CN-121998188-A - Construction engineering quality risk prediction method and system based on big data and artificial intelligence

CN121998188ACN 121998188 ACN121998188 ACN 121998188ACN-121998188-A

Abstract

The application relates to the technical field of data processing, discloses a construction engineering quality risk prediction method and system based on big data and artificial intelligence, and aims to solve the problems that a traditional method depends on manpower, information fragmentation and risk early warning is lagged. The method comprises the steps of collecting and cleaning multi-source heterogeneous engineering data, extracting multi-dimensional feature vectors from standardized data based on a knowledge graph, calculating features by utilizing a hybrid intelligent model, outputting risk prediction probability and grades, visualizing results and generating a structured early warning report, pushing the report through a mobile terminal and tracking treatment feedback to form closed-loop management. The system comprises a data acquisition and integration module, a data center and feature engineering module, an intelligent risk prediction model module, a risk visualization and decision support module and a mobile terminal early warning and co-processing module. The method can realize prospective and intelligent prediction and closed loop control of engineering quality risks, and improves risk identification precision and response efficiency.

Inventors

YANG DIMING
CHEN RONGXI
FENG WENHUI

Assignees

广州中天工程检测服务有限公司

Dates

Publication Date: 20260508
Application Date: 20260127

Claims (10)

1. The construction engineering quality risk prediction system based on big data and artificial intelligence is characterized by comprising the following components: The system comprises a data acquisition and integration module, a data analysis module and a data analysis module, wherein the data acquisition and integration module is used for acquiring original engineering data from a plurality of heterogeneous data sources in real time or near real time, and cleaning, format conversion and standardization processing are carried out on the original engineering data to generate data to be analyzed in a unified format; The data center and feature engineering module is used for receiving and storing the data to be analyzed and constructing a theme data warehouse facing the engineering quality risk analysis, extracting multidimensional feature vectors from the theme data warehouse based on a preset engineering quality risk knowledge graph, wherein the feature vectors comprise static features, time sequence features and associated features; The intelligent risk prediction model module is used for receiving the multidimensional feature vector, constructing and training a risk prediction model through an integrated learning algorithm and a deep learning algorithm, and outputting a prediction probability and a risk level aiming at a specific quality risk event; the intelligent risk prediction model module comprises a plurality of sub-models, wherein the sub-models respectively carry out special modeling for different types of quality risks, carry out classification prediction of static and associated features by adopting a gradient lifting tree algorithm, and carry out trend deduction and anomaly detection on the time-series features by adopting a long-period memory network algorithm; The risk visualization and decision support module is used for mapping the prediction probability and the risk level into a visualized risk thermodynamic diagram, a trend curve and an early warning signal, and automatically generating a structured early warning report containing a risk position, a cause analysis and a treatment suggestion when the risk level exceeds a preset threshold; The mobile terminal early warning and cooperative disposal module is used for distributing the structured early warning report to mobile terminals of relevant responsible persons in real time through a message pushing service, receiving disposal feedback information from the mobile terminals, forming a closed-loop management flow of risk early warning, task dispatch, on-site disposal and result feedback, and integrating a geofence and personnel positioning function of the mobile terminal early warning and cooperative disposal module, so that the early warning information can be accurately contacted with the on-site responsible persons.
2. The construction engineering quality risk prediction system based on big data and artificial intelligence according to claim 1 is characterized in that an internet of things sensor network in the data acquisition and integration module is deployed at a key position of a construction site, sensor types comprise strain gauges, inclinometers, temperature and humidity sensors, pressure sensors and image acquisition equipment, and the data adapter cluster supports MQTT, OPC UA, RESTful API and database direct connection with various data exchange protocols.
3. The construction engineering quality risk prediction system based on big data and artificial intelligence according to claim 1 is characterized in that a subject data warehouse constructed by a data center and a characteristic engineering module is organized by adopting a star model, a fact table records monitoring index values taking time stamps and space positions as dimensions, a dimension table comprises engineering stage dimensions, responsibility unit dimensions, component type dimensions and risk type dimensions, the characteristic engineering process specifically comprises the steps of carrying out missing value filling and smooth denoising processing on original monitoring data, extracting mean values, variances, slopes and periodic characteristics from time sequence data based on a sliding window technology, mining the association strength among entities through a graph traversal algorithm based on an engineering quality risk knowledge graph, and quantifying the association strength into association characteristic values.
4. The system for predicting the quality risk of construction engineering based on big data and artificial intelligence according to claim 1, wherein the objective function of the gradient lifting tree sub-model in the intelligent risk prediction model module is defined as a weighted sum of minimized prediction error and model complexity, and the specific formula is: , Wherein, the Representing a loss function; Representative sample Is a real risk tag of (1); Representative model pair sample Is used for predicting the probability of (1); is applied to the first Decision tree Is a regularization term of (2); is the total number of decision trees in the gradient lifting process.
5. The construction engineering quality risk prediction system based on big data and artificial intelligence according to claim 1, wherein the long-term memory network sub-model in the intelligent risk prediction model module is used for processing continuous time sequence data monitored by a sensor, a cell state update mechanism can effectively capture long-term dependency, and a calculation formula of a gating unit is as follows: , , , , , , Wherein, the , , Representing a forget gate, an input gate and an output gate respectively, In the state of a cell, the cell is in a state of being, In order to be in a hidden state, Is that The input characteristics of the time of day, And As a parameter of the model, it is possible to provide, The function is activated for sigmoid.
6. The system for predicting the quality risk of construction engineering based on big data and artificial intelligence according to claim 1, wherein the content of the structured early warning report generated by the risk visualization and decision support module at least comprises a risk event identifier, a three-dimensional coordinate or two-dimensional plan marking of a risk occurrence position, a risk level, a confidence level, a key index for triggering early warning and a historical change curve thereof, a most probable cause deduced based on a knowledge map, and a recommended treatment measure list matched from a historical treatment case library.
7. A construction engineering quality risk prediction method based on big data and artificial intelligence is characterized by comprising the following specific steps: S110, acquiring multi-source heterogeneous original engineering data in parallel from an internet-of-things sensor network, a building information model platform, a project management system and a material supply chain database through a data adapter cluster deployed in a data acquisition and integration module; S120, cleaning the original engineering data in a data acquisition and integration module, removing abnormal values and repeated records, and carrying out unified coding and unit conversion on data in different formats to generate a standardized data stream to be analyzed; S130, inputting the data stream to be analyzed into a data center and a feature engineering module, storing the data stream to be analyzed into a corresponding fact table and a dimension table of a subject data warehouse according to a preset data model, executing feature engineering based on an engineering quality risk knowledge graph, and extracting multidimensional feature vectors containing static features, time sequence features and associated features; S140, inputting the multidimensional feature vector into an intelligent risk prediction model module which is trained in advance, respectively carrying out fusion calculation on features by a gradient lifting tree sub-model and a long-period memory network sub-model in the model module, outputting prediction probabilities for various quality risks, and dividing risk grades according to probability values; S150, in a risk visualization and decision support module, performing visualization rendering on the risk level and the prediction result to generate a risk thermodynamic diagram and a trend analysis chart, and when the risk level exceeds a preset threshold, automatically triggering early warning and generating a structured report containing cause analysis and treatment suggestions; And S160, pushing the structured early warning report to a mobile terminal of a relevant responsible person through a mobile terminal early warning and cooperative treatment module, and tracking and recording treatment feedback information to complete a closed loop process of identifying the risk from the treatment verification.
8. The method for predicting the quality risk of construction engineering based on big data and artificial intelligence according to claim 7, wherein in step S120, the cleaning process is performed by adopting outlier rejection based on a statistical three-sigma principle and repeated record deduplication based on a time stamp and a data source identifier, and the unified coding and unit conversion are performed according to a preset engineering data standardization dictionary.
9. The method for predicting the quality risk of construction engineering based on big data and artificial intelligence according to claim 7, wherein in step S130, the sliding window technique is used for extracting the time sequence features from the time sequence data, and specifically comprises sliding a sliding window with a preset length and a step length on the time sequence data, and calculating the mean value, variance and slope obtained by linear fitting of the data in each window.
10. The method for predicting the quality risk of construction engineering based on big data and artificial intelligence according to claim 7, wherein in step S160, the pushing process integrates the functions of geofence and personnel positioning, and when the pre-warning is related to a specific construction area, pre-warning information is preferentially pushed to the mobile terminals of responsible persons currently located within the range of the geofence.

Description

Construction engineering quality risk prediction method and system based on big data and artificial intelligence Technical Field The invention belongs to the technical field of data processing, and particularly relates to a construction engineering quality risk prediction method and system based on big data and artificial intelligence. Background Along with the digital transformation acceleration promotion of the construction industry, the quality risk management and control of construction engineering becomes a core link for guaranteeing engineering safety and sustainable development. Modern engineering construction involves a large amount of heterogeneous multi-source data, including construction process records, environmental monitoring information, material supply chain data, BIM model parameters and real-time state indexes collected by various Internet of things sensors, which together form basic resources for engineering quality assessment and risk identification. However, the current industry generally lacks the capability of efficient integration and deep mining of the data, so that quality risk identification still highly depends on manual experience judgment and staged field inspection, and systematic and prospective risk early warning and intervention are difficult to realize. The construction engineering quality risk prediction technology based on big data and artificial intelligence aims at opening up an information island in the whole life cycle of a project by constructing a unified data medium-level architecture, fusing structured and unstructured data, and carrying out dynamic modeling and trend deduction on key quality indexes by utilizing a machine learning and time sequence modeling method. The technical direction focuses on automatically extracting risk features from complex, high-dimensional and time-varying engineering data and identifying potential quality hidden danger modes, so that an intelligent risk assessment and decision response mechanism is supported. The method has the obvious defects in the aspect of coping with the construction engineering quality risk prediction, that firstly, the data integration capability is weak, heterogeneous data from multiple channels such as a BIM platform, internet of things equipment, a project management system and the like are difficult to collect efficiently, so that information fragmentation is serious, secondly, the analysis model generalization capability is limited, most systems still adopt a rule engine or static threshold value to judge and cannot adapt to dynamic change rules under different engineering scenes, thirdly, risk early warning is lagged and lack of interpretability, a closed loop logic link from original data to risk level to processing advice is not established, and finally, a real-time pushing and cooperative processing mechanism facing a mobile terminal is lacking, so that the risk response efficiency is low, and the requirement of modern engineering fine management is difficult to meet. These problems severely restrict the improvement of the engineering quality from 'post deviation correction' to 'pre prevention', and an integrated risk prediction system integrating big data management and intelligent algorithm driving is needed to solve the problems. Disclosure of Invention The invention aims to make up the defects of the prior art, provides a construction engineering quality risk prediction method and a construction engineering quality risk prediction system based on big data and artificial intelligence, and can effectively solve the problems in the background art. In order to achieve the above purpose, the invention adopts the technical proposal that, A construction engineering quality risk prediction system based on big data and artificial intelligence comprises the following components: The system comprises a data acquisition and integration module, a data analysis module and a data analysis module, wherein the data acquisition and integration module is used for acquiring original engineering data from a plurality of heterogeneous data sources in real time or near real time, and cleaning, format conversion and standardization processing are carried out on the original engineering data to generate data to be analyzed in a unified format; The data center and feature engineering module is used for receiving and storing the data to be analyzed and constructing a theme data warehouse facing the engineering quality risk analysis, extracting multidimensional feature vectors from the theme data warehouse based on a preset engineering quality risk knowledge graph, wherein the feature vectors comprise static features, time sequence features and associated features; The intelligent risk prediction model module is used for receiving the multidimensional feature vector, constructing and training a risk prediction model through an integrated learning algorithm and a deep learning algorithm, and outputting a prediction probabili