Construction project quality whole-process monitoring method and system based on Internet of Things

Abstract

The invention discloses a construction project quality whole-process monitoring method and system based on the Internet of Things, and relates to the technical field of intelligent supervision. Comprising the following steps: S1, determining a quality monitoring target and requirement, and formulating a quality monitoring scheme based on the Internet of Things; s2, implementing the formulated quality monitoring scheme; s3, performing quality analysis processing on the construction data based on a machine learning algorithm, and performing corresponding quality early warning; and S4, carrying out whole-process collaborative management by utilizing the whole-process collaborative management platform. A large number of sensors are deployed in combination with the Internet of Things technology, engineering quality related data can be obtained in real time in the whole construction process, the data collection speed and precision are high, data processing and analysis are conducted in combination with a machine learning algorithm, and the quality problem in engineering can be rapidly and accurately analyzed; corresponding quality early warning is carried out; and a whole-process collaborative management platform is further established to realize information sharing and collaborative decision among all parties, so that the quality problem processing efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent monitoring technology, and in particular to a method and system for monitoring the whole process of construction project quality based on the Internet of Things. Background Art

[0002] In traditional construction project quality monitoring, data collection mainly relies on manual regular inspections and sampling tests. This method has a time lag and cannot reflect the quality changes of concrete during the pouring process and the initial curing period in real time. In addition, there may be operational errors during manual sampling, resulting in inaccurate test results, making it difficult to fully and truly reflect the quality of the project.

[0003] Traditional methods for monitoring deformation of building structures often use instruments such as total stations for regular measurements. However, this method can only obtain data at discrete time points, and the slight deformation of the structure that may occur between two measurements cannot be detected in time, which may miss the opportunity for early warning of potential quality problems.

[0004] Traditional monitoring often focuses on certain key links or parts in the construction process, making it difficult to achieve comprehensive coverage of the entire construction project. However, factors that are not fully monitored may affect the quality of the foundation and superstructure during the subsequent construction process. Moreover, the quality monitoring data between different construction stages and professional fields are often independent of each other, lacking effective integration and correlation analysis. For example, the quality data of the civil construction stage and the data of the electromechanical installation stage are not organically combined, making it impossible to comprehensively consider the interaction of various factors when evaluating the overall quality of the project, making it difficult to grasp the overall picture of the project quality from a system level, and not conducive to timely discovery and resolution of cross-stage and cross-professional quality problems.

[0005] Most of the data collected in traditional quality monitoring is stored in paper records or simple spreadsheets, and data processing mainly relies on manual statistics and analysis. Faced with a large amount of complex quality data, this processing method is inefficient and prone to human errors. For example, when counting the steel bar welding quality inspection data of multiple construction sites, manual calculation and summary may result in data omissions or incorrect entry, affecting the accuracy of data analysis.

[0006] The existing feedback mechanism for dealing with quality issues is also relatively slow. When a quality problem is discovered, it usually takes a series of cumbersome processes, such as filling out a quality problem report, approval at each level, and communicating rectification requirements, before it can reach the construction site for rectification. This process is time-consuming and may cause the quality problem to further deteriorate during the construction process, increasing the cost of repair and the impact on the progress of the project.

[0007] In order to solve the above problems, the present invention proposes a method and system for monitoring the whole process of construction project quality based on the Internet of Things. Summary of the invention

[0008] The main purpose of the present invention is to provide a method and system for monitoring the quality of construction projects throughout the entire process based on the Internet of Things, which can effectively solve the problems in the background technology.

[0009] In order to achieve the above object, the present invention adopts the following technical solutions: The whole process monitoring method of construction project quality based on the Internet of Things includes the following steps: S1: Clarify the quality monitoring objectives and requirements of various monitoring indicators in construction projects, and formulate a quality monitoring plan based on the Internet of Things; S2: According to the established quality control plan, evaluate, select and install IoT devices at the construction site, build the IoT communication network, and debug the installed equipment and communication network; S3: IoT devices collect construction data in real time during the construction process, perform quality analysis and processing on the construction data based on machine learning algorithms, and issue corresponding quality warnings based on the data analysis results based on the quality warning mechanism; S4: Use the full-process collaborative management platform to conduct collaborative management of the entire process. When quality problems occur, collaborative processing is performed based on the full-process collaborative management platform to summarize and evaluate the quality monitoring process; The monitoring indicators in S1 include structural engineering indicators, geotechnical engineering indicators, building material indicators, and construction technology and environmental indicators; The quality monitoring plan formulated in S1 includes monitoring objectives and scope, sensor layout, data collection frequency, communication network architecture, data analysis process and the establishment of a full-process collaborative management platform.

[0010] Preferably, the Internet of Things communication network in S2 selects 5G communication modules and wireless ad hoc network sensors for installation.

[0011] Preferably, in S3, the construction data during the construction process is collected by task, stage and area according to different construction tasks, different construction stages and different areas of the construction project.

[0012] Preferably, the construction data collected in S3 includes engineering data and image data, and quality analysis of the collected engineering data and image data is performed based on a machine learning algorithm, as follows: Based on the data feature extraction module and the image feature extraction module, feature extraction is performed on the engineering data and the image data respectively, wherein the data feature extraction module extracts the periodic features of the engineering data based on Fourier transform or autocorrelation function; the image feature extraction module extracts the image features based on the corner point detection method; The feature data extracted by the data feature extraction module and the image feature extraction module are further selected based on the improved SOM network, wherein the improved SOM network includes an input layer and an output layer, and the calculation is specifically as follows: Initialization process: weights of output layer neurons Set an initial value and preset the end condition of network training; Get the input vector in the input , find the input vector The shortest distance between neurons , and calculate and The connection weight vector with the shortest distance between ; ;in, is the shortest distance function; Defined as the winning neuron, for The adjacent area introduces the gradient adjustment coefficient Correct the error in the adjustment process; all neurons in the adjacent area gradually move closer to the input vector in the following way : ; in, For SOM network The learning rate at the moment; for Neighboring function of the winning neuron at the moment; The gradient adjustment coefficient The calculation is as follows: ; ; ; ; ;in, and They are the first-order moment estimate and the second-order moment estimate of the gradient respectively; and are the corrected first-order moment estimates and second-order moment estimates, respectively; is the number of gradient iterations; Iterate gradients for weights; is the learning rate; and is the decay rate; A constant to make the denominator non-zero; Continue to input new samples and repeat the calculation steps above; Gradually reduce the adjacent area, reduce the learning rate, and repeat the above steps; When the weight error is lower than the allowed error value or meets the pre-set training length, the training is stopped and the output layer outputs the final feature selection result; The output of the improved SOW network is then input into the NFM network for quality analysis of the selected features. Before quality analysis, the NFM network also divides the feature data into a training set and a test set in a 1:1 ratio; The NFM network includes an embedding layer, a double interaction layer, a hidden layer and an output layer; The embedding layer is used for the input feature vector , by embedding the matrix Convert it to an embedding vector ; The dual interaction layer calculates the second-order interaction information of the features, assuming that the set of embedded vectors is , then the output of the double interaction layer for: ; The hidden layer is composed of several fully connected layers, and the number of hidden layers is , No. The output of the hidden layer is , the activation function is , then the calculation process of the hidden layer is: ; ; in, is the input of the hidden layer; For the The weight matrix of the hidden layer; For the The bias vector of the hidden layer; No. The output of the hidden layer; The output layer linearly transforms the output of the last hidden layer to obtain the final quality analysis result. : ; in, is the weight matrix of the output layer; is the bias vector of the output layer; The mean square error of the quality analysis result is defined as the loss function, and the loss function is minimized based on the improved snake optimization algorithm. The improved snake optimization algorithm is as follows: Randomly initialize the snake population and divide it into two parts: female snakes and male snakes; The inverse of the quality analysis error of the NFM network is used as the fitness of the algorithm. The fitness of the female snake and the male snake are calculated respectively. The individual optimal value and the global optimal value are calculated by the fitness function, and the best positions of the female snake and the male snake are recorded. Use the following formula to calculate the temperature threshold and food quantity , where the temperature threshold corresponds to the abnormality of the construction data, and the amount of food corresponds to the importance of the construction data: ; ;in, is the number of iterations; is the maximum number of iterations; determine the amount of food Values: when When , the female snake and the male snake look for food and determine whether the termination condition is met; when When the temperature threshold is determined Value: If , then the female and male snakes move closer to the optimal food location through the following formula: in, for female or male position; is the position of the best individual; is the learning constant; when When the random number Values: like , the female snake and the male snake begin to mate and produce the next generation: ; ;in, The female group the location of each individual; The male group the location of each individual; , The mating abilities of males and females, respectively; ; ; The next generation is replaced by the worst male and female individuals: ; ;in, and For the worst female and male individuals; and are the upper and lower limits of the individual, respectively; when When the female and male snakes fight with the same sex, the better snakes are produced: ; ; in, and Respectively represent the newer and better the location of the females and males; and Indicates the position of the best individuals in the female and male groups; and Represent the mating ability of females and males respectively; ; ;in, represents the fitness of the best individual in the female population; represents the fitness of the best individual in the male group; Update the positions of the female and male snakes, calculate their fitness values, and record their fitness values ​​and optimal positions; Determine whether the algorithm meets the termination conditions: if the iteration conditions or convergence accuracy are met, then output the final solution; if not, return to recalculate the fitness of the female and male snakes.

[0013] The NFM network is also trained using the training set data, and the model parameters are continuously updated iteratively until the loss function converges or the preset number of iterations is reached.

[0014] Preferably, in S3, based on the analysis results of the machine learning algorithm and in combination with the engineering quality standards, the warning thresholds of the quality indicators are set, and the warning thresholds are divided into different levels to provide warnings of quality problems of different degrees. Corresponding warning response processes are also established for warnings of quality problems of different degrees.

[0015] Preferably, in S3, digital twin technology is also combined to build a three-dimensional virtual construction model of the construction project based on the collected construction project process data, and VR technology and AR technology are also combined for virtual display.

[0016] Preferably, the full-process collaborative management platform described in S4 is used to incorporate all parties involved in the collaborative construction project into a unified management system, and all parties share construction project information in real time through the platform and perform collaborative work.

[0017] The whole process monitoring system of construction project quality based on the Internet of Things includes: Construction plan planning module: used to clarify the quality monitoring objectives and requirements of various monitoring indicators in the construction project, and formulate a quality monitoring plan based on the Internet of Things; Infrastructure construction module: used to install IoT devices at the construction site according to the established quality monitoring plan, build the IoT communication network, and debug the installed equipment and communication network; Data collection and analysis module: used for IoT devices to collect construction data in real time during the construction process, perform quality analysis and processing on the construction data based on machine learning algorithms, and issue corresponding quality warnings based on the quality warning mechanism according to the data analysis results; Engineering collaborative management module: used to use the full-process collaborative management platform to conduct collaborative management of the entire process. When quality problems occur, collaborative processing is performed based on the full-process collaborative management platform to summarize and evaluate the quality monitoring process. The construction scheme planning module includes: Target control unit: used to clarify the quality control targets and requirements of various monitoring indicators in construction projects; Plan formulation unit: used to formulate quality monitoring plans based on the Internet of Things; The infrastructure building module includes: Facility construction unit: used to evaluate, select and install IoT devices at the construction site according to the established quality monitoring plan, and build an IoT communication network; Facility debugging unit: used to debug installed equipment and communication networks; The data acquisition and analysis module includes: Data collection unit: used to collect construction data during the construction process by task, stage and area according to different construction tasks, different construction stages and different areas of the construction project; Data analysis unit: used to analyze and process construction data based on machine learning algorithms; Quality warning unit: used to establish a quality warning mechanism based on data analysis results, and to make corresponding quality warnings based on data analysis results; The engineering collaborative management module includes: Project information management unit: used to centrally manage basic project information and project documents of construction projects; Project schedule management unit: used to provide construction project planning tools to formulate project plans, track and provide feedback on project progress, and make corresponding resource allocation and progress association for the project; Project quality management unit: used to store quality standards and specification documents of construction projects, coordinate the implementation of the quality monitoring plan, and conduct statistical analysis and storage of data during the implementation of the quality monitoring plan; Project cost management unit: used to assist the construction unit and the construction unit in budget preparation and cost estimation, and to control and calculate project costs based on the quality monitoring plan, and further analyze and predict project cost data; Project safety management unit: used to store safety management documents for construction projects and conduct corresponding safety training management according to the construction stage and type of work; at the same time, formulate safety inspection plans, conduct regular hidden danger inspections, and formulate emergency response procedures for safety accidents; Project communication and collaboration unit: used to comprehensively connect all parties involved in the construction project for real-time communication, and provide a collaborative work platform for all parties, as well as notification and announcement functions.

[0018] Compared with the prior art, the present invention provides a method and system for monitoring the quality of construction projects throughout the entire process based on the Internet of Things, which has the following beneficial effects: The present invention deploys a large number of sensors in combination with Internet of Things technology, which can be used throughout the entire construction process to obtain project quality-related data in real time, and the data collection speed and accuracy are both high. It also combines machine learning algorithms to process and analyze data, and can analyze quality problems in the project relatively quickly and accurately; and carry out corresponding quality warnings; further build a full-process collaborative management platform to achieve information sharing and collaborative decision-making among all parties, improve the efficiency of quality problem handling, and avoid decision-making delays and deterioration of quality problems due to poor information. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a flow chart of the method mentioned in Example 1 of the present invention; Figure 2 This is a schematic diagram of the machine learning algorithm framework mentioned in Example 1 of the present invention; Figure 3 This is the feature selection flow chart mentioned in Example 1 of the present invention; Figure 4 This is a system block diagram mentioned in Example 2 of the present invention.

[0020] In the figure: 100. Construction plan planning module; 110. Target control unit; 120. Plan formulation unit; 200. Infrastructure construction module; 210. Facility construction unit; 220. Facility commissioning unit; 300. Data collection and analysis module; 310. Data collection unit; 320. Data analysis unit; 330. Quality warning unit; 400. Engineering collaborative management module; 410. Project information management unit; 420. Project progress management unit; 430. Project quality management unit; 440. Project cost management unit; 450. Project safety management unit; 460. Project communication and collaboration unit. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0022] Embodiment 1:

[0023] See also Figure 1-Figure 3 The present invention provides a method for monitoring the quality of a construction project in the whole process based on the Internet of Things, comprising the following steps: S1: Clarify the quality monitoring objectives and requirements of various monitoring indicators in construction projects, and formulate a quality monitoring plan based on the Internet of Things; the details are as follows: Clarify the quality monitoring goals and requirements for construction projects. For example, for a high-rise building, it may be necessary to focus on monitoring the quality of concrete pouring and steel structure welding. Based on these goals, design the architecture of the IoT monitoring system, including determining the type of data to be collected (such as temperature, pressure, displacement and other physical quantities), the selection of data acquisition equipment, and the installation location.

[0024] Taking concrete pouring as an example, it is necessary to consider installing temperature sensors inside the concrete and on the surface of the formwork to monitor the temperature changes of the concrete during the solidification process to prevent cracks caused by temperature stress. The selection of sensors should consider factors such as measurement accuracy and working environment (such as waterproof, pressure resistance, etc.). S2: According to the established quality monitoring plan, evaluate and select and install IoT devices at the construction site, build an IoT communication network, and debug the installed equipment and communication network; the details are as follows: Evaluate various IoT technologies and devices, and select appropriate IoT devices based on the system design plan, including sensors (such as smart sensor fusion devices, wireless ad hoc network sensors, etc.), data acquisition terminals (with data collection, preliminary processing and transmission functions), communication modules (such as Zigbee, LoRa, 4G / 5G, etc., used to send data to the server), etc.

[0025] For example, when selecting a data acquisition terminal, its compatibility should be considered to support the access of multiple types of sensors. At the same time, it is necessary to ensure that the signal coverage of the communication module can meet the needs of the construction site, especially for large-scale construction projects, which may require multiple communication nodes to ensure the stability of data transmission.

[0026] Build the necessary infrastructure for the operation of the IoT system. This includes installing data center servers (for storing and processing large amounts of monitoring data), network equipment (such as routers, switches, etc.), and network cabling to ensure smooth communication between the construction site and the data center. In addition, a stable power supply is required for sensors and data acquisition terminals. A combination of mains access and backup batteries can be used to prevent power outages from affecting data acquisition.

[0027] Install IoT devices at the construction site according to the design plan. Sensors should be installed in key locations. For example, during the foundation construction phase, settlement sensors should be installed at the corners and key load-bearing locations of the building foundation; when pouring concrete for beams and columns, concrete strain sensors should be installed on the steel skeleton. After the installation is complete, debug the equipment. Check whether the data collection of the sensor is normal through the data acquisition terminal, and whether the communication module can send the data to the data center accurately. For example, when debugging a temperature sensor, you can place the sensor in an environment with a known temperature, compare the actual temperature with the data collected by the sensor, and calibrate the sensor.

[0028] S3: IoT devices collect construction data in real time during the construction process, perform quality analysis and processing on the construction data based on machine learning algorithms, and issue corresponding quality warnings based on the data analysis results based on the quality warning mechanism; the details are as follows: During the construction process, IoT devices collect construction data in real time. Set the working mode and parameters of the data acquisition terminal to ensure the continuity and stability of data acquisition. The data acquisition terminal collects sensor data at a set frequency (such as collecting concrete temperature data every 10 minutes) and sends it to the data center server through the communication module. The collected construction data includes engineering data and image data. The quality analysis of the collected engineering data and image data is performed based on the machine learning algorithm, as follows: Reference Figure 2 , based on the data feature extraction module and the image feature extraction module, feature extraction is performed on the engineering data and the image data respectively, wherein the data feature extraction module extracts the periodic features of the engineering data based on Fourier transform or autocorrelation function; the image feature extraction module extracts the image features based on the corner point detection method; Reference Figure 3 , based on the improved SOM network, further feature selection is performed on the feature data extracted by the data feature extraction module and the image feature extraction module. The improved SOM network includes an input layer and an output layer, and its calculation is as follows: Initialization process: weights of output layer neurons Set an initial value and preset the end condition of network training; Get the input vector in the input , find the input vector The shortest distance between neurons , and calculate and The connection weight vector with the shortest distance between ; ; in, is the shortest distance function; Defined as the winning neuron, for The adjacent area introduces the gradient adjustment coefficient Correct the error in the adjustment process; all neurons in the adjacent area gradually move closer to the input vector in the following way : ; in, For SOM network The learning rate at the moment; for Neighboring function of the winning neuron at the moment; Gradient adjustment coefficient The calculation is as follows: ; ; ; ; ;in, and They are the first-order moment estimate and the second-order moment estimate of the gradient respectively; and They are the corrected first-order moment estimate and second-order moment estimate respectively; is the number of gradient iterations; Iterate gradients for weights; is the learning rate; and is the decay rate; A constant to make the denominator non-zero; Continue to input new samples and repeat the calculation steps above; Gradually reduce the adjacent area, reduce the learning rate, and repeat the above steps; When the weight error is lower than the allowed error value or meets the pre-set training length, the training is stopped and the output layer outputs the final feature selection result; The output of the improved SOW network is then input into the NFM network for quality analysis of the selected features. Before quality analysis, the NFM network also divides the feature data into a training set and a test set in a 1:1 ratio. The NFM network includes an embedding layer, a dual interaction layer, a hidden layer, and an output layer; The embedding layer takes the input feature vector , by embedding the matrix Convert it to an embedding vector ; The second-order interaction information of the double interaction layer calculation features, assuming that the set of embedded vectors is , then the output of the double interaction layer for: ; The hidden layer consists of several fully connected layers, and the number of hidden layers is , No. The output of the hidden layer is , the activation function is , then the calculation process of the hidden layer is: ; ;in, is the input of the hidden layer; For the The weight matrix of the hidden layer; For the The bias vector of the hidden layer; No. The output of the hidden layer; The output layer linearly transforms the output of the last hidden layer to obtain the final quality analysis result. : ;in, is the weight matrix of the output layer; is the bias vector of the output layer; The mean square error of the quality analysis results is defined as the loss function, and the loss function is minimized based on the improved snake optimization algorithm. The improved snake optimization algorithm is as follows: Randomly initialize the snake population and divide it into two parts: female snakes and male snakes; The inverse of the quality analysis error of the NFM network is used as the fitness of the algorithm. The fitness of the female snake and the male snake are calculated respectively. The individual optimal value and the global optimal value are calculated by the fitness function, and the best positions of the female snake and the male snake are recorded. Use the following formula to calculate the temperature threshold and food quantity , where the temperature threshold corresponds to the abnormality of the construction data, and the amount of food corresponds to the importance of the construction data: ; ;in, is the number of iterations; is the maximum number of iterations; Judging food quantities Value: When When , the female snake and the male snake look for food and determine whether the termination condition is met; when When the temperature threshold is determined Value: If , then the female and male snakes move closer to the optimal food location through the following formula: in, for female or male position; is the position of the best individual; is the learning constant; when When the random number Values: like , the female snake and the male snake begin to mate and produce the next generation: ; ;in, The female group the location of each individual; The male group the location of each individual; , The mating abilities of males and females, respectively; ; ; The next generation is replaced by the worst male and female individuals: ; ;in, and For the worst female and male individuals; and are the upper and lower limits of the individual, respectively; when When the female and male snakes fight with the same sex, the better snakes are produced: ; ; in, and Respectively represent the newer and better the location of the females and males; and Indicates the position of the best individuals in the female and male groups; and Represent the mating ability of females and males respectively; ; ;in, represents the fitness of the best individual in the female population; represents the fitness of the best individual in the male group; Update the positions of the female and male snakes, calculate their fitness values, and record their fitness values ​​and optimal positions; Determine whether the algorithm meets the termination conditions: if the iteration conditions or convergence accuracy are met, then output the final solution; if not, return to recalculate the fitness of the female and male snakes.

[0029] The NFM network is also trained using the training set data, and the model parameters are continuously updated iteratively until the loss function converges or the preset number of iterations is reached.

[0030] Based on the analysis results of the machine learning algorithm and combined with the engineering quality standards, the warning thresholds of the quality indicators are set, and the warning thresholds are divided into different levels to provide warnings for quality problems of different degrees. In addition, corresponding warning response processes are established for warnings of quality problems of different degrees.

[0031] Once the early warning information is received, the construction unit should promptly organize personnel to investigate and deal with the quality problems. By viewing the detailed monitoring data and the actual situation on site, the cause of the problem can be analyzed and corresponding remedial measures can be taken, such as reworking unqualified welding parts. In addition, digital twin technology is combined to build a three-dimensional virtual construction model of the construction project based on the collected construction process data, and VR technology and AR technology are combined for virtual display. The use of VR and AR technology can provide quality monitoring personnel with a more intuitive visual experience. AR technology can also superimpose monitoring data on actual objects in the form of virtual labels at the construction site, so that construction personnel can more intuitively understand the quality parameters of the current part.

[0032] S4: Use the full-process collaborative management platform to conduct collaborative management of the entire process. When quality problems occur, collaborative processing is carried out based on the full-process collaborative management platform to summarize and evaluate the quality monitoring process. The details are as follows: Establish a collaborative management platform for the entire construction process based on the Internet of Things, and incorporate all parties such as construction units, design units, construction units, and supervision units into a unified management system. All parties can share monitoring data and quality information in real time through the platform to achieve collaborative work. The platform also has task management and process approval functions, which can ensure the standardization and standardization of quality monitoring work.

[0033] The full-process collaborative management platform for construction projects collects all monitoring data during the construction phase for comprehensive collation and analysis. By comparing data from different stages, the stability of the project quality can be comprehensively evaluated. The project quality can also be comprehensively evaluated based on the data analysis results and combined with the construction project quality acceptance standards. The large amount of data provided by the Internet of Things system can serve as an important basis for quality evaluation. For example, for the installation quality of the electrical system in a smart building, the operating parameters of the monitoring equipment (such as voltage, current, etc.) can be used to determine whether it meets the design and specification requirements. If all monitoring data and indicators meet the acceptance standards, the project can pass the acceptance smoothly; otherwise, the existing quality problems need to be rectified until the acceptance requirements are met.

[0034] After the acceptance is completed, the IoT monitoring system is maintained. The operating status of the equipment is checked, and the equipment that may fail is repaired or replaced. At the same time, all data from the project quality monitoring process is archived and saved for subsequent review and reference. These data are of great value for quality tracing, maintenance and reinforcement of the building during use.

[0035] Embodiment 2:

[0036] See also Figure 4 The present invention provides a whole-process monitoring system for construction project quality based on the Internet of Things, including: a construction plan planning module 100 including: Target control unit 110: used to clarify the quality control targets and requirements of various monitoring indicators in the construction project; A plan formulation unit 120 is used to formulate a quality monitoring plan based on the Internet of Things; The infrastructure building module 200 includes: Facility construction unit 210: used to evaluate, select and install IoT devices at the construction site according to the established quality control plan, and build an IoT communication network; Facility debugging unit 220: used to debug the installed equipment and communication network; The data collection and analysis module 300 includes: Data collection unit 310: used to collect construction data in the construction process by task, stage and area according to different construction tasks, different construction stages and different areas of the construction project; Data analysis unit 320: used for performing quality analysis on construction data based on machine learning algorithms; Quality warning unit 330: used to establish a quality warning mechanism according to the data analysis results, and to make corresponding quality warnings according to the data analysis results; The engineering collaborative management module 400 includes: Project information management unit 410: used to store basic details of construction projects, such as project name, location, type (residential, commercial, industrial, etc.), scale (building area, number of floors, etc.), investment amount, etc. This information provides basic background data for the entire project, making it convenient for all parties to quickly understand the project overview. It is also used to centrally manage various project documents, including design drawings, construction plans, contract documents, technical specifications, acceptance reports, etc. It supports document upload, download, version control and permission management.

[0037] Project schedule management unit 420: used to provide schedule planning tools, supporting Gantt charts, critical path method (CPM) and other methods. Project managers can develop detailed construction schedules based on project goals and task decomposition, and clarify the start time, completion time and duration of each process. Real-time tracking of project progress, the construction unit can regularly update the actual progress, the system automatically compares the actual progress with the planned progress, and timely discovers delays or advances through progress deviation analysis. When the progress deviation exceeds a certain threshold, the system automatically issues an early warning. For example, if the main structure construction progress lags behind the planned progress by more than one week, the system will send an early warning message to the relevant managers to remind them to take measures.

[0038] The allocation of resources such as manpower, materials, and equipment is also combined with the schedule. The system can calculate the amount of resources required for each stage based on the schedule and monitor the actual input of resources.

[0039] Project Quality Management Unit 430: It is used to provide quality standards and specification documents related to construction projects, which are convenient for all parties to consult. At the same time, the quality standards are linked to specific construction processes and acceptance links to ensure that quality control during the construction process has rules to follow. For example, in reinforced concrete construction, the system can automatically associate relevant specifications such as concrete strength standards and steel bar anchorage length standards.

[0040] Support the coordinated implementation of quality monitoring programs, including daily inspections, hidden project inspections, and acceptance of sub-items and sub-projects. Inspectors can use mobile devices to record inspection results on site, upload photos and descriptions of quality problems, and the system automatically generates quality inspection reports. For situations that do not meet quality standards, the system initiates the rectification process and tracks the rectification until the problem is resolved. For example, during the acceptance of hidden projects, the supervision personnel record the quality of steel bar binding through the platform. After discovering problems, the construction unit receives rectification notices on the platform. After the rectification is completed, upload photos and materials after the rectification for review.

[0041] It is also used to collect and analyze quality data, such as the distribution of quality problem types, quality qualification rate, and frequency of quality problems. Through data mining and visualization technology, it provides decision support for quality management. Project cost management unit 440: It is used to assist construction units and construction units in preparing project budgets, including bill of quantities pricing, quota pricing, etc. According to the design drawings and relevant specifications, the budget cost of each project content is calculated. At the same time, during the project, the cost can be dynamically estimated according to the actual situation. Based on the quality monitoring plan, the expenditure of project costs is monitored in real time, and the actual cost is compared and analyzed with the budget cost. Early warning of cost overruns is issued, and the reasons for cost overruns are analyzed, such as rising material prices and increased engineering volume. Construction units can use the platform to perform cost accounting, including the accounting of various cost elements such as labor costs, material costs, and machinery costs. Using historical cost data and the current cost situation of the project, the future cost trend of the project is predicted through data analysis and prediction models. Provide reference for project cost control and decision-making.

[0042] Project safety management unit 450: used to store documents such as the safety management system, operating procedures and emergency plans for construction projects. At the same time, organize safety training plans and record training content, training personnel and training results. The system can push corresponding safety operating procedures training courses according to the construction stage and type of work. After the construction personnel complete the training, the system records the training results as one of the bases for taking up the post. Formulate a safety inspection plan, including regular inspections and special inspections. Inspectors record safety hazards on site through mobile devices, such as whether the protective facilities at the construction site are in place and whether there are safety failures in the construction machinery. The system generates a safety inspection report and tracks and corrects the hazards. When a safety accident occurs, the system starts the safety accident handling process. Record information such as the time, location, process, cause and loss of the accident. Analyze the accident, summarize the lessons learned, and provide a basis for improvement for subsequent safety management.

[0043] Project communication and collaboration unit 460: used to provide an instant messaging tool similar to chat software, which comprehensively connects construction units, design units, construction units, supervision units and other personnel. A project group can be established to discuss specific issues or tasks. A collaborative work platform for all parties is also provided to support multiple people to edit documents, drawings and other files online at the same time to achieve collaborative work. Used to publish important notifications, announcements and file update information of the project. Personnel from all parties can receive notifications in a timely manner through the platform to ensure timely communication of information. The above is only a preferred specific implementation method of the present invention, but the protection scope of the present invention is not limited to this. Any technician familiar with the technical field within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, is equivalently replaced or changed, which should be covered within the protection scope of the present invention.

Claims

1. A method for monitoring the quality of construction projects throughout the entire process based on the Internet of Things, characterized by: The steps include: S1: Clarify the quality monitoring objectives and requirements of various monitoring indicators in construction projects, and formulate a quality monitoring plan based on the Internet of Things; S2: According to the established quality control plan, evaluate, select and install IoT devices at the construction site, build the IoT communication network, and debug the installed equipment and communication network; S3: IoT devices collect construction data in real time during the construction process, perform quality analysis and processing on the construction data based on machine learning algorithms, and issue corresponding quality warnings based on the data analysis results based on the quality warning mechanism; S4: Use the full-process collaborative management platform to conduct collaborative management of the entire process. When quality problems occur, collaborative processing is performed based on the full-process collaborative management platform to summarize and evaluate the quality monitoring process; The monitoring indicators in S1 include structural engineering indicators, geotechnical engineering indicators, building material indicators, and construction technology and environmental indicators; The quality monitoring plan formulated in S1 includes monitoring objectives and scope, sensor layout, data collection frequency, communication network architecture, data analysis process and the establishment of a full-process collaborative management platform.

2. The method for monitoring the quality of construction projects throughout the entire process based on the Internet of Things according to claim 1 is characterized in that: The Internet of Things communication network in S2 selects 5G communication modules and wireless ad hoc network sensors for installation.

3. The method for monitoring the quality of construction projects throughout the entire process based on the Internet of Things according to claim 1 is characterized in that: In S3, according to different construction tasks, different construction stages and different areas of the construction project, the construction data during the construction process is collected by task, stage and area.

4. The method for monitoring the quality of construction projects throughout the entire process based on the Internet of Things according to claim 3 is characterized in that: The construction data collected in S3 includes engineering data and image data. The quality analysis of the collected engineering data and image data is performed based on the machine learning algorithm, as follows: Based on the data feature extraction module and the image feature extraction module, feature extraction is performed on the engineering data and the image data respectively, wherein the data feature extraction module extracts the periodic features of the engineering data based on Fourier transform or autocorrelation function; the image feature extraction module extracts the image features based on the corner point detection method; The feature data extracted by the data feature extraction module and the image feature extraction module are further selected based on the improved SOM network, wherein the improved SOM network includes an input layer and an output layer, and the calculation is specifically as follows: Initialization process: weights of output layer neurons Set an initial value and preset the end condition of network training; Get the input vector in the input , find the input vector The shortest distance between neurons , and calculate and The connection weight vector with the shortest distance between ; ;in, is the shortest distance function; Defined as the winning neuron, for The adjacent area introduces the gradient adjustment coefficient Correct the error in the adjustment process; all neurons in the adjacent area gradually move closer to the input vector in the following way : ;in, For SOM network The learning rate at the moment; for Neighboring function of the winning neuron at the moment; The gradient adjustment coefficient The calculation is as follows: ; ; ; ; ;in, and They are the first-order moment estimate and the second-order moment estimate of the gradient respectively; and They are the corrected first-order moment estimate and second-order moment estimate respectively; is the number of gradient iterations; Iterate gradients for weights; is the learning rate; and is the decay rate; A constant to make the denominator non-zero; Continue to input new samples and repeat the calculation steps above; Gradually reduce the adjacent area, reduce the learning rate, and repeat the above steps; When the weight error is lower than the allowed error value or meets the pre-set training length, the training is stopped and the output layer outputs the final feature selection result; The output of the improved SOW network is then input into the NFM network for quality analysis of the selected features. Before quality analysis, the NFM network also divides the feature data into a training set and a test set in a 1:1 ratio; The NFM network includes an embedding layer, a double interaction layer, a hidden layer and an output layer; The embedding layer is used for the input feature vector , by embedding the matrix Convert it to an embedding vector ; The dual interaction layer calculates the second-order interaction information of the features, assuming that the set of embedded vectors is , then the output of the double interaction layer for: ; The hidden layer is composed of several fully connected layers, and the number of hidden layers is , No. The output of the hidden layer is , the activation function is , then the calculation process of the hidden layer is: ; ;in, is the input of the hidden layer; For the The weight matrix of the hidden layer; For the The bias vector of the hidden layer; No. The output of the hidden layer; The output layer linearly transforms the output of the last hidden layer to obtain the final quality analysis result. : ;in, is the weight matrix of the output layer; is the bias vector of the output layer; The mean square error of the quality analysis result is defined as the loss function, and the loss function is minimized based on the improved snake optimization algorithm. The improved snake optimization algorithm is as follows: Randomly initialize the snake population and divide it into two parts: female snakes and male snakes; The inverse of the quality analysis error of the NFM network is used as the fitness of the algorithm. The fitness of the female snake and the male snake are calculated respectively. The individual optimal value and the global optimal value are calculated by the fitness function, and the best positions of the female snake and the male snake are recorded. Use the following formula to calculate the temperature threshold and food quantity , where the temperature threshold corresponds to the abnormality of the construction data, and the amount of food corresponds to the importance of the construction data: ; ;in, is the number of iterations; is the maximum number of iterations; determine the amount of food Values: when When , the female snake and the male snake look for food and determine whether the termination condition is met; when When the temperature threshold is determined Value: If , then the female and male snakes move closer to the optimal food location through the following formula: ;in, for female or male position; is the position of the best individual; is the learning constant; when When the random number Values: like , the female snake and the male snake begin to mate and produce the next generation: ; ;in, The female group the location of each individual; The male group the location of each individual; , The mating abilities of males and females, respectively; ; ; The next generation is replaced by the worst male and female individuals: ; ;in, and For the worst female and male individuals; and are the upper and lower limits of the individual respectively; when When the female and male snakes fight with the same sex, the better snakes are produced: ; ;in, and Respectively represent the newer and better the location of the females and males; and Indicates the position of the best individuals in the female and male groups; and Represent the mating ability of females and males respectively; ; ;in, represents the fitness of the best individual in the female population; represents the fitness of the best individual in the male group; Update the positions of the female and male snakes, calculate their fitness values, and record their fitness values ​​and optimal positions; Determine whether the algorithm meets the termination conditions: if the iteration conditions or convergence accuracy are met, then output the final solution; if not, return to recalculate the fitness of the female and male snakes; The NFM network is also trained using the training set data, and the model parameters are continuously updated iteratively until the loss function converges or the preset number of iterations is reached.

5. The method for monitoring the quality of construction projects throughout the entire process based on the Internet of Things according to claim 4 is characterized in that: In S3, based on the analysis results of the machine learning algorithm and in combination with the engineering quality standards, the warning thresholds of the quality indicators are set, and the warning thresholds are divided into different levels to provide warnings of quality problems of different degrees. Corresponding warning response processes are also established for warnings of quality problems of different degrees.

6. The method for monitoring the quality of construction projects throughout the entire process based on the Internet of Things according to claim 5 is characterized by: In the S3, digital twin technology is also combined to build a three-dimensional virtual construction model of the construction project based on the collected construction project process data, and VR technology and AR technology are also combined for virtual display.

7. The method for monitoring the quality of construction projects throughout the entire process based on the Internet of Things according to claim 1 is characterized in that: The full-process collaborative management platform described in S4 is used to incorporate all parties involved in the collaborative construction project into a unified management system. All parties share construction project information in real time through the platform and work collaboratively.

8. The Internet of Things-based construction project quality whole-process monitoring system applied to the method described in any one of claims 1 to 7 is characterized in that: include: Construction plan planning module (100): used to clarify the quality monitoring objectives and requirements of various monitoring indicators in the construction project, and formulate a quality monitoring plan based on the Internet of Things; Infrastructure construction module (200): used to install IoT devices at the construction site according to the established quality monitoring plan, build an IoT communication network, and debug the installed devices and communication network; Data collection and analysis module (300): used for collecting construction data in real time during the construction process using IoT devices, performing quality analysis and processing on the construction data based on a machine learning algorithm, and issuing corresponding quality warnings based on the data analysis results based on a quality warning mechanism; Engineering collaborative management module (400): used to use the full-process collaborative management platform to perform collaborative management of the entire process. When quality problems occur, collaborative processing is performed based on the full-process collaborative management platform to summarize and evaluate the quality monitoring process. The construction scheme planning module (100) comprises: Target Control Unit (110): used to define the quality control targets and requirements of various monitoring indicators in the construction project; Plan formulation unit (120): used to formulate a quality monitoring plan based on the Internet of Things; The infrastructure construction module (200) includes: Facility construction unit (210): used to evaluate, select and install IoT devices at the construction site according to the established quality control plan, and build an IoT communication network; Facility debugging unit (220): used to debug installed equipment and communication networks; The data collection and analysis module (300) comprises: Data collection unit (310): used for collecting construction data in the construction process according to different construction tasks, different construction stages and different areas of the construction project, by task, stage and area; Data analysis unit (320): used for analyzing and processing construction data based on a machine learning algorithm; Quality warning unit (330): used to establish a quality warning mechanism according to the data analysis results, and to issue corresponding quality warnings according to the data analysis results; The engineering collaborative management module (400) includes: Project information management unit (410): used to centrally manage basic project information and project documents of the construction project; Project progress management unit (420): used to provide a project planning tool for the construction project to formulate a project plan, track and provide feedback on the project progress, and perform corresponding resource allocation and progress association for the project; Project quality management unit (430): used to store quality standards and specification documents of construction projects, coordinate the implementation of the quality monitoring plan, and perform statistical analysis and storage of data during the implementation of the quality monitoring plan; Project cost management unit (440): used to assist the construction unit and the construction unit in budget preparation and cost estimation, and to control and calculate project costs based on the quality control plan, and further to analyze and predict project cost data; Project Safety Management Unit (450): used to store safety management documents for construction projects and conduct corresponding safety training management according to the construction stage and type of work; at the same time, formulate safety inspection plans, conduct regular hidden danger inspections, and formulate emergency response procedures for safety accidents; Project Communication and Collaboration Unit (460): It is used to comprehensively connect all parties involved in the construction project for real-time communication, and provides a collaborative work platform for all parties, as well as notification and announcement functions.

Images