Intelligent construction site data acquisition and visual management method based on Internet of Things

Abstract

The invention discloses an intelligent construction site data collection and visual management method based on the Internet of Things, and the method comprises the following steps: collecting real-time multi-dimensional data of a construction site, carrying out the joint optimization of node selection and multi-dimensional resource distribution, carrying out the data size optimization, and transmitting the data to a data processing center; constructing a risk evaluation index system of the construction site, inputting the optimized real-time multi-dimensional data and the risk evaluation index system into a construction site analysis model for identification, and extracting potential risk factors in the construction site; inputting the potential risk factors into a construction site early warning model for analysis, and generating risk early warning information; performing multi-modal data fusion on the real-time multi-dimensional data, the potential risk factors and the risk early warning information to generate a construction site visual data model; and analyzing the construction site visual data model through the associated construction site optimization model, and generating a dynamic optimization path of the construction site environment and the construction process. According to the invention, the management level of the intelligent construction site is obviously improved, and the construction process and resource allocation are optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent construction site management, and in particular to a smart construction site data collection and visualization management method based on the Internet of Things. Background Art

[0002] With the continuous development of the construction industry, the scale of modern construction projects is expanding, the construction environment is becoming more complex, and the safety management of construction sites and the improvement of construction efficiency have become the focus of the industry. However, the traditional construction site management method still relies on manual inspections, paper records and other means, with low information collection efficiency and insufficient accuracy, and it is difficult to grasp the dynamic changes of the construction site in real time. This traditional management model has problems such as information lag, difficulty in fully perceiving and managing the construction site environment and potential risks in the construction process, which makes it difficult to discover and deal with safety hazards in the construction process in a timely manner, thus posing a serious threat to project quality and personnel safety.

[0003] In recent years, the rapid development of Internet of Things (IoT) technology has provided technical support for the construction of smart construction sites. By combining sensors, terminal devices, communication networks and data analysis technology, IoT can realize the real-time collection, transmission and analysis of multi-dimensional data such as construction site environment, equipment operating status, construction progress, etc. However, relying solely on a single data collection and monitoring function is difficult to meet the multi-dimensional management needs of smart construction sites. For example, in a complex and changeable construction site environment, how to efficiently select appropriate collection nodes and reasonably allocate limited IoT resources has become an important challenge facing the IoT data collection system; at the same time, relying solely on a single data dimension or simple data analysis methods is difficult to accurately identify potential risks, which puts higher requirements on the safety management of the construction site.

[0004] In the construction of smart construction sites, how to deeply integrate the Internet of Things technology with big data analysis, visualization technology, artificial intelligence and other technologies to achieve comprehensive perception of construction site dynamics, accurate identification and early warning of potential risks, and intelligent management and optimization of the construction process is a hot topic in current research and application. Existing smart construction site management methods generally have the following shortcomings: (1) Insufficient real-time data collection and optimization: Existing methods do not fully utilize the perception capabilities of IoT devices to comprehensively collect multi-dimensional data, and lack effective optimization methods in node selection and resource allocation, resulting in low data collection efficiency and waste of resources.

[0005] (2) Potential risk identification is not accurate enough: Traditional risk assessment methods do not make sufficient use of the multi-dimensional data of construction sites. Risk identification is based only on a single dimension or simple statistical methods, making it difficult to accurately extract potential risk factors, resulting in delayed and incomplete risk warnings.

[0006] (3) Insufficient data fusion and visualization: Existing technologies lack deep mining and correlation analysis when integrating multimodal data from construction sites, and fail to effectively realize the visual dynamic management of multidimensional data and risk information, affecting the real-time monitoring and optimization decision-making of the construction process.

[0007] (4) Insufficient dynamic optimization of the construction process: In the management and optimization of the construction process, existing methods often lack systematic and intelligent path optimization means, and fail to effectively combine real-time data with factors such as construction progress and risk information to dynamically optimize the site environment and construction process. Summary of the invention

[0008] In order to overcome the shortcomings of the prior art, the present invention provides a smart construction site data collection and visualization management method based on the Internet of Things, which is used to solve the shortcomings of the existing smart construction site management methods in real-time data collection, construction safety monitoring, risk warning and resource optimization allocation, thereby significantly improving the management level of the smart construction site, which can not only ensure the safety of the construction site, but also optimize the construction process and resource allocation.

[0009] To solve the above problems, the technical solution adopted by the present invention is as follows: A method for data collection and visualization management of smart construction sites based on the Internet of Things, comprising the following steps: Collect real-time multi-dimensional data of the construction site through various types of IoT terminal devices distributed on the construction site; Jointly optimizing node selection and multi-dimensional resource allocation for the real-time multi-dimensional data, and transmitting the data to a data processing center after optimizing the data volume; Identify and screen the risk assessment indicators of the construction site, and build a risk assessment indicator system for the construction site. Input the optimized real-time multi-dimensional data and risk assessment indicator system into the construction site analysis model for identification and extraction of potential risk factors in the construction site. Inputting the potential risk factors into the construction site early warning model for analysis to generate risk early warning information; Perform multimodal data fusion on real-time multidimensional data, potential risk factors and risk warning information, generate a construction site visualization data model, and conduct dynamic visualization monitoring and management of the construction site environment and construction process; Analyze the site visualization data model through the associated site optimization model to generate a dynamic optimization path for the site environment and construction process; Among them, when generating the dynamic optimization path of the construction site environment and the construction process, the dynamic optimization of the path is achieved through population generation and population evolution based on the objective function.

[0010] As a preferred embodiment of the present invention, when collecting real-time multi-dimensional data of a construction site, it includes: Obtain the construction site coverage of the current construction site and determine the type of IoT terminal device based on the real-time multi-dimensional data to be collected; Obtaining the node unit sensing distance of each type of IoT terminal device, and determining the setting position of each type of IoT terminal device within the coverage range of the construction site according to the node unit sensing distance, and calibrating the same; The corresponding types of IoT terminal devices are set at the calibration positions, and the coverage and collection range of the IoT terminal devices at each calibration position are calibrated, and the optimal data collection frequency of each type of IoT terminal device is determined.

[0011] As a preferred implementation mode of the present invention, when performing joint optimization of node selection and multi-dimensional resource allocation, it includes: By establishing The cluster configures a fixed resource allocation strategy for the collected real-time multi-dimensional data; according to The resource allocation strategy of the cluster configuration obtains the first state set and inputs it into the node selection model to obtain Node selection strategy; based on Node selection strategy selected The existing resource state of the node is used to obtain a second state set, which is input into a resource allocation model to obtain a multi-dimensional resource allocation strategy; Inputting the multi-dimensional resource allocation strategy into the node selection model as a third state set to adjust the model result; The node selection model and the resource allocation model are iterated alternately until the total objective function reaches the set maximum allowable error; Among them, in all In the node, based on the first state set, select the node with the largest first reward function result Node is the final selected node, and Node selection strategy; During the resource allocation process, a resource allocation strategy with the largest second reward function result is selected as the multi-dimensional resource allocation strategy based on the second state set.

[0012] As a preferred embodiment of the present invention, when the data volume is optimized and transmitted to the data processing center, it includes: Integrating the real-time multi-dimensional data using Horner function parameters to obtain integrated real-time multi-dimensional data; Obtaining an optimized estimated value, and optimizing the data volume of the integrated real-time multi-dimensional data by using the optimized estimated value to obtain optimized real-time multi-dimensional data; The optimized real-time multi-dimensional data is transmitted to the data processing center through the selected nodes and based on the determined multi-dimensional resource allocation strategy.

[0013] As a preferred embodiment of the present invention, when extracting potential risk factors in a construction site, it includes: Will The subjective weighting method is combined with the objective weighting method of anti-entropy weighting to combine and weight the risk assessment index system to obtain the comprehensive weight of the index; According to the comprehensive weight of the indicators, the relative closeness of the grayscale association is obtained; Determine the risk assessment level of each risk assessment indicator according to the relative proximity of the grayscale association and sort them, and obtain the risk assessment indicators with risk assessment levels greater than a threshold from the sorting as potential risk indicators; Real-time multi-dimensional data associated with the potential risk indicator is obtained as the potential risk factor.

[0014] As a preferred embodiment of the present invention, when obtaining the relative proximity of grayscale association, it includes: The comprehensive weights of the indicators are combined with the standardized fuzzy decision matrix to obtain a weighted normalized matrix; Select the optimal indicator set as the reference sequence for comparison and evaluation of risk assessment indicators; Establishing a grey correlation coefficient matrix based on the weighted normalization matrix, and obtaining a positive ideal solution and a negative ideal solution in the grey correlation coefficient matrix; The distances from the risk assessment index values ​​under each assessment target to the positive ideal solution and the negative ideal solution are obtained, and the relative proximity of the grey correlation is obtained.

[0015] As a preferred embodiment of the present invention, when generating risk warning information, it includes: Constructing a training sample set, and discretizing and simplifying the training sample set, deleting incomplete data therein, and sorting the training sample set from large to small to obtain an optimal attribute set; Reference the cross-validation method to select kernel function parameters and penalty parameters, and use Programming to solve the model parameters; A construction site early warning model based on support vector machine is constructed, the potential risk factors are input into the construction site early warning model, and risk early warning information is output.

[0016] As a preferred embodiment of the present invention, when constructing a construction site early warning model based on a support vector machine, it includes: Using nonlinear mapping to map the optimal attribute set from the original space to a high-dimensional space, and performing linear regression in the high-dimensional space to obtain a regression function; Transforming the regression function according to Lagrangian to obtain a dual form, and obtaining a regression prediction function according to training and learning; The regression prediction function is combined with the error detection function to obtain the construction site early warning model.

[0017] As a preferred embodiment of the present invention, when generating a construction site visualization data model, it includes: Constructing a three-dimensional digital model of the construction site and several construction site equipment models, and combining them to generate the construction site visualization data model; Among them, when constructing a three-dimensional digital model of the construction site, it includes: Use 3D digital models to simulate construction progress and predict progress plans. By integrating real-time multi-dimensional data, potential risk factors and risk warning information, we can focus on monitoring risk areas, automatically compare planned progress with actual progress, and generate progress deviations. The components in the 3D digital model are associated with the risk operations involved in the potential risk factors and the standard processes used in the construction procedures. The real-time process parameters of the risk operations are automatically compared with the standard process parameters to generate process deviations.

[0018] As a preferred embodiment of the present invention, when the dynamic optimization of the path is realized by population generation and population evolution based on the objective function, it includes: Providing a site environmental objective function based on the Euclidean distance between facility centroids and safety factors; Provide construction process objective function based on progress deviation value and progress correction weight; Performing array encoding on the design variables, generating an initial population using random numbers, treating each chromosome in the initial population as a feasible solution, and substituting the feasible solution into the construction site environment objective function and the construction process objective function to obtain the objective function value; The selection operator is used to select some chromosomes to enter the mating pool, and the chromosomes in the mating pool are allowed to perform crossover operations according to the crossover probability to generate offspring individuals; Some individuals in the population are randomly mutated according to the mutation probability to detect whether the population meets the preset termination iteration conditions. If so, the population evolution is terminated and the dynamic optimization path of the construction site environment and construction process is output. If not, a new initial population is generated using random numbers.

[0019] Compared with the prior art, the present invention has the following beneficial effects: (1) Efficient data collection and optimization The present invention maximizes the use of the resources of the collection equipment by rationally configuring the type, location and collection frequency of the IoT terminal equipment, combined with the joint optimization of node selection and multi-dimensional resource allocation, and significantly improves the efficiency and quality of real-time multi-dimensional data collection. Through the data volume optimization method, while ensuring data accuracy, the data transmission volume is reduced, the pressure of network transmission is reduced, and the real-time performance and stability of the system are improved.

[0020] (2) Accurate risk identification and early warning The present invention will The subjective weighting method is combined with the objective weighting method of anti-entropy weighting, and the multi-dimensional data of the construction site is combined and weighted to analyze, build a scientific risk assessment index system, accurately extract potential risk factors, and improve the accuracy of risk identification. The construction site early warning model based on support vector machine, using regression prediction and error detection in high-dimensional space, can quickly and accurately generate risk warning information, providing reliable protection for construction safety.

[0021] (3) Comprehensive data fusion and dynamic visualization Through multimodal data fusion, the present invention effectively integrates real-time multidimensional data, potential risk factors and risk warning information, generates a three-dimensional visualization data model of the construction site, and realizes dynamic monitoring and intelligent management of the construction site environment and construction process. The visualization model uses three-dimensional digital technology to intuitively present the construction progress, risk areas and process deviations, allowing managers to clearly understand the real-time status of the construction site and improve management efficiency.

[0022] (4) Intelligent dynamic optimization of the construction process The present invention is based on the dynamic objective function of the construction site environment and the construction process, combined with the facility centroid distance, safety factor, progress deviation value, and progress correction weight, and uses an intelligent optimization algorithm to generate a dynamic optimization path for the construction site environment and the construction process. The generation process of the optimization path ensures the global optimality of the solution through the selection, crossover and mutation operations of the genetic algorithm, and can dynamically adjust the resource allocation and operation process during the construction process, effectively improving the construction efficiency and engineering quality.

[0023] (5) Improving construction safety and engineering efficiency The present invention comprehensively considers the dynamic changes of multi-dimensional data and its impact on construction site safety and construction efficiency in the whole process of risk identification, early warning, monitoring and optimization, and effectively reduces safety hazards and resource waste through systematic and intelligent management. The real-time comparison function of process parameters and standard parameters can help the construction team to promptly discover deviations in construction and take corrective measures, thereby reducing the uncertainty in construction site operations and ensuring the timely delivery and construction quality of the project.

[0024] (6) Strong flexibility and scalability The multi-dimensional data collection method and optimization algorithm of the present invention can adapt to construction site scenarios of different scales and types, and support the access of multiple IoT devices and multimodal data, providing high flexibility and scalability for smart construction site management. The standardized design and visual presentation of the data model support seamless connection with other management systems, laying a solid foundation for the digital and intelligent transformation of the construction industry.

[0025] (7) Save costs and resources By dynamically optimizing nodes and resources, the present invention avoids the setting of redundant nodes and irrational allocation of resources, significantly reducing the cost of deploying and operating IoT devices. The optimized design of data transmission and processing reduces the consumption of bandwidth and storage resources, improving the economy and sustainability of the overall system.

[0026] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a step diagram of the method for data collection and visualization management of smart construction sites based on the Internet of Things provided by the present invention; Figure 2 It is an example diagram of the risk assessment index system for the construction site constructed by the present invention. DETAILED DESCRIPTION

[0028] The present invention provides a method for collecting and visualizing data on a smart construction site based on the Internet of Things, such as Figure 1 As shown, the following steps are included: Step S1: collecting real-time multi-dimensional data of the construction site through various types of IoT terminal devices distributed on the construction site, wherein the real-time multi-dimensional data includes real-time construction site environment data and real-time construction data; Step S2: Jointly optimize node selection and multi-dimensional resource allocation for real-time multi-dimensional data, and transmit the data to the data processing center after optimizing the data volume; Step S3: Identify and screen construction site risk assessment indicators, and build a construction site risk assessment indicator system; Step S4: inputting the optimized real-time multi-dimensional data and risk assessment index system into the construction site analysis model for identification, and extracting potential risk factors in the construction site; Step S5: inputting potential risk factors into the construction site early warning model for analysis to generate risk warning information including safety accident risk warning and construction delay risk warning; Step S6: Perform multimodal data fusion on real-time multidimensional data, potential risk factors and risk warning information to generate a construction site visualization data model; Step S7: Implementing dynamic visualization monitoring and management of the construction site environment and construction process based on the construction site visualization data model; Step S8: Analyze the construction site visualization data model through the associated construction site optimization model to generate a dynamic optimization path for the construction site environment and the construction process.

[0029] Specifically, various types of IoT terminal devices include: environmental monitoring equipment, video surveillance equipment, smart sensors, construction equipment monitoring devices, and personnel positioning equipment.

[0030] Among them, the environmental monitoring equipment includes: temperature and humidity sensors, gas sensors, dust sensors and noise detection modules, which are used to collect environmental parameters including temperature, humidity, air quality (PM2.5, PM10), harmful gas concentration and noise decibels.

[0031] Video surveillance equipment includes: high-definition cameras and thermal imaging cameras. Based on the combination of high-definition cameras and thermal imaging cameras, real-time video streams and thermal imaging data of key areas of the construction site are collected.

[0032] Smart sensors include: vibration sensors, inclination sensors and stress sensors, which are used to monitor the equipment operating status, structural stress changes and vibration abnormalities in the construction area.

[0033] The construction equipment monitoring device collects operating parameters of construction equipment such as excavators and cranes through embedded sensor modules, including working hours, mechanical temperature and mechanical load conditions.

[0034] The personnel positioning device is based on ultra-wideband (UWB) positioning technology or GPS positioning technology to track the location and movement trajectory of construction workers on the construction site in real time. At the same time, it combines the construction site safety helmet wearing sensor to detect whether the personnel are wearing safety helmets.

[0035] In the above step S1, when collecting real-time multi-dimensional data of the construction site through multiple types of IoT terminal devices distributed on the construction site, it includes: Obtain the construction site coverage of the current construction site and determine the type of IoT terminal device based on the real-time multi-dimensional data to be collected; Obtain the node unit sensing distance of each type of IoT terminal equipment, and determine the setting location of each type of IoT terminal equipment within the coverage area of ​​the construction site according to the node unit sensing distance, and perform calibration; The corresponding types of IoT terminal devices are set at the calibration positions, and the coverage and collection range of the IoT terminal devices at each calibration position are calibrated, and the optimal data collection frequency of each type of IoT terminal device is determined.

[0036] Furthermore, when obtaining the sensing distance of the node unit of each type of IoT terminal device, it includes: According to the preset node spacing of each type of IoT terminal equipment and the preset sampling frequency of each type of IoT terminal equipment, the node unit sensing distance is obtained, as shown in Formula 1: (1); In the formula, The sensing distance of the node units of various types of IoT terminal devices, Preset node spacing for various types of IoT terminal devices. It is the initial collection scope of various types of IoT terminal devices. The actual collection range of various types of IoT terminal devices. Preset sampling frequencies for various types of IoT terminal devices. Provide the shortest coverage distance for data collection of various types of IoT terminal devices; When determining the optimal data collection frequency for each type of IoT terminal device, include: According to the coverage and collection range of each type of IoT terminal equipment at the calibrated position and the total collection range of each type of IoT terminal equipment within the coverage range of the construction site, the optimal data collection frequency is obtained, as shown in Formula 2: (2); In the formula, The optimal data collection frequency for various types of IoT terminal devices. The total collection range of various types of IoT terminal devices within the coverage area of ​​the construction site, It is the basic collection unit difference of various types of IoT terminal devices. The data collection limits for various types of IoT terminal devices are: Calibrate the repeated collection range between locations for various types of IoT terminal devices. Directed data collection for various types of IoT terminal devices.

[0037] In the above step S2, when joint optimization of node selection and multi-dimensional resource allocation is performed on real-time multi-dimensional data, it includes: By establishing The cluster configures a fixed resource allocation strategy for the collected real-time multi-dimensional data; according to The resource allocation strategy of the cluster configuration obtains the first state set and inputs it into the node selection model to obtain Node selection strategy; based on Node selection strategy selected The existing resource state of the node is used to obtain a second state set, which is input into a resource allocation model to obtain a multi-dimensional resource allocation strategy; Input the multi-dimensional resource allocation strategy into the node selection model as the third state set to adjust the model results; Let the node selection model and resource allocation model iterate alternately until the overall objective function reaches ; in, is the overall objective function, is the maximum allowable error of the setting; for The utility load status at the moment, for The utility load state at the moment is shown in Formula 3: (3); In the formula, is the weight of system utility, for The weight of the overall load balancing of the cluster, for The system utility at the moment, for At the moment The overall load balance of the cluster, for The system utility at the moment, for At the moment The overall load balance of the cluster, To get the maximum value.

[0038] Further, when obtaining the first state set, it includes: According to the CPU resources, memory resources and network resources consumed by the node and the CPU resources, memory resources and network resources possessed by the node, the available CPU resources, memory resources and network resources of the node are obtained, as shown in Formula 4: (4); In the formula, The CPU resources, memory resources, and network resources available to the node. is the CPU resources available for the node, is the memory resources available to the node, is the network resources available to the node, is the CPU resource of the node, is the memory resource of the node, is the network resources of the node, is the CPU resources consumed by the node, is the memory resources consumed by the node, The network resources consumed by the node; according to The resource allocation strategy, storage and communication resources of the cluster configuration, and the CPU, memory and network resources available to the node, obtain the first state set, as shown in Formula 5: (5); In the formula, is the first state set, for The CPU resources configured for the cluster, for The memory resources configured for the cluster, for The network resources configured by the cluster; In getting When selecting a node, the strategy includes: In all In the node, based on the first state set, select the node with the largest first reward function result The node is the final selected node, as shown in Formula 6 and Formula 7: (6); In the formula, Indicates that the first reward function with the largest result is selected node, Select a model for the node; The first reward function is shown in Formula 7: (7); In the formula, is the first reward function, for The overall load balance of the cluster; Among them, when When the overall load balance of the cluster is the smallest, the result of the first reward function is the largest.

[0039] Further, when obtaining the second state set, it includes: Get the selected The node's transmission rate benefit, computational benefit, and execution cost are used to obtain the second state set, as shown in Formula 8: (8); In the formula, is the second state set, For selected The node's transmission rate gain, For selected The computational income of the node, For selected The execution cost of the node; When obtaining a multi-dimensional resource allocation strategy, it includes: In the resource allocation process based on system utility, the resource allocation strategy with the largest second reward function result is selected as the multi-dimensional resource allocation strategy based on the second state set, as shown in Formula 9: (9); In the formula, represents the resource allocation strategy that selects the maximum result of the second reward function, Model for resource allocation; The second reward function is shown in Formula 10: (10); In the formula, is the second reward function, is the transmission rate benefit weight, To calculate the return weight, is the execution cost weight.

[0040] In the above step S2, when the data volume is optimized and transmitted to the data processing center, it includes: The real-time multi-dimensional data is integrated using the Horner function parameters to obtain integrated real-time multi-dimensional data, as shown in Formula 11: (11); In the formula, To integrate real-time multi-dimensional data, is the first Horner function parameter, is the second Horner function parameter, For real-time multidimensional data to be integrated, is the dimension of real-time multidimensional data; The optimized estimated value is obtained, and the data volume of the integrated real-time multi-dimensional data is optimized by the optimized estimated value to obtain the optimized real-time multi-dimensional data, as shown in Formula 12 and Formula 13: (12); (13); In the formula, is the estimated value of each dimension data, To optimize the estimate, It is optimized real-time multi-dimensional data; Transmitting the optimized real-time multi-dimensional data to the data processing center through the selected nodes and based on the determined multi-dimensional resource allocation strategy; Among them, the optimized estimated value is used to reduce the overall data volume.

[0041] In the above step S4, when extracting potential risk factors in the construction site, it includes: Will The subjective weighting method is combined with the objective weighting method of anti-entropy weighting to combine and weight the risk assessment index system to obtain the comprehensive weight of the index; The relative closeness of gray-scale correlation is obtained according to the comprehensive weight of the indicators; Determine the risk assessment level of each risk assessment indicator according to the relative proximity of grayscale association and sort them, and obtain the risk assessment indicators with risk assessment levels greater than the threshold from the sorting as potential risk indicators; Obtain real-time multi-dimensional data associated with potential risk indicators as potential risk factors.

[0042] Specifically, construction site risk assessment indicators include: personnel risk, technical risk, management risk and environmental risk.

[0043] Among them, personnel risks include: construction workers' skill level, construction workers' safety awareness, and construction workers' violations. Technical risks include: equipment failure, data security and privacy protection, system stability and compatibility. Management risks include: on-site management system and system, teamwork and information sharing, safety monitoring and protection measures. Environmental risks include: natural disasters, environmental pollution, transportation conditions, geological hydrology and meteorological conditions.

[0044] Based on the above construction site risk assessment indicators, a construction site risk assessment indicator system is constructed, such as Figure 2 shown.

[0045] Furthermore, when obtaining the relative proximity of grayscale association, it includes: Combine the comprehensive weights of the indicators with the standardized fuzzy decision matrix to obtain a weighted normalized matrix; Select the optimal indicator set as the reference sequence for comparison and evaluation of risk assessment indicators; A grey correlation coefficient matrix is ​​established based on the weighted normalization matrix, and positive ideal solutions and negative ideal solutions in the grey correlation coefficient matrix are obtained; Obtain the distances of the risk assessment index values ​​under each assessment target to the positive ideal solution and the negative ideal solution, and obtain the relative closeness of the grey correlation, as shown in Formula 14 and Formula 15: (14); In the formula, is the distance from the risk assessment index value under each assessment target to the positive ideal solution, is the distance from the risk assessment index value under each assessment target to the negative ideal solution, is the number of risk assessment indicators in the grey correlation coefficient matrix, is the first evaluation objectives, is the first The first evaluation target Risk assessment index value, For the A positive ideal solution, For the A negative ideal solution; (15); In the formula, is the relative closeness of grey relation, is the distance from the risk assessment index value in the reference sequence to the positive ideal solution under each assessment target, is the distance from the risk assessment index value in the reference sequence to the negative ideal solution under each assessment target; Among them, the grey correlation relative closeness is used to measure the closeness between the risk assessment index sequence and the reference sequence under each evaluation target; When determining the risk assessment level of each risk assessment indicator, it includes: Divide the preset evaluation score range into several sub-ranges, and assign a corresponding risk evaluation level to each sub-range; The grey relational relative proximity is mapped to obtain the corresponding integer evaluation value, and the risk evaluation level is determined according to the sub-interval where the integer evaluation value is located; The mapping relationship is obtained by mapping a given grey relational relative proximity to an integer evaluation value in a preset evaluation score range.

[0046] In the above step S5, when generating risk warning information, it includes: Constructing a training sample set ,in, For the The input value of the learning sample, For the The target value corresponding to the learning sample, is the number of learning samples in the training sample set; pass The mean method discretizes the training sample set. The theory simplifies the training sample set, deletes incomplete data, and sorts them from large to small to obtain the optimal attribute set; Reference the cross-validation method to select kernel function parameters and penalty parameters, and use Programming to solve the model parameters; Construct a construction site early warning model based on support vector machine, input potential risk factors into the construction site early warning model, and output risk warning information.

[0047] Furthermore, when constructing a construction site early warning model based on support vector machine, it includes: Use nonlinear mapping to map the optimal attribute set from the original space to the high-dimensional space, and perform linear regression in the high-dimensional space to obtain the regression function, as shown in Formula 16: (16); In the formula, is the top item coefficient, is the number of learning samples, As potential risk factors, is the value range of the attribute set; The regression function is transformed according to Lagrangian to obtain the dual form, as shown in formula 17: (17); In the formula, is the number of attributes in the attribute set, As potential risk factors, is the number of support vector machines, The attribute set properties, The attribute set The optimal solution for each attribute; The regression prediction function is obtained according to training and learning, as shown in Formula 18: (18); In the formula, is the regression prediction function, For the The input value of the learning sample, For the The input value of the learning sample, is the top item coefficient, is the kernel function; Combining the regression prediction function with the error detection function, we get the construction site early warning model, as shown in Formula 19: (19); In the formula, For the construction site early warning model, is the number of learning samples in the training sample set, is the regression prediction function, For the Error test results; Among them, the error check function is shown in formula 20: (20); In the formula, For the The actual value of the learning samples, For the The predicted value of the learning sample.

[0048] In the above step S6, when generating the construction site visualization data model, it includes: Construct a three-dimensional digital model of the construction site and several construction site equipment models, and combine the three-dimensional digital model of the construction site and several construction site equipment models to generate a construction site visualization data model; Among them, when constructing a three-dimensional digital model of the construction site, it includes: Use 3D digital models to simulate construction progress and predict progress plans. By integrating real-time multi-dimensional data, potential risk factors and risk warning information, we can focus on monitoring risk areas, automatically compare planned progress with actual progress, and generate progress deviations. Associate the components in the 3D digital model with the risk operations involved in the potential risk factors and identify key points, and form a safety sandbox based on the 3D digital model from the perspective of safety management; The components in the 3D digital model are associated with the standard processes used in the construction process, and the logos, instructions, key points of standard process practices and picture information are loaded into the cloud. During the on-site construction process, the standard process is automatically pushed to provide visual guidance for construction; Automatically compare the real-time process parameters of risk operations with standard process parameters to generate process deviations; When building several site equipment models, including: Provide an equipment model library, and associate the equipment models in the equipment model library with the real-time data and standard data collected by the corresponding type of IoT terminal devices in the real-time multi-dimensional data, automatically compare the real-time data with the standard data, generate data deviations, and obtain several construction site equipment models.

[0049] In the above step S8, when generating the dynamic optimization path of the construction site environment and the construction process, it includes: Provide the site environment objective function and the construction process objective function, as shown in Formula 21 and Formula 22 respectively: (twenty one); In the formula, is the construction site environment objective function, For facilities and facilities The Euclidean distance between centroids, For the Facilities under construction and facilities The Euclidean distance between the centroids is The safety factor when is the number of construction stages, is the number of facilities; Among them, when two facilities are close to each other at the construction site, unsafe factors such as noise, uncomfortable temperature and pollution will be generated; The larger the value, the more potential safety hazards there are between the two facilities. The distance between the two facilities should be increased to increase the safety factor of the site. The smaller the value, the fewer potential safety hazards there are between the two facilities in the site. The facilities can be appropriately moved closer. (twenty two); In the formula, is the objective function of the construction process, For real-time progress Progress with goals During the construction phase The progress deviation value in For real-time progress Progress with goals The progress modification weight between For the The time taken for each construction phase, For construction stage To the construction stage The progress deviation value, For construction stage To the construction stage The progress modifier weight of is the progress number, is the number of construction stages; Array encoding is performed on the design variables, and the initial population is generated using random numbers. Each chromosome in the initial population is regarded as a feasible solution, and the feasible solution is substituted into the site environment objective function and the construction process objective function to obtain the objective function value; The selection operator is used to select some chromosomes to enter the mating pool, and the chromosomes in the mating pool are allowed to perform crossover operations according to the crossover probability to generate offspring individuals; Some individuals in the population are randomly mutated according to the mutation probability to detect whether the population meets the preset termination iteration conditions. If so, the population evolution is terminated and the dynamic optimization path of the construction site environment and construction process is output. If not, a new initial population is generated using random numbers.

[0050] The above-mentioned embodiments are only preferred embodiments of the present invention and cannot be used to limit the scope of protection of the present invention. Any non-substantial changes and substitutions made by technicians in this field on the basis of the present invention shall fall within the scope of protection required by the present invention.

Claims

1. A method for data collection and visualization management of smart construction sites based on the Internet of Things, characterized in that: The following steps are involved: Collect real-time multi-dimensional data of the construction site through various types of IoT terminal devices distributed on the construction site; Jointly optimizing node selection and multi-dimensional resource allocation for the real-time multi-dimensional data, and transmitting the data to a data processing center after optimizing the data volume; Identify and screen the risk assessment indicators of the construction site, and build a risk assessment indicator system for the construction site. Input the optimized real-time multi-dimensional data and risk assessment indicator system into the construction site analysis model for identification and extraction of potential risk factors in the construction site. Inputting the potential risk factors into the construction site early warning model for analysis to generate risk early warning information; Perform multimodal data fusion on real-time multidimensional data, potential risk factors and risk warning information, generate a construction site visualization data model, and conduct dynamic visualization monitoring and management of the construction site environment and construction process; Analyze the site visualization data model through the associated site optimization model to generate a dynamic optimization path for the site environment and construction process; Among them, when generating the dynamic optimization path of the construction site environment and the construction process, the dynamic optimization of the path is achieved through population generation and population evolution based on the objective function.

2. According to the method for data collection and visualization management of smart construction sites based on the Internet of Things according to claim 1, it is characterized in that: When collecting real-time multi-dimensional data on the construction site, it includes: Obtain the construction site coverage of the current construction site and determine the type of IoT terminal device based on the real-time multi-dimensional data to be collected; Obtaining the node unit sensing distance of each type of IoT terminal device, and determining the setting position of each type of IoT terminal device within the coverage range of the construction site according to the node unit sensing distance, and calibrating the same; The corresponding types of IoT terminal devices are set at the calibration positions, and the coverage and collection range of the IoT terminal devices at each calibration position are calibrated, and the optimal data collection frequency of each type of IoT terminal device is determined.

3. The method for data collection and visualization management of smart construction sites based on the Internet of Things according to claim 2 is characterized in that: When performing joint optimization of node selection and multi-dimensional resource allocation, it includes: By establishing The cluster configures a fixed resource allocation strategy for the collected real-time multi-dimensional data; according to The resource allocation strategy of the cluster configuration obtains the first state set and inputs it into the node selection model to obtain Node selection strategy; based on Node selection strategy selected The existing resource state of the node is used to obtain a second state set, which is input into a resource allocation model to obtain a multi-dimensional resource allocation strategy; Inputting the multi-dimensional resource allocation strategy into the node selection model as a third state set to adjust the model result; The node selection model and the resource allocation model are iterated alternately until the total objective function reaches the set maximum allowable error; Among them, in all In the node, based on the first state set, select the node with the largest first reward function result Node is the final selected node, and we get Node selection strategy; During the resource allocation process, a resource allocation strategy with the largest second reward function result is selected as the multi-dimensional resource allocation strategy based on the second state set.

4. The method for data collection and visualization management of smart construction sites based on the Internet of Things according to claim 3 is characterized in that: When data is optimized and transmitted to the data processing center, it includes: Integrating the real-time multi-dimensional data using Horner function parameters to obtain integrated real-time multi-dimensional data; Obtaining an optimized estimated value, and optimizing the data volume of the integrated real-time multi-dimensional data by using the optimized estimated value to obtain optimized real-time multi-dimensional data; The optimized real-time multi-dimensional data is transmitted to the data processing center through the selected nodes and based on the determined multi-dimensional resource allocation strategy.

5. The method for data collection and visualization management of smart construction sites based on the Internet of Things according to any one of claims 1 to 4, characterized in that: When extracting potential risk factors on the construction site, include: Will The subjective weighting method is combined with the objective weighting method of anti-entropy weighting to combine and weight the risk assessment index system to obtain the comprehensive weight of the index; According to the comprehensive weight of the indicators, the relative closeness of the grayscale association is obtained; Determine the risk assessment level of each risk assessment indicator according to the relative proximity of the grayscale association and sort them, and obtain the risk assessment indicators with risk assessment levels greater than a threshold from the sorting as potential risk indicators; Real-time multi-dimensional data associated with the potential risk indicator is obtained as the potential risk factor.

6. The method for data collection and visualization management of smart construction sites based on the Internet of Things according to claim 5 is characterized in that: When obtaining the relative proximity of grayscale association, it includes: The comprehensive weights of the indicators are combined with the standardized fuzzy decision matrix to obtain a weighted normalized matrix; Select the optimal indicator set as the reference sequence for comparison and evaluation of risk assessment indicators; Establishing a grey correlation coefficient matrix based on the weighted normalization matrix, and obtaining a positive ideal solution and a negative ideal solution in the grey correlation coefficient matrix; The distances from the risk assessment index values ​​under each assessment target to the positive ideal solution and the negative ideal solution are obtained, and the relative proximity of the grey correlation is obtained.

7. The method for data collection and visualization management of smart construction sites based on the Internet of Things according to claim 6 is characterized in that: When generating risk warning information, include: Constructing a training sample set, and discretizing and simplifying the training sample set, deleting incomplete data therein, and sorting the training sample set from large to small to obtain an optimal attribute set; The cross-validation method is used to select kernel function parameters and penalty parameters, and the Programming to solve the model parameters; A construction site early warning model based on support vector machine is constructed, the potential risk factors are input into the construction site early warning model, and risk early warning information is output.

8. The method for data collection and visualization management of smart construction sites based on the Internet of Things according to claim 7 is characterized in that: When building a construction site early warning model based on support vector machine, it includes: Using nonlinear mapping to map the optimal attribute set from the original space to a high-dimensional space, and performing linear regression in the high-dimensional space to obtain a regression function; Transforming the regression function according to Lagrangian to obtain a dual form, and obtaining a regression prediction function according to training and learning; The regression prediction function is combined with the error detection function to obtain the construction site early warning model.

9. The method for data collection and visualization management of smart construction sites based on the Internet of Things according to claim 1 is characterized in that: When generating a construction site visualization data model, include: Constructing a three-dimensional digital model of the construction site and several construction site equipment models, and combining them to generate the construction site visualization data model; Among them, when constructing a three-dimensional digital model of the construction site, it includes: Use 3D digital models to simulate construction progress and predict progress plans. By integrating real-time multi-dimensional data, potential risk factors and risk warning information, we can focus on monitoring risk areas, automatically compare planned progress with actual progress, and generate progress deviations. The components in the 3D digital model are associated with the risk operations involved in the potential risk factors and the standard processes used in the construction procedures. The real-time process parameters of the risk operations are automatically compared with the standard process parameters to generate process deviations.

10. The method for data collection and visualization management of smart construction sites based on the Internet of Things according to claim 9 is characterized in that: When dynamic path optimization is achieved through population generation and population evolution based on the objective function, it includes: Providing a site environmental objective function based on the Euclidean distance between facility centroids and safety factors; Provide construction process objective function based on progress deviation value and progress correction weight; Performing array encoding on the design variables, generating an initial population using random numbers, treating each chromosome in the initial population as a feasible solution, and substituting the feasible solution into the construction site environment objective function and the construction process objective function to obtain the objective function value; The selection operator is used to select some chromosomes to enter the mating pool, and the chromosomes in the mating pool are allowed to perform crossover operations according to the crossover probability to generate offspring individuals; Some individuals in the population are randomly mutated according to the mutation probability to detect whether the population meets the preset termination iteration conditions. If so, the population evolution is terminated and the dynamic optimization path of the construction site environment and construction process is output. If not, a new initial population is generated using random numbers.

Images