Intelligent construction novel industrial trend prediction and resource optimization method and system

Abstract

The invention discloses an intelligent construction novel industrial trend prediction and resource optimization method and system, and relates to the technical field of intelligent construction, and the method comprises the steps: collecting construction data, and carrying out the preprocessing of the data. On the basis of a GIS technology, surrounding construction data are collected, a CIM dynamic simulation library is constructed, and multi-stage simulation change signals are generated in combination with a layered self-adaptive dynamic modeling method. And performing trend prediction on the multi-stage simulation change signal, comparing the trend with actual construction data, adjusting a construction plan and optimizing resource allocation. According to the invention, intelligentization, precision and high efficiency of construction management are realized. The method has remarkable technical advantages in the aspects of construction prediction, plan adjustment, resource optimization and the like. Through the combination of multi-level modeling and an intelligent scheduling system, the construction efficiency is improved, the adaptability and sustainability of construction management are enhanced, and a more scientific and reasonable construction optimization scheme is provided for the intelligent construction industry.

Description

Technical Field

[0001] This invention relates to the field of intelligent construction technology, specifically to a method and system for predicting new industrialization trends and optimizing resources in intelligent construction. Background Technology

[0002] In recent years, with the informatization, digitalization, and intelligentization of the construction industry, intelligent construction technology has gradually become an important means of construction project management. The traditional construction industry relies on manual experience for construction scheduling and resource allocation, resulting in problems such as information lag, low resource utilization, and unstable construction progress. To improve construction efficiency, reduce costs, and ensure project quality, technologies such as Building Information Modeling (BIM), Geographic Information Systems (GIS), and Computer Integrated Manufacturing (CIM) are widely used to achieve digital management throughout the entire construction lifecycle. Furthermore, with the continuous maturation of technologies such as the Internet of Things (IoT), big data analytics, and artificial intelligence (AI), intelligent construction scheduling, construction process monitoring, and resource optimization have become research hotspots. Intelligent construction combines sensing technology, deep learning algorithms, and reinforcement learning methods, enabling construction companies to monitor construction status in real time, predict construction progress, and optimize construction resource scheduling, thereby improving the scientific nature of project management and the accuracy of construction decisions. However, despite the significant progress made in intelligent construction technology, current construction management still faces many challenges.

[0003] While existing intelligent construction methods have leveraged BIM, GIS, and data analytics to improve construction management, several key issues remain. First, the ability to acquire and preprocess construction data is limited. Current construction data acquisition systems suffer from high data noise, high sensor measurement errors, and weak anomaly identification capabilities, impacting the accuracy of subsequent modeling and optimization. Second, construction trend prediction models lack multi-level dynamic modeling capabilities. Traditional time-series prediction methods based on historical data struggle to adapt to dynamic changes in the construction environment, resulting in low accuracy in construction progress prediction. Third, existing construction scheduling optimization methods are relatively simplistic, mostly employing rule-based static scheduling methods that cannot respond in real-time to fluctuations in construction progress and uncertainties in resource consumption, hindering intelligent adjustments to construction tasks. Fourth, existing construction resource optimization methods also have limitations. Traditional construction resource allocation is often based on static rules, failing to adequately consider real-time data feedback, leading to unreasonable resource allocation and difficulty in optimizing construction efficiency. In summary, existing intelligent construction technologies have not yet fully utilized the dynamic modeling method combining GIS and CIM, and lack accurate prediction of multi-level simulated change signals and adaptive optimization of construction resources. There is an urgent need to propose more accurate, intelligent, and efficient methods for predicting construction trends and optimizing resources in order to improve the scientific nature of construction management and the level of intelligence in engineering implementation. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by this invention is that existing construction trend prediction methods suffer from low accuracy in acquiring construction data, insufficient accuracy in predicting construction progress, and limited ability to optimize construction scheduling. It also addresses the question of how to improve the efficiency of construction resource optimization and allocation through intelligent algorithms.

[0006] To address the aforementioned technical problems, this invention provides the following technical solution: a method for predicting industry trends and optimizing resources in intelligent construction new industrialization, comprising collecting construction data and preprocessing the data. Based on GIS technology, surrounding construction data is collected, a CIM dynamic simulation library is constructed, and multi-level simulated change signals are generated using a hierarchical adaptive dynamic modeling method. Trend prediction is performed on the multi-level simulated change signals and compared with actual construction data to adjust the construction plan and optimize resource allocation. Constructing the CIM dynamic simulation library includes establishing a construction status dataset based on historical construction data, classifying it according to construction category, resource allocation, and construction environment characteristics, performing similarity analysis using unsupervised learning methods, and classifying construction scenarios using density-based spatial clustering to form a construction condition classification model. A construction parameter correlation model is constructed, calculating the probability distribution of construction status based on Bayesian inference to predict changes in construction progress under different working conditions. Simulated change signals of the construction process are generated using dynamic modeling methods, and optimization and adjustment are performed using reinforcement learning methods, where construction progress deviation is used as the reward signal for reinforcement learning, and a deep Q-network training and adjustment strategy is used to gradually approximate the predicted results to the actual construction progress. Anomaly detection is performed on the generated simulated change signals. A variational autoencoder is used to learn the potential distribution of construction data, identify abnormal data points, and calculate their deviations. When the deviation exceeds a set threshold, an automatic correction mechanism is triggered to fill in the missing data. A dynamic adjustment model is constructed based on the optimized simulated change signals. The parameters of construction variable factors are optimized through dynamic weighted regression, and the optimal allocation of construction resources is achieved by combining resource scheduling rules.

[0007] As a preferred embodiment of the intelligent construction new industrialization trend prediction and resource optimization method described in this invention, the collection of construction data includes real-time monitoring of the construction site environment, equipment status, and personnel operation information using multi-source data acquisition equipment. Environmental data at the construction site includes temperature, humidity, and wind speed in the construction area. Equipment status data includes the operating status of the construction equipment. Personnel operation information includes the location information of the construction personnel.

[0008] As a preferred embodiment of the intelligent construction new industrialization trend prediction and resource optimization method described in this invention, the data preprocessing includes: using a Kalman filter algorithm to denoise the data and reduce measurement errors; identifying extreme data using an anomaly detection algorithm; setting corresponding first thresholds for different data; and triggering a data correction mechanism if the threshold is exceeded.

[0009] As a preferred embodiment of the intelligent construction new industrialization industry trend prediction and resource optimization method described in this invention, the step of collecting surrounding construction data based on GIS technology includes collecting construction progress, equipment operating status, and resource consumption, evaluating future construction trends through a time-series prediction model, and calculating adjustment factors based on construction deviations.

[0010] As a preferred embodiment of the intelligent construction new industrialization trend prediction and resource optimization method described in this invention, the method of generating multi-level simulated change signals by combining hierarchical adaptive dynamic modeling includes: constructing a CIM dynamic simulation library; establishing a construction status dataset based on historical construction data; classifying the dataset; dividing it into multiple subsets according to construction category, resource allocation, and construction environment characteristics; performing similarity analysis on each subset using an unsupervised learning method; and classifying data from similar construction scenarios using a density-based spatial clustering method to form a construction condition classification model. For each construction condition category, a construction parameter correlation model is established, using construction progress, construction equipment utilization rate, resource consumption rate, and construction environmental factors as variables to construct a construction state probability model. The construction state probability model calculates the probability distribution of each construction state based on a Bayesian inference method to predict the possibility of changes in construction progress under different conditions. Based on the construction parameter correlation model, a dynamic modeling method is used to generate simulated change signals of the construction process. These simulated change signals are optimized and adjusted using reinforcement learning, with construction progress deviation as the reward signal. A deep Q-network method is used to train the adjustment strategy. During training, the adjustment reward value is calculated based on construction progress deviations in historical data, and the optimal strategy is used to select adjustment modeling parameters, gradually bringing the predicted construction progress closer to the actual construction progress. Anomaly detection is performed on the generated simulated change signals. A probabilistic generation model is used to estimate the latent distribution of construction data, and detected anomalies are corrected. Anomaly detection employs a variational autoencoder method, which learns the latent feature distribution of construction data to identify anomaly points and calculates the probability deviation of anomaly points relative to historical data. When the probability deviation exceeds a set threshold, an automatic correction mechanism is triggered to automatically fill in missing data. Based on the optimized simulated change signals, a dynamic adjustment model is constructed, incorporating construction variable factors into the adjustment model. The parameters of the construction variable factors are optimized using a dynamic weighted regression method, and combined with resource scheduling rules, the optimal allocation of construction resources is achieved.

[0011] As a preferred embodiment of the intelligent construction new industrialization trend prediction and resource optimization method described in this invention, the step of predicting the trend of multi-level simulated change signals and comparing them with actual construction data includes: the trend prediction adopts an ARIMA-LSTM hybrid model, combining time series analysis and deep learning, to calculate the future construction progress change trend. If the prediction result shows that a certain construction task is delayed by more than a preset time, the construction plan is re-optimized using an integer programming method.

[0012] As a preferred embodiment of the intelligent construction new industrialization trend prediction and resource optimization method described in this invention, the optimized resource allocation includes: adjusting the construction plan based on the prediction results and optimizing resource allocation through a real-time monitoring and feedback system. RFID sensors are used to monitor the supply of building materials; if material inventory is detected to be below the minimum safe level, an intelligent procurement and scheduling system is automatically triggered to optimize the material transportation plan. The status of construction equipment is remotely monitored through IoT sensors; when the equipment operating load exceeds a safe threshold, the allocation of construction tasks is automatically adjusted.

[0013] Another objective of this invention is to provide a smart construction new industrialization trend prediction and resource optimization system. This system can collect surrounding construction data based on GIS technology, construct a CIM dynamic simulation library, and generate multi-level simulated change signals by combining a hierarchical adaptive dynamic modeling method. This solves the problems of low accuracy in construction data acquisition, insufficient accuracy in construction progress prediction, limited construction scheduling optimization capabilities, and how to improve the efficiency of construction resource optimization and allocation through intelligent algorithms in existing construction trend prediction methods.

[0014] As a preferred embodiment of the intelligent construction new industrialization trend prediction and resource optimization system described in this invention, it includes: a data acquisition and preprocessing module, an intelligent construction data analysis and dynamic modeling module, and a trend prediction and resource optimization module. The data acquisition and preprocessing module is used to collect construction data and preprocess it. The intelligent construction data analysis and dynamic modeling module is used to collect surrounding construction data based on GIS technology, construct a CIM dynamic simulation library, and generate multi-level simulated change signals using a hierarchical adaptive dynamic modeling method. The trend prediction and resource optimization module is used to predict the trends of the multi-level simulated change signals and compare them with actual construction data to adjust the construction plan and optimize resource allocation.

[0015] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement a method for predicting and optimizing resource allocation in the new industrialization of intelligent construction.

[0016] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of a method for predicting and optimizing the trend of intelligent construction of new industrialization.

[0017] The beneficial effects of this invention are as follows: The intelligent construction new industrialization trend prediction and resource optimization method provided by this invention collects construction data and performs preprocessing. It ensures the comprehensiveness and accuracy of the data by real-time monitoring of construction environment data, equipment status, and personnel operation information. Noise is eliminated through the Kalman filter algorithm, improving the stability and reliability of the measurement data. This improves the quality and accuracy of construction data, reducing management deviations caused by data errors. It ensures the stability and continuity of construction status data, providing reliable input data for intelligent construction decisions. It avoids construction scheduling errors due to missing or incorrect data, improving the accuracy and intelligence level of construction management.

[0018] By collecting surrounding construction data using GIS technology, a CIM dynamic simulation library is constructed. Combined with a hierarchical adaptive dynamic modeling method, multi-level simulated change signals are generated, enabling construction management to evolve from static rules to dynamic adaptive adjustments, thus improving the accuracy of intelligent construction. Through multi-level signal prediction, deviations in construction progress can be identified in advance, and construction plans can be dynamically adjusted to reduce the risk of construction delays. This also improves the rationality of construction resource allocation, making construction planning more scientific and flexible.

[0019] By predicting trends in multi-level simulated change signals and comparing them with actual construction data, construction plans can be adjusted and resource allocation optimized, improving the accuracy of construction trend prediction and making construction management more scientific and intelligent. Dynamic adjustments to construction plans improve the efficiency of construction task execution and reduce schedule delays. Optimized construction resource allocation reduces construction costs, improves resource utilization efficiency, and ensures smooth construction progress. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 The overall flowchart of the intelligent construction new industrialization industry trend prediction and resource optimization method provided in the first embodiment of the present invention is shown. Detailed Implementation

[0022] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0023] Example 1, referring to Figure 1 As an embodiment of the present invention, a method for predicting and optimizing the industrial trend of intelligent construction and new industrialization is provided, comprising: S1: Collect construction data and preprocess the data.

[0024] Multi-source data acquisition equipment is used to monitor the construction site's environment, equipment status, and personnel operation information in real time. Environmental data includes temperature, humidity, and wind speed in the construction area. Equipment status data includes the operating status of construction equipment. Personnel operation information includes the location information of construction personnel. A Kalman filter algorithm is used to denoise the data, reducing measurement errors. An anomaly detection algorithm identifies extreme data, and a corresponding first threshold is set for different data points; if the threshold is exceeded, a data correction mechanism is triggered.

[0025] It should be noted that a preferred approach to data preprocessing includes, firstly, denoising using a Kalman filter algorithm to reduce measurement errors and improve data stability. The prediction error covariance matrix is ​​used to dynamically correct outlier data, improving signal accuracy. Subsequently, the data enters an anomaly detection process, where a statistical anomaly identification algorithm analyzes various data types and sets first thresholds for different data types. If data exceeds the threshold, a data correction mechanism is triggered, employing time-series interpolation to ensure data integrity and accuracy, providing reliable input for subsequent construction status analysis and modeling.

[0026] Furthermore, a preferred embodiment of the first threshold specifically includes the following: In this invention, the first threshold for the environmental conditions, equipment status, and personnel operation information at the construction site is set according to industry standards and engineering experience to ensure the accuracy of data anomaly detection and reduce the possibility of false alarms and missed alarms. Specific threshold settings include: a temperature threshold set to -10℃ to 50℃; temperatures below -10℃ may cause abnormal solidification of construction materials, while temperatures above 50℃ may affect the health of construction personnel and the stability of equipment operation. Data exceeding this range is considered abnormal. A humidity threshold set to 10% to 95%; humidity below 10% may affect concrete curing and construction accuracy, while humidity above 95% may cause materials to absorb water and deform or affect the safety of electrical equipment. A wind speed threshold set to 0m / s to 20m / s; wind speeds exceeding 20m / s may affect the safety of high-altitude operations, especially hoisting and welding operations. The equipment vibration frequency threshold is set to ±15% of the rated vibration value. During normal operation, the vibration frequency of construction equipment (such as pile drivers, cranes, and concrete mixers) will fluctuate within ±15% of the rated vibration value. If the vibration frequency exceeds this range, it may indicate equipment malfunction or overload. The equipment energy consumption threshold is set to ±10% of the rated power. The power consumption of construction equipment typically fluctuates within ±10% of the rated power. Abnormal increases or decreases in power consumption may indicate energy efficiency issues, excessive load, or malfunction. The equipment temperature threshold is set to 30℃ to 90℃. The safe operating temperature for equipment is generally between 30℃ and 90℃. Excessive temperature may cause overheating, affecting lifespan or causing safety problems; excessively low temperature may affect the efficiency of the hydraulic system. The construction worker heart rate threshold is set to 50 bpm to 160 bpm. The normal heart rate range for construction workers is generally 50 to 160 bpm (heartbeats per minute). A heart rate below 50 bpm may indicate health problems, while a heart rate above 160 bpm may indicate excessive workload or a risk of heatstroke. The threshold for variation in the number of steps taken by construction workers is set at ±30% of the average number of steps. The number of steps taken by construction workers typically varies depending on the work content, but if the variation exceeds 30% of the average, it may indicate an abnormal situation, such as prolonged inactivity (potential accident) or overwork (health risk). The threshold for location data drift is set at ±2m. Under normal circumstances, the accuracy of GPS location data for construction workers is within ±2m. If it exceeds this range, there may be positioning errors or data anomalies.

[0027] It should be noted that S1, based on multi-source data acquisition technology, monitors the environment, equipment status, and personnel operation information at the construction site in real time, and improves data quality through data preprocessing. First, a Kalman filter algorithm is used to remove measurement noise, ensuring data stability. Then, an anomaly detection algorithm identifies abnormal data and sets a first threshold for different types of data. If data exceeds the threshold range, a data correction mechanism is triggered, using time series interpolation to correct the abnormal data, ensuring data integrity and reliability. This step improves the accuracy and stability of construction data, reduces data deviations caused by noise and outliers, and enhances the reliability of construction progress prediction and CIM modeling. Simultaneously, threshold setting prevents false alarms and missed alarms, making construction management more intelligent, ensuring the safety of construction personnel, stable equipment operation, and optimizing resource scheduling. The construction environment is complex, and sensor data is easily affected by external interference; directly using unprocessed data may lead to model misjudgments. Kalman filtering can dynamically adjust prediction errors and improve measurement accuracy, while anomaly detection and threshold setting can effectively identify equipment failures, changes in the construction environment, and personnel safety risks. The data correction mechanism ensures data continuity, providing a reliable foundation for subsequent intelligent analysis.

[0028] S2: Based on GIS technology, collect surrounding construction data, construct a CIM dynamic simulation library, and generate multi-level simulated change signals using a hierarchical adaptive dynamic modeling method. Collect data on construction progress, equipment operating status, and resource consumption, assess future construction trends through a time-series prediction model, and calculate adjustment factors based on construction deviations.

[0029] It should be noted that a preferred approach for assessing future construction trends using a time-series forecasting model and calculating adjustment factors based on construction deviations specifically includes: utilizing GIS technology to collect data on construction progress, equipment operating status, and resource consumption; constructing a time-series dataset; and then employing a time-series forecasting model to assess future construction trends and calculate adjustment factors based on construction deviations. (Construction progress) Equipment status and resource consumption The relationship that changes over time is expressed as follows: , in, This represents the construction status variable. This indicates measurement error. A nonlinear function representing the evolution of the construction state.

[0030] Predicting future trends using an autoregressive integral moving average model: , in, Indicates time Predicted construction status. express The construction status. This indicates the number of autoregressive terms, representing how much the current state is influenced by past states. This indicates the number of terms in the moving average, representing how much of the past error has influenced the current state. Represents the autoregressive coefficient. This represents the moving average coefficient. express Noise at any given moment. This indicates a trend, representing the long-term trend of construction progress. This represents the noise term. After predicting construction over-potential, the construction deviation is calculated. : , in, This indicates the construction progress at the previous moment.

[0031] like If the set threshold is exceeded, the adjustment factor will be triggered. : , in, These represent weighting coefficients, which respectively measure the impact of construction deviations, equipment condition, and resource consumption on the adjustment factor. Adjustment Factor Used for subsequent modeling and optimization.

[0032] The construction of the CIM dynamic simulation library includes: establishing a construction status dataset based on historical construction data; classifying the dataset; dividing it into multiple subsets according to construction category, resource allocation, and construction environment characteristics; using unsupervised learning methods to perform similarity analysis on each subset; and using density-based spatial clustering methods to classify data with similar construction scenarios to form a construction condition classification model.

[0033] It should be noted that a preferred approach to constructing a CIM dynamic simulation library specifically includes, firstly, classifying historical construction data within the CIM dynamic simulation library to improve modeling accuracy. The data is divided into multiple subsets based on construction category, resource allocation, and construction environment characteristics. Unsupervised learning methods are used for similarity analysis, and density-based spatial classification (DBSCAN) is employed for data categorization, forming a construction condition classification model. The core computation of DBSCAN is represented as follows: , in, express The neighborhood of a point. The distance between data points is represented by the neighborhood radius; data outside this range is considered noise. This represents the construction status dataset. Indicates the centrality of the dataset and Relevant construction status data points. The neighborhood radius determines whether a data point belongs to the core, thus affecting the density clustering results of data classification. After classification, the construction status data of each category forms an independent model input, providing support for subsequent modeling.

[0034] For each construction condition category, a construction parameter correlation model is established, taking construction progress, construction equipment utilization rate, resource consumption rate, and construction environmental factors as variables to construct a construction state probability model. The construction state probability model calculates the probability distribution of each construction state based on the Bayesian inference method and predicts the possibility of changes in construction progress under different conditions.

[0035] It should be noted that a preferred approach to establishing a construction parameter correlation model specifically includes, for each construction condition category, establishing a construction parameter correlation model, defining construction state variables, including construction progress, equipment utilization rate, resource consumption rate, and construction environmental factors, and using Bayesian inference to calculate the probability distribution of construction states. The construction state probability model is expressed as: , in, Indicates that under known construction variables Under what circumstances, construction status The probability of occurrence. Represents construction variables In a specific construction state The probability of it occurring. This represents the prior probability, i.e., the probability when no prior observation is made. Before, The likelihood of it happening. This represents the normalization constant.

[0036] In the modeling process, the probability density of the construction state is represented as: , in, This represents the number of different construction status categories in the construction status classification, with each category corresponding to a probability density function. This indicates the weight of the construction variable. This represents the probability density function of the construction state. The construction state probability model is used to predict the likelihood of changes in construction progress under different working conditions, providing a basis for generating simulation change signals.

[0037] Based on the construction parameter association model, a dynamic modeling method is used to generate simulated change signals of the construction process. The simulated change signals are optimized and adjusted based on the reinforcement learning method. The construction progress deviation is used as the reward signal for reinforcement learning. The deep Q-network method is used to train the adjustment strategy. During the training process, the adjustment reward value is calculated based on the construction progress deviation in historical data. The optimal strategy is used to select and adjust the modeling parameters so that the predicted construction progress gradually approaches the actual construction progress.

[0038] It should be noted that the preferred scheme for generating simulated change signals of the construction process using dynamic modeling methods, and optimizing these simulated change signals based on reinforcement learning methods, specifically includes: optimizing the simulated change signals using reinforcement learning methods based on a construction parameter correlation model; using construction progress deviation as the reward signal for reinforcement learning; and using a deep Q-network (DQN) for policy optimization. The loss function of reinforcement learning is defined as: , in, Indicates the state Take action The Q-value is the payoff of the strategy. This indicates the current construction status. This indicates an adjustment strategy. This represents the discount factor. This indicates a reward signal, calculated based on construction progress deviations: , in, This indicates the actual construction progress. This indicates the predicted construction progress. After multiple training sessions, it can generate more accurate simulated change signals based on historical construction data.

[0039] Anomaly detection is performed on the generated simulated change signals. A probabilistic generation model is used to estimate the potential distribution of construction data. The detected abnormal signals are corrected. Anomaly detection adopts the variational autoencoder method. By learning the potential feature distribution of construction data, abnormal data points are identified and the probability deviation of abnormal data points relative to historical data is calculated. When the probability deviation exceeds a set threshold, an automatic correction mechanism is triggered to automatically fill in the missing data.

[0040] It should be noted that a preferred scheme for anomaly detection specifically includes performing anomaly detection on the generated simulated change signal, using a variational autoencoder (VAE) to learn the latent distribution of the construction data, and the criteria for judging anomaly data being: , in, This represents the potential distribution of construction data. Representing latent variables, it represents the feature space of the construction state (the latent structure of the construction data). Representing a given latent variable The probability distribution of construction data at that time. This represents the prior distribution of the latent variable. If If the data falls below a set threshold, it is considered abnormal. Abnormal data points are corrected through an automatic signal correction mechanism, and historical similar data is used to complete the data to ensure data continuity.

[0041] Furthermore, in this invention, for the detection of anomalies in construction data, the set probability deviation threshold is 2.5 times the standard deviation (σ). That is, when the probability of an abnormal data point deviates from the mean of historical data by more than 2.5σ, the system determines that the data is abnormal and triggers an automatic correction mechanism.

[0042] Based on the optimized simulated change signal, a dynamic adjustment model is constructed, incorporating construction variable factors into the adjustment model. The parameters of the construction variable factors are optimized through a dynamic weighted regression method, and the optimal allocation of construction resources is achieved by combining resource scheduling rules.

[0043] It should be noted that a preferred approach to constructing a dynamic adjustment model specifically includes optimizing the parameters of construction variable factors and combining this with resource scheduling rules to achieve optimal allocation of construction resources. The optimization of construction variable factors employs a dynamic weighted regression method. , in, This represents the optimized construction variables. This indicates the number of construction variable factors, such as construction progress, equipment load, and resource consumption. This represents a dynamic weighting factor that is dynamically adjusted based on historical data. Indicates the first One construction variable.

[0044] Based on resource scheduling rules, a resource allocation optimization equation is constructed with the goal of minimizing the total cost of construction resources while meeting construction schedule requirements. , in, This indicates the number of construction tasks. This indicates the cost of construction resources consumed. Indicates the assignment of construction tasks The amount of resources. Ultimately, through dynamic weight adjustment and optimized scheduling, efficient allocation of construction resources is achieved, thereby improving construction efficiency.

[0045] It should be noted that S2 utilizes GIS technology to collect data on construction progress, equipment status, and resource consumption, and combines this with a CIM dynamic simulation library for modeling and optimization. First, a time-series prediction model is used to assess construction trends, and adjustment factors are calculated based on construction deviations. Then, a construction state classification model is constructed, classifying construction scenarios based on unsupervised learning methods, and establishing a construction state probability model using Bayesian inference. Further, reinforcement learning methods are used to optimize simulated change signals, adjusting construction states through a deep Q-network (DQN) to make the prediction results more accurate. Finally, variational autoencoders (VAEs) are used for anomaly detection, and dynamic weighted regression is used to optimize construction variable factors, achieving optimal resource scheduling. This step can dynamically predict construction trends and optimize construction scheduling through reinforcement learning, improving the accuracy of construction progress prediction. Simultaneously, the anomaly detection mechanism can effectively identify data anomalies, ensuring data quality, optimizing resource allocation, improving construction efficiency, and reducing project costs. The construction environment is complex, and the data is highly time-varying; therefore, it is necessary to combine time-series prediction and reinforcement learning for optimization. Unsupervised learning is used for classification to improve the accuracy of construction status modeling, while Bayesian inference can infer the probability distribution of construction status based on different variables. Furthermore, VAEs can effectively identify abnormal data, ensuring the stability of construction data and ultimately improving the intelligence level of resource scheduling and construction management.

[0046] S3: Perform trend prediction on multi-level simulated change signals and compare them with actual construction data to adjust construction plans and optimize resource allocation.

[0047] Trend prediction employs an ARIMA-LSTM hybrid model, combining time series analysis and deep learning to calculate future construction progress trends. If the prediction indicates that a construction task is delayed beyond a preset time, the construction plan is re-optimized using integer programming. Construction plans are adjusted based on the prediction results, and resource allocation is optimized through a real-time monitoring and feedback system. RFID sensors monitor the supply of building materials; if material inventory falls below the minimum safety level, an intelligent procurement and scheduling system is automatically triggered to optimize material transportation plans. Construction equipment status is remotely monitored via IoT sensors; when equipment operating load exceeds a safety threshold, construction task allocation is automatically adjusted.

[0048] It should be noted that a preferred scheme for calculating future construction progress trends specifically includes, to improve the accuracy of construction progress prediction, a hybrid prediction method based on an autoregressive integral moving average model and a long short-term memory network. The time series analysis method is mainly used to extract the long-term trend of construction progress and analyze its periodic and random fluctuation characteristics, while the deep learning method is used to learn nonlinear influencing factors and combine them with historical data for trend prediction. First, the stationarity of construction progress, equipment operating status, and resource consumption data is tested, and the influence of trends is eliminated through differencing, making the data more stable. Then, a time series prediction framework based on an autoregressive model is constructed. By setting a backtracking window for historical data, the dependency of construction progress on several past time steps is calculated, thereby establishing a prediction model. To improve prediction accuracy, this scheme further introduces a long short-term memory network to model the construction data. This method extracts features from historical data through an input layer, hidden layer, and output layer structure, and stores long-term dependency information through memory units. During training, historical construction data is used to optimize model parameters, minimizing prediction errors. Finally, by combining the prediction results of the time series model and the deep learning model, a weighted average method is used to fuse the predictions of the two models in order to improve the accuracy of construction progress prediction.

[0049] It should be noted that a preferred approach to re-optimizing the construction plan using integer programming methods specifically includes the following: when prediction results indicate that the schedule of a construction task is delayed beyond a set time threshold, the system automatically triggers a construction plan adjustment process. First, the affected construction tasks are identified, and their dependencies with other tasks are analyzed to ensure the adjusted plan conforms to construction logic. Then, the priorities of the construction tasks are calculated, and the reordering of construction tasks is optimized based on integer programming methods. During the task adjustment process, the pre-constraints between tasks are analyzed to ensure the adjusted plan still conforms to construction logic. For example, if a task can only begin after another task is completed, this order must still be maintained after adjustment. Second, the availability of construction resources is calculated to ensure the adjusted construction plan does not exceed the capacity of equipment and manpower. Based on the optimal order of construction tasks, the system calculates the task adjustment scheme and generates a new construction plan. During the optimization of the construction plan, the system prioritizes optimizing tasks on the critical path, i.e., adjusting tasks that have the greatest impact on the overall construction schedule. If a delay in a task will cause the entire project to be delayed, additional resources are allocated to that task to ensure its timely completion. The adjustment plan includes increasing the number of construction personnel, optimizing the usage time of construction equipment, or adjusting the parallel execution strategy of construction tasks.

[0050] It should be noted that the preferred solution, which adjusts the construction plan based on the forecast results and optimizes resource allocation through a real-time monitoring and feedback system, specifically includes the following: To ensure efficient allocation of construction resources, this solution employs an intelligent scheduling system based on real-time monitoring. First, radio frequency identification (RFID) sensors are deployed at the construction site to monitor the inventory of building materials in real time. When the inventory of a certain building material falls below a safe level, the system automatically triggers an intelligent procurement scheduling mechanism, calculates the required quantity of materials based on the needs of the construction task, and automatically generates a purchase order. Regarding material transportation, this solution optimizes supply chain management by calculating the optimal transportation route and time schedule to ensure that materials are delivered to the construction site on time. The procurement scheduling system selects the optimal supplier based on the supplier's delivery capacity and monitors the transportation progress through a logistics tracking system to ensure that materials arrive at the construction site on time and avoid construction delays due to material shortages. In terms of construction equipment management, the system uses IoT sensors to monitor the operating status of construction equipment, including equipment load, energy consumption levels, and operating temperature. When the equipment load exceeds a set safety threshold, the system automatically adjusts the allocation of construction tasks, reducing the workload of high-load equipment to prevent equipment failure due to overload. Meanwhile, if the equipment's energy consumption level becomes abnormal, the system will trigger a maintenance alert and schedule equipment repairs to minimize the impact of equipment failure on the construction schedule. Furthermore, the system dynamically monitors the work status of construction personnel and adjusts personnel arrangements according to the needs of the construction tasks. If a task requires additional personnel, the system will automatically allocate available personnel from other tasks to improve construction efficiency. For high-intensity construction tasks, the system will consider the workers' work schedules to ensure that their workload is within a reasonable range, avoiding safety accidents caused by fatigue.

[0051] Furthermore, S3 enhances the intelligence level of construction management by constructing a construction progress prediction model and optimizing construction plans and resource scheduling. First, it extracts long-term trends in construction progress using time series analysis and combines this with deep learning algorithms to improve prediction accuracy. Then, based on the prediction results, it determines whether any construction tasks are delayed and uses integer programming to optimize the sequence of construction tasks and resource allocation, ensuring that critical tasks are completed on time. Finally, it dynamically optimizes construction resources using a real-time monitoring and feedback system, including intelligent procurement of building materials, monitoring of construction equipment load, and personnel allocation, to ensure stable construction progress. This step accurately predicts construction progress and quickly adjusts construction plans when tasks are delayed, avoiding project delays. Simultaneously, the intelligent scheduling system optimizes resource allocation, ensuring the rational use of building materials, construction equipment, and personnel, improving construction efficiency, and reducing costs. In addition, the real-time monitoring and feedback mechanism continuously optimizes construction strategies, improving the flexibility and automation level of construction management. Construction progress is affected by various factors and has a high degree of uncertainty; therefore, it is necessary to combine time series analysis and deep learning for accurate prediction. Construction plan optimization ensures reasonable adjustments through integer programming, while intelligent scheduling systems can dynamically optimize resource allocation, reducing construction delays caused by material shortages, equipment overload, or improper personnel deployment, thereby improving construction efficiency and safety.

[0052] Example 2, an embodiment of the present invention, provides an intelligent construction new industrialization trend prediction and resource optimization system, including a data acquisition and preprocessing module, an intelligent construction data analysis and dynamic modeling module, and a trend prediction and resource optimization module. The data acquisition and preprocessing module collects construction data and preprocesses it. The intelligent construction data analysis and dynamic modeling module collects surrounding construction data based on GIS technology, constructs a CIM dynamic simulation library, and generates multi-level simulated change signals using a hierarchical adaptive dynamic modeling method. The trend prediction and resource optimization module predicts trends in the multi-level simulated change signals and compares them with actual construction data to adjust construction plans and optimize resource allocation.

Claims

1. A method for predicting industry trends and optimizing resources in intelligent construction and new industrialization, characterized in that: include: Collect construction data and preprocess the data; Based on GIS technology, surrounding construction data is collected, a CIM dynamic simulation library is constructed, and multi-level simulated change signals are generated by combining a hierarchical adaptive dynamic modeling method. Trend prediction of multi-level simulated change signals is performed and compared with actual construction data to adjust construction plans and optimize resource allocation; The construction of the CIM dynamic simulation library includes: establishing a construction status dataset based on historical construction data, classifying it according to construction type, resource allocation, and construction environment characteristics; performing similarity analysis using unsupervised learning methods and categorizing construction scenarios using density-based spatial clustering to form a construction condition classification model; constructing a construction parameter association model, calculating the probability distribution of construction status based on Bayesian inference, and predicting changes in construction progress under different conditions; generating simulated change signals of the construction process using dynamic modeling methods and optimizing them using reinforcement learning methods, with construction progress deviation as the reward signal for reinforcement learning, and employing a deep Q-network training and adjustment strategy to gradually approximate the actual construction progress; performing anomaly detection on the generated simulated change signals, using a variational autoencoder to learn the latent distribution of construction data, identifying abnormal data points and calculating their deviations, and triggering an automatic correction mechanism to fill in missing data when the deviation exceeds a set threshold; and constructing a dynamic adjustment model based on the optimized simulated change signals, optimizing construction variable factor parameters through dynamic weight regression, and achieving optimal allocation of construction resources in conjunction with resource scheduling rules.

2. The method for predicting industry trends and optimizing resources in intelligent construction and new industrialization as described in claim 1, characterized in that: The collected construction data includes, Multi-source data acquisition equipment is used to monitor the environment, equipment status, and personnel operation information at the construction site in real time; Environmental data at the construction site includes temperature, humidity, and wind speed in the construction area; equipment status data includes the operating status of construction equipment; and personnel operation information includes the location information of construction personnel.

3. The method for predicting industry trends and optimizing resources in intelligent construction and new industrialization as described in claim 2, characterized in that: The data preprocessing includes, The Kalman filter algorithm is used to denoise the data and reduce measurement errors; Extreme data is identified through anomaly detection algorithms, and a corresponding first threshold is set for different data. If the threshold is exceeded, a data correction mechanism is triggered.

4. The intelligent construction new industrialization industry trend prediction and resource optimization method as described in claim 3, characterized in that: The GIS-based collection of surrounding construction data includes, Data on construction progress, equipment operating status, and resource consumption are collected. Future construction trends are assessed using a time-series forecasting model, and adjustment factors are calculated based on construction deviations.

5. The method for predicting industry trends and optimizing resources in intelligent construction and new industrialization as described in claim 4, characterized in that: The method of generating multi-level simulated change signals by combining hierarchical adaptive dynamic modeling includes... The construction of the CIM dynamic simulation library includes: establishing a construction status dataset based on historical construction data; classifying the dataset; dividing it into multiple subsets according to construction category, resource allocation, and construction environment characteristics; using unsupervised learning methods to perform similarity analysis on each subset; and using density-based spatial clustering methods to classify data with similar construction scenarios to form a construction condition classification model. For each construction condition category, a construction parameter correlation model is established, taking construction progress, construction equipment utilization rate, resource consumption rate, and construction environmental factors as variables to construct a construction state probability model. The construction state probability model calculates the probability distribution of each construction state based on the Bayesian inference method and predicts the possibility of changes in construction progress under different conditions. Based on the construction parameter association model, a dynamic modeling method is used to generate simulated change signals of the construction process. The simulated change signals are optimized and adjusted based on the reinforcement learning method. The construction progress deviation is used as the reward signal for reinforcement learning. The deep Q-network method is used to train the adjustment strategy. During the training process, the adjustment reward value is calculated based on the construction progress deviation in historical data. The optimal strategy is used to select and adjust the modeling parameters so that the predicted construction progress gradually approaches the actual construction progress. Anomaly detection is performed on the generated simulated change signals. A probabilistic generation model is used to estimate the potential distribution of construction data. The detected abnormal signals are corrected. Anomaly detection adopts the variational autoencoder method. By learning the potential feature distribution of construction data, abnormal data points are identified and the probability deviation of abnormal data points relative to historical data is calculated. When the probability deviation exceeds a set threshold, an automatic correction mechanism is triggered to automatically fill in the missing data. Based on the optimized simulated change signal, a dynamic adjustment model is constructed, incorporating construction variable factors into the adjustment model. The parameters of the construction variable factors are optimized through a dynamic weighted regression method, and the optimal allocation of construction resources is achieved by combining resource scheduling rules.

6. The method for predicting industry trends and optimizing resources in intelligent construction new industrialization as described in claim 5, characterized in that: The process of predicting trends in multi-level simulated change signals and comparing them with actual construction data includes... Trend prediction uses an ARIMA-LSTM hybrid model, combining time series analysis and deep learning, to calculate future construction progress trends. If the prediction results show that a certain construction task is delayed beyond the preset time, the construction plan is re-optimized using integer programming.

7. The method for predicting industry trends and optimizing resources in intelligent construction new industrialization as described in claim 6, characterized in that: The optimized resource allocation includes, The construction plan was adjusted based on the forecast results, and resource allocation was optimized through a real-time monitoring and feedback system. RFID sensors are used to monitor the supply of building materials. If the material inventory is detected to be lower than the minimum safety stock, the intelligent procurement and scheduling system will be automatically triggered to optimize the material transportation plan. The status of construction equipment is remotely monitored through IoT sensors, and the allocation of construction tasks is automatically adjusted when the equipment operating load exceeds the safety threshold.

8. A smart construction new industrialization trend prediction and resource optimization system, characterized in that: It includes modules for data acquisition and preprocessing, intelligent construction data analysis and dynamic modeling, and trend prediction and resource optimization. The data acquisition and preprocessing module is used to acquire construction data and preprocess the data. The intelligent construction data analysis and dynamic modeling module is used to collect surrounding construction data based on GIS technology, build a CIM dynamic simulation library, and generate multi-level simulated change signals by combining a hierarchical adaptive dynamic modeling method. The trend prediction and resource optimization module is used to predict the trends of multi-level simulated change signals and compare them with actual construction data to adjust the construction plan and optimize resource allocation.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the intelligent construction new industrialization industry trend prediction and resource optimization method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent construction new industrialization industry trend prediction and resource optimization method as described in any one of claims 1 to 7.

Images