twin bodies by fusing the data, the multi-dimensional repair feature extraction module is used for quantitatively analyzing and generating a structural multi-dimensional repair feature vector and comprises a geometric deviation quantization unit, a material performance evaluation unit and a value intervention quantization unit, the fusion evaluation engine is used for inputting the feature vector into the machine learning model to obtain a comprehensive repair quality grade, and the visual interaction and management terminal is used for presenting a digital twin body and evaluation details. The invention realizes multidimensional quantitative evaluation and intelligent comprehensive judgment of the repair result, and visually displays the repair result in a visual mode, thereby providing scientific basis for acceptance and archiving of the repair result of the historical building.

Inventors

• HUANG ZHILONG
• ZHONG QIJUN
• LU XIAOTONG
• XIE TINGTING

Assignees

• 福建省工大规划设计院有限公司

Dates

Publication Date

20260512

Application Date

20260416

Claims (9)

1. 1. An intelligent evaluation management system for historical building repair results is characterized by comprising the following components: The system comprises a repair data acquisition module, a target historical building repair module, a target data processing module and a target data processing module, wherein the repair data acquisition module is used for acquiring multi-mode repair data after the target historical building repair is completed, wherein the multi-mode repair data at least comprises space geometric data, apparent image data and material performance data; The digital twin body construction module is connected with the repair data acquisition module and used for fusing the multi-mode repair data to construct a repair completion digital twin body representing the real state of the target historical building after repair; The multi-dimensional repair feature extraction module is connected with the digital twin body construction module and is used for carrying out multi-dimensional quantitative analysis on the repaired completion digital twin body to generate a structural multi-dimensional repair feature vector, and the multi-dimensional repair feature extraction module comprises: The geometric deviation quantifying unit is used for extracting the geometric characteristics of the repair completion digital twin body, comparing the geometric characteristics with a preset repair design digital model and generating a geometric deviation characteristic value; The material performance evaluation unit is used for extracting material performance data related to the repair completion digital twin body, and comparing the material performance data with a standard value in a preset historical value evaluation reference library to generate a material performance conformity characteristic value; the value intervention quantification unit is used for identifying and quantifying the intervention degree of the core value element reflected by the completion digit twin body based on a preset value element identification model, and generating a value intervention index characteristic value; The fusion evaluation engine is respectively connected with the geometric deviation quantification unit, the material performance evaluation unit and the value intervention quantification unit and is used for combining the geometric deviation characteristic value, the material performance conformity characteristic value and the value intervention index characteristic value to form a structural multidimensional maintenance characteristic vector, and inputting the structural multidimensional maintenance characteristic vector into a pre-trained maintenance quality comprehensive evaluation machine learning model for processing so as to obtain the comprehensive maintenance quality grade of the target historical building; The visual interaction and management terminal is connected with the fusion evaluation engine and is used for displaying the repair completion digital twin body in a visual mode, and displaying the comprehensive repair quality grade and the corresponding multidimensional repair feature vector details thereof for the manager to review and archive.
2. 2. The intelligent evaluation management system for historical building repair results according to claim 1, wherein the repair data acquisition module is used for acquiring multi-mode repair data after the target historical building repair is completed, and specifically comprises: The three-dimensional laser scanning unit is used for carrying out three-dimensional laser scanning on the target historical building, and acquiring space point cloud data after the repair is completed as space geometric data; The high-definition image acquisition unit is used for shooting multi-view high-definition images of the target historical building to obtain apparent image data; the material performance detection unit is used for carrying out on-site nondestructive detection or sampling laboratory detection on key parts of the target historical building to obtain material performance data; The data synchronization and preprocessing unit is respectively connected with the three-dimensional laser scanning unit, the high-definition image acquisition unit and the material performance detection unit and is used for performing space-time registration, denoising and normalization on the acquired space point cloud data, apparent image data and material performance data to form multi-mode repair data after registration and alignment, and transmitting the multi-mode repair data to the digital twin body construction module.
3. 3. The intelligent evaluation and management system for historical building repair results according to claim 1, wherein the digital twin body construction module is used for fusing multi-mode repair data to construct a repair completion digital twin body representing the real state of a target historical building after repair, and specifically comprises: the geometric model reconstruction unit is used for preprocessing the space geometric data in the multi-mode repair data, including point cloud filtering, splicing and triangular gridding, and generating a basic geometric model which reflects the repaired building geometric form with high precision; The texture mapping fusion unit is connected with the geometric model reconstruction unit and is used for receiving the apparent image data in the basic geometric model and the multi-mode repair data, mapping the multi-view high-definition image to the surface of the basic geometric model through feature point matching and camera pose resolving, and generating a textured geometric model with real apparent textures after texture fusion and optimization; And the performance information association unit is connected with the texture mapping fusion unit and is used for receiving the material performance data in the textured geometric model and the multi-mode repair data, and binding the material performance data to corresponding building components or parts in the textured geometric model one by one based on the space position coordinates to form a repair completion digital twin body with integrated geometric, texture and material performance information.
4. 4. The intelligent evaluation and management system for historical building repair results according to claim 1, wherein the geometric deviation quantification unit is used for extracting geometric characteristics of a repair completion digital twin body and comparing the geometric characteristics with a preset repair design digital model to generate geometric deviation characteristic values, and specifically comprises the following steps: The geometric deviation quantification unit receives the repaired completion digital twin body, performs overall grid light weight treatment on the repaired completion digital twin body, extracts a geometric feature surface set to be measured reflecting the actual form of the repaired building, performs spatial registration alignment on the geometric feature surface set to be measured and a corresponding design geometric feature surface set in a pre-stored repaired design digital model to enable the geometric feature surface set to be measured and the corresponding design geometric feature surface set to be positioned under the same spatial coordinate system, calculates the spatial distance deviation and the angle deviation between each corresponding feature surface based on the registered geometric feature surface set to be measured and the design geometric feature surface set, and forms an initial deviation data set; And respectively calculating the overall average geometric deviation, the local maximum geometric deviation of the key node part and the overall deviation distribution uniformity based on the effective deviation data set, and carrying out weighted fusion on the overall average geometric deviation, the local maximum geometric deviation and the overall deviation distribution uniformity to generate a geometric deviation characteristic value for representing the coincidence degree between the repaired building geometric form and the design intention.
5. 5. The intelligent evaluation and management system for historical building repair results according to claim 4, wherein the material performance evaluation unit is used for extracting material performance data associated with a repair completion digital twin body, comparing the material performance data with standard values in a preset historical value evaluation reference library, and generating a material performance conformity characteristic value, and specifically comprises: The material performance evaluation unit receives the repair completion digital twin body, traverses all building components bound with material performance data in the repair completion digital twin body, and extracts the material type, the detection position and the material performance actual measurement value of the corresponding position of each building component; according to the historical age, building style and material process characteristics of the target historical building, searching from a preset historical value evaluation reference library and determining a historical standard performance interval or reference value corresponding to each material type; Comparing each extracted measured value of the material performance with a corresponding historical standard performance interval or reference value item by item, calculating single-point coincidence degree of each performance index according to the degree that the measured value falls into the standard interval or the degree close to the reference value, and correcting the single-point coincidence degree by combining the confidence degree of the material performance detection method; according to the value importance level of the building components in the historical building, corresponding weight coefficients are given to the single-point coincidence degree of different building components; and carrying out fusion statistics on the weighted single-point coincidence degree of all building components, and calculating to obtain an overall material performance comprehensive score which is used as a material performance coincidence degree characteristic value for representing the coincidence degree between the repaired building material performance and the historical primitive requirement.
6. 6. The intelligent evaluation and management system for historical building repair results according to claim 5, wherein the value intervention quantifying unit is configured to identify and quantify a core value element intervention degree reflected by a repair completion digital twin based on a preset value element identification model, and generate a value intervention index characteristic value, and specifically comprises: The value intervention quantification unit receives the repaired completion digital twin, and invokes a preset value element identification model to carry out traversal analysis on the repaired completion digital twin so as to identify various core value elements contained in the repaired completion digital twin, wherein the core value elements at least comprise original historical components, traditional process traces, key decorative patterns and building details with times of characteristics; For each identified core value element, the value intervention quantifying unit quantitatively analyzes the processing mode and the degree of the core value element in the current repair process by comparing the repair completion digital twin with the historical state information recorded in the historical archive data of the core value element before repair, wherein the processing mode at least comprises original retention, repair reinforcement, local replacement or overall restoration; the value intervention quantification unit calculates a single intervention degree value of each core value element according to a preset intervention coefficient corresponding to each processing mode and combining the proportion of the actual processing area or volume of the core value element to the whole body quantity of the core value element; Determining a weight coefficient of each core value element in the overall value composition of the target historical building according to a pre-constructed value hierarchy evaluation system, wherein the value hierarchy evaluation system at least comprises three dimensions of comprehensive scores of value scarcity, historical representativeness and spatial zone significance; And carrying out weighted accumulation on the single intervention degree value of each core value element and the corresponding weight coefficient, and carrying out normalization processing on the accumulation result according to a preset value intervention tolerance threshold range to generate a value intervention index characteristic value for representing the integral influence degree of the repair engineering on the core value elements of the historical building.
7. 7. The intelligent evaluation management system for repairing achievement of historical architecture according to claim 6, wherein the fusion evaluation engine is used for combining the geometric deviation characteristic value, the material performance conformity characteristic value and the value intervention index characteristic value to form a structural multidimensional repairing characteristic vector, inputting the structural multidimensional repairing characteristic vector into a pre-trained repairing quality comprehensive evaluation machine learning model for processing so as to obtain a comprehensive repairing quality grade of a target historical architecture, and specifically comprises: The fusion evaluation engine firstly receives the geometric deviation characteristic value from the geometric deviation quantification unit, the material performance conformity characteristic value from the material performance evaluation unit and the value intervention index characteristic value from the value intervention quantification unit, respectively performs dimensionless normalization processing on the three received characteristic values, eliminates the difference of different dimensionalities and magnitude orders, and obtains normalized geometric deviation characteristic components, material performance conformity characteristic components and value intervention index characteristic components; The method comprises the steps of inputting a generated structured multi-maintenance repair feature vector into a pre-trained repair quality comprehensive evaluation machine learning model, wherein the repair quality comprehensive evaluation machine learning model is a multi-classification model obtained by training by adopting a gradient lifting tree algorithm based on sample data in a historical repair case library, and the sample data comprises a plurality of structured multi-maintenance repair feature vectors corresponding to evaluated historical building repair projects and comprehensive repair quality grade labels thereof assessed by experts; The fusion evaluation engine calls the machine learning model to perform forward calculation on the current structured multidimensional repairing feature vector, and integrated voting or weighted summation of a plurality of decision trees is performed in the machine learning model to output probability distribution that target historical buildings respectively belong to a plurality of preset comprehensive repairing quality grades; The fusion evaluation engine selects the grade with the maximum probability value as the comprehensive repair quality grade of the target historical building according to the output probability distribution, and stores the grade, the corresponding geometric deviation characteristic value, the material performance conformity characteristic value and the detail of the value intervention index characteristic value in a correlated manner so as to enable the visualized interaction and the management terminal to call and display; Meanwhile, the fusion evaluation engine also calculates the contribution weights of the geometric deviation characteristic value, the material performance coincidence degree characteristic value and the value intervention index characteristic value to the current grade judgment result by utilizing the characteristic importance analysis function of the machine learning model, and outputs the contribution weights as explanatory information for assisting management personnel in understanding the influence degree of each maintenance dimension on the final quality grade.
8. 8. The intelligent evaluation management system for historical building repair results according to claim 1, wherein the visual interaction and management terminal is used for displaying repair completion digital twin in a visual manner and displaying comprehensive repair quality grades and corresponding multidimensional repair feature vector details for manager to review and archive, and specifically comprises: the data receiving and analyzing unit is connected with the fusion evaluation engine and is used for receiving the repair completion digital twin, the comprehensive repair quality grade, the multidimensional repair feature vector details and the weight interpretation information of each feature value contribution output by the fusion evaluation engine, analyzing and converting the received data to generate a data structure required by visual rendering; The three-dimensional visual rendering unit is connected with the data receiving and analyzing unit and is used for constructing a three-dimensional visual scene based on the analyzed repair completion digital twin body and rendering the three-dimensional visual scene in real time, displaying a three-dimensional model of the object history building after repair on the interactive interface, and supporting a user to perform rotation, translation, scaling and sectioning operations in a mouse or touch mode so as to observe building details from any view angle; The evaluation information superposition display unit is connected with the three-dimensional visual rendering unit and is used for superposing and displaying comprehensive repair quality grade, multidimensional repair feature vector details and contribution weight information thereof on the periphery of the three-dimensional visual scene or the floating panel in a graphical mode, endowing corresponding color highlighting marks for different component parts of the three-dimensional model according to the numerical intervals of each feature value, and visually reflecting the geometric deviation, the material performance conformity and the spatial distribution of the value intervention degree; The interactive detail query unit is connected with the evaluation information superposition display unit and is used for responding to the selection operation of a user on a specific component or part in the three-dimensional model, querying and popup window to display detailed repair detection data, material performance actual measurement values, deviation values between the components and the design model, core value element intervention details and the contribution degree of the component to the whole evaluation level, so that the linkage display of the whole-to-local multi-level information is realized; And the evaluation report generation and archiving unit is connected with the interactive detail query unit and is used for integrating the currently displayed repair completion digital twin view, comprehensive repair quality grade, multidimensional repair feature vector details, each feature value contribution weight and component detailed information queried by a user to generate a repair result evaluation report in a standard format according to a user instruction, and providing export and online archiving functions for a manager to review and store for a long time.
9. 9. A method for intelligently evaluating and managing historical building repair results, which is executed by the system of any one of claims 1-8, and is characterized in that the method comprises the following steps: Step S1, acquiring multi-mode repair data after finishing repairing a target historical building through a repair data acquisition module, wherein the multi-mode repair data at least comprises space geometric data, apparent image data and material performance data; s2, fusing the multi-mode repair data through a digital twin body construction module to construct a repair completion digital twin body representing the real state of the target historical building after repair; and S3, carrying out multidimensional quantitative analysis on the repaired completion digital twin body through a multidimensional repairing feature extraction module to generate a structured multidimensional repairing feature vector, wherein the steps specifically comprise: S3-1, extracting geometric characteristics of a repair completion digital twin body through a geometric deviation quantification unit, performing spatial registration alignment with a preset repair design digital model, calculating spatial distance deviation and angle deviation between corresponding characteristic surfaces, removing outlier deviation through statistical analysis, and fusing to generate overall average geometric deviation, local maximum geometric deviation and overall deviation distribution uniformity as geometric deviation characteristic values; S3-2, traversing and repairing all building components bound with material performance data in a completion digital twin body through a material performance evaluation unit, extracting the material type and the material performance actual measurement value of each building component, retrieving a corresponding historical standard performance interval from a preset historical value evaluation reference library, comparing the actual measurement value with the standard interval item by item to calculate single-point coincidence degree, and carrying out weighted fusion by combining with component value importance weights to obtain a material performance coincidence degree characteristic value; S3-3, traversing and analyzing the repair completion digital twin body by calling a preset value element identification model through a value intervention quantification unit, identifying core value elements in the repair completion digital twin body, quantifying the processing mode and the degree of each core value element in the repair according to historical archival data before the repair, carrying out weighted accumulation according to a preset intervention coefficient and the value weight of the element, and carrying out normalization processing to obtain a value intervention index characteristic value; Step S4, performing dimensionless normalization processing on the geometric deviation characteristic value, the material performance conformity characteristic value and the value intervention index characteristic value through a fusion evaluation engine, splicing and combining according to a preset sequence to form a structural multidimensional maintenance characteristic vector, inputting the structural multidimensional maintenance characteristic vector into a pre-trained maintenance quality comprehensive evaluation machine learning model, outputting probability distribution of a target historical building belonging to preset comprehensive maintenance quality grades through forward calculation of the model, selecting the grade with the highest probability as the comprehensive maintenance quality grade, and calculating contribution weight of each characteristic value to a judging result by utilizing a characteristic importance analysis function of the model as explanatory information; And S5, receiving and analyzing repair completion digital twin bodies, comprehensive repair quality grades, multidimensional repair feature vector details and contribution weight interpretation information through a visual interaction and management terminal, constructing a three-dimensional visual scene, rendering and displaying a repaired three-dimensional model in real time, superposing and displaying evaluation information on the periphery of the model in a graphical mode, giving color highlighting identification to different component parts of the model according to numerical intervals of each feature value, inquiring and displaying detailed detection data and evaluation details of the component in response to selection operation of the component by a user, and finally integrating and generating a repair result evaluation report in a standard format according to a user instruction and providing a export and archiving function.

Description

Intelligent evaluation management system and method for historical building repair results Technical Field The invention relates to the technical field of building repair result evaluation management, in particular to an intelligent evaluation management system and method for historical building repair results. Background The existing historical building repair result evaluation technology mainly relies on manual field investigation, two-dimensional drawing comparison and empirical judgment, and the evaluation process lacks effective integration and quantitative analysis of multi-source heterogeneous data. The traditional method is difficult to fuse and utilize space geometric information, apparent image information and material performance information, so that the evaluation result is high in subjectivity and poor in repeatability. In addition, the prior art lacks quantification means in the aspect of evaluating the intervention degree of the repair engineering on the historical value of the building core, and cannot scientifically measure the influence of the repair behavior on the original components, the traditional process and other valuable elements. Most of the assessment results are presented in a paper report or a simple two-dimensional chart, and lack of an intuitive three-dimensional visual carrier, so that a manager is difficult to quickly master the overall state and the details of all parts after the construction is repaired. Therefore, a system and a method for evaluating and managing repair results, which can realize multi-modal data fusion, multi-dimensional quantitative analysis, intelligent comprehensive evaluation and visual interaction, are needed. Disclosure of Invention The invention aims to provide an intelligent evaluation management system and method for historical building repair results, which are used for solving the problems in the background technology. In order to achieve the above purpose, the present invention provides the following technical solutions: An intelligent evaluation management system for historical building repair results, comprising: The system comprises a repair data acquisition module, a target historical building repair module, a target data processing module and a target data processing module, wherein the repair data acquisition module is used for acquiring multi-mode repair data after the target historical building repair is completed, wherein the multi-mode repair data at least comprises space geometric data, apparent image data and material performance data; The digital twin body construction module is connected with the repair data acquisition module and used for fusing the multi-mode repair data to construct a repair completion digital twin body representing the real state of the target historical building after repair; The multi-dimensional repair feature extraction module is connected with the digital twin body construction module and is used for carrying out multi-dimensional quantitative analysis on the repaired completion digital twin body to generate a structural multi-dimensional repair feature vector, and the multi-dimensional repair feature extraction module comprises: The geometric deviation quantifying unit is used for extracting the geometric characteristics of the repair completion digital twin body, comparing the geometric characteristics with a preset repair design digital model and generating a geometric deviation characteristic value; The material performance evaluation unit is used for extracting material performance data related to the repair completion digital twin body, and comparing the material performance data with a standard value in a preset historical value evaluation reference library to generate a material performance conformity characteristic value; the value intervention quantification unit is used for identifying and quantifying the intervention degree of the core value element reflected by the completion digit twin body based on a preset value element identification model, and generating a value intervention index characteristic value; The fusion evaluation engine is respectively connected with the geometric deviation quantification unit, the material performance evaluation unit and the value intervention quantification unit and is used for combining the geometric deviation characteristic value, the material performance conformity characteristic value and the value intervention index characteristic value to form a structural multidimensional maintenance characteristic vector, and inputting the structural multidimensional maintenance characteristic vector into a pre-trained maintenance quality comprehensive evaluation machine learning model for processing so as to obtain the comprehensive maintenance quality grade of the target historical building; The visual interaction and management terminal is connected with the fusion evaluation engine and is used for displaying the repair completion digital twin body in a visual