Steel structure construction component tracking and tracing method based on Internet of Things

Abstract

The invention relates to the technical field of steel structure construction traceability, and discloses a steel structure construction component tracking and traceability method based on the Internet of Things. The method comprises the following steps: acquiring historical violation data and normal traffic data under multiple types of traffic scenes, and forming a standardized scene data set through format specification and interference filtering processing; then, illegal features and passing features in the data are converted into a preset feature space through a feature extraction and conversion module, and scene feature vectors are generated; and constructing a violation triggering judgment model based on the feature vectors, and obtaining a cross-scene unified feature identifier by matching difference features. And further training a violation probability prediction model, and predicting the violation occurrence probability by using the cross-scene unified feature identifier and the real-time data of the target scene. And calculating a scene-level violation identification threshold according to historical violation data probability distribution, judging whether to trigger snapshot or not by combining with a real-time prediction probability, and updating model parameters regularly. According to the method, the violation snapshot accuracy and the scene applicability are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of steel structure construction traceability, and in particular to a steel structure construction component tracking and tracing method based on the Internet of Things. Background Art

[0002] Steel structures are widely used in modern construction projects, and the tracking and tracing of their construction components is crucial to project quality, progress, and safety management. Traditional methods for tracking and tracing steel structure construction components have numerous flaws. Information collection primarily relies on manual recording, which is not only inefficient but also prone to human error. This results in incomplete and untimely component perception and construction process information, failing to accurately reflect the actual status of the components.

[0003] Most existing tracking methods lack deep correlation and multi-dimensional analysis of information, making it difficult to build comprehensive and accurate tracking models and effectively capture the evolution of components during construction. Furthermore, traditional methods often lack dynamic adjustment mechanisms and cannot promptly update tracking thresholds based on actual construction conditions. This leads to inaccurate judgments on component traceability status and difficulty in timely identification and resolution of issues. While some IoT-based tracking methods have been proposed with the development of IoT technology, most of these methods focus solely on single-dimensional information collection and processing, failing to fully leverage the advantages of IoT technology to achieve multi-dimensional, full-process tracking and traceability of components.

[0004] There is an urgent need for a steel structure construction component tracking and tracing method that can comprehensively collect information, deeply analyze correlations, and dynamically adjust thresholds to improve the efficiency and quality of construction management and ensure the smooth progress of the project. Summary of the Invention

[0005] The purpose of the present invention is to provide a steel structure construction component tracking and tracing method based on the Internet of Things to solve the problems raised in the above background technology.

[0006] To achieve the above objectives, the present invention provides a steel structure construction component tracking and tracing method based on the Internet of Things, the method comprising: Collect IoT perception information and construction process information of steel structure construction components, associate the perception information with the process information, and build a component tracking feature set; Based on the component tracking feature set, tracking nodes are defined. According to the construction process information, associated edges and edge weights are defined to build a multi-dimensional tracking network. Based on the temporal tracking network model, the evolution law of tracking nodes in the multi-dimensional tracking network is learned to generate component tracking feature vectors. Train a traceability prediction model for predicting the real-time traceability status of components. This model uses component tracking feature vectors and real-time construction data as input data, and component traceability status indicators as output data to predict the traceability status indicators of steel structure construction components at the current construction stage. Based on the component tracking feature vector, all construction states are divided into several state groups. The group traceability threshold is calculated for each state group. Based on the group traceability threshold and the matching degree between the component tracking feature vector of each construction state and the typical vector of the state group to which it belongs, the dynamic traceability threshold of the construction state is calculated. The traceability status indicator output by the traceability prediction model is compared with the dynamic traceability threshold of the construction status. If the traceability status indicator exceeds the dynamic traceability threshold, a corresponding tracking adjustment instruction is generated, and the Internet of Things is triggered to execute terminal actions and regularly update the dynamic traceability threshold of each construction status.

[0007] Preferably, the specific method of collecting IoT perception information and construction process information of steel structure construction components, and performing feature association between the perception information and the process information to construct a component tracking feature set is as follows: Acquire sensory information and construction process information from IoT sensor terminals of steel structure construction components. The sensory information includes: component unique identification code, installation coordinate positioning, and material detection data; the construction process information includes: transportation timestamp, hoisting sequence record, and welding quality log; The characteristic attributes of the perception information and the construction process information are extracted and associated to obtain the component tracking feature set of each construction status.

[0008] Preferably, the specific method of defining tracking nodes based on the component tracking feature set, defining associated edges and edge weights according to the construction process information, and constructing a multi-dimensional tracking network is as follows: Taking the perception information as the basic attribute of the tracking node, the tracking node is constructed based on the component tracking feature set and the corresponding basic attributes of each tracking node; The correlation edges between tracking nodes are constructed based on the construction process information, and the temporal correlation between tracking nodes is calculated based on the temporal record data in the construction process information. The temporal record data of tracking node p is the information sequence within the continuous construction period, and the temporal record data of tracking node q is the information sequence within the corresponding period. The time series alignment algorithm is used to assign adjustment weights to the correlation of temporal data in different periods based on the time decay parameter, and the temporal correlation between tracking node p and tracking node q is calculated. Converting the temporal correlation into the tracking correlation; Based on the historical collaborative construction records of the components, the adjacent node set of tracking nodes p and q in the collaborative construction network is obtained. Based on the adjacent node set, the collaborative matching index between tracking nodes is calculated. Different weights are assigned to nodes with different collaborative distances to obtain the weighted collaborative matching index. The weighted collaborative matching index is normalized to the interval [0,1] to obtain the collaborative correlation degree between tracking nodes p and q. The edge weights of the associated edges between tracking nodes p and q are calculated based on the tracking association degree and collaborative association degree, and the tracking nodes, associated edges and their edge weights are used to form a multi-dimensional tracking network.

[0009] Preferably, the training method of the time series tracking network model is: Step B1: Construct a time series tracking network model, including a node encoding layer, a time series memory layer, an attention module, an output layer, and a parameter optimization module. The model takes the multi-dimensional tracking network at each time point as input and the evolutionary characteristics of the tracking nodes at each time point as output. The node encoding layer is used to map the features of each tracking node and associated edges to a unified feature space. Based on the node encoding layer, a temporal memory layer is introduced to capture the dynamic evolution pattern of the tracking features. It takes the multi-dimensional tracking network sequence at each time point as input, outputs the evolution features of the tracking node at each time point, and describes the trajectory of the evolution features of the tracking node over time. The attention module is used to adaptively assign the influence weights of different adjacent nodes to highlight the role of key adjacent nodes. The output layer outputs the evolution features of each tracking node at each time point. The parameter optimization module is used to calculate the gradient of the loss function with respect to the model parameters using the backpropagation algorithm. Step B2: For each tracking node p, construct a positive sample set and a negative sample set based on the tracking correlation and collaborative correlation, set the correlation threshold and the matching threshold, and take the tracking nodes whose tracking correlation with the tracking node p is greater than the correlation threshold or whose collaborative correlation is greater than the matching threshold as positive samples. The tracking nodes that do not belong to the positive sample set are included in the negative sample set. The positive and negative samples are used as training data. An unsupervised training method is adopted, with the nodes in the multidimensional tracking network as input and the evolution characteristics of the nodes as output. The time series tracking network model is applied to the multidimensional tracking network, and the evolution characteristics of all tracking nodes at each time point t are generated by forward propagation. Step B3: Using the evolutionary features of the tracking node p at time point t, calculate the feature similarity between the tracking node p and the tracking nodes in the positive sample set and the negative sample set. Based on the feature similarity, calculate the contrast loss function at the current time point t. Average the contrast loss functions of all tracking nodes and all time points to obtain the final loss function. Minimizing the loss function is the training goal. Step B4: Calculate the gradient of the loss function with respect to the model parameters through the back-propagation algorithm and use the optimizer to update the model parameters; Step B5: Repeat steps B2 to B4 until the loss function converges.

[0010] Preferably, the specific method of learning the evolution law of tracking nodes in the multi-dimensional tracking network based on the time-series tracking network model and generating the component tracking feature vector is: On a multi-dimensional tracking network, a temporal tracking network model is used to learn the evolutionary characteristics of tracking nodes. The evolutionary characteristics of tracking node p at the nth layer are obtained through forward propagation. An attention module is introduced to calculate the attention weights of adjacent tracking nodes with different collaborative relationships for tracking node p. In the temporal memory layer, a gated memory unit aggregation function is defined to fuse the evolutionary features of the adjacent tracking node q at the current time point and the evolutionary features of the tracking node p at the previous time point to obtain the evolutionary features of the tracking node p at the current time point t. The evolution characteristics of the tracking node p at time point t are spliced ​​with the perception information of the tracking node to obtain the component tracking feature vector of the tracking node p.

[0011] Preferably, the training of the traceability prediction model for predicting the real-time traceability status of components uses the component tracking feature vector and real-time construction data as input data, and uses the component traceability status index as output data. The specific method for predicting the traceability status index of the steel structure construction component at the current construction stage is: Obtain a batch of historical construction records of steel structure components, extract their component tracking feature vectors, mark the status indicators of abnormal traceability as abnormal traceability labels, construct training samples based on the component tracking feature vectors and abnormal traceability labels, use the component tracking feature vectors and real-time construction data as input data, and the abnormal traceability labels as output data to train the traceability prediction model; For each training sample, the prediction target is to accurately predict the anomaly traceability label. The cross entropy loss function is used as the loss function for the training model. The training target is to minimize the value of the loss function. The training is completed when the loss function converges. Using the trained traceability prediction model, for the current real-time construction status, the component tracking feature vector and real-time construction data of the corresponding tracking node are extracted as input data, and the traceability status indicator of the real-time construction status is output.

[0012] Preferably, all construction states are divided into several state groups based on the component tracking feature vector, and a group traceability threshold is calculated for each state group. Based on the group traceability threshold and the matching degree between the component tracking feature vector of each construction state and the typical vector of the state group to which it belongs, a specific method for calculating the dynamic traceability threshold of the construction state is as follows: Based on the component tracking feature vector, all construction states are divided into several state groups through density clustering algorithm; For each state group, the component tracking feature vector and historical construction records of the construction state are extracted. The component tracking feature vector and each historical construction data are used as input data to input the traceability prediction model to obtain the traceability state index of each historical construction. Collect statistics on the traceability status indicators of each historical construction of the construction state to obtain the traceability status indicator distribution of the construction state; calculate the statistical characteristics of the traceability status indicator distribution of the construction state, including the average value and dispersion; Calculate the statistical characteristics of the traceability status indicator distribution of each construction status in the status group to obtain the group traceability threshold of the status group; For each construction state, the matching degree between its component tracking feature vector and the typical vector of the state group to which it belongs is calculated, where the component tracking feature vector represents the feature set of the construction state, and the typical vector represents the typical feature set of the state group; The dynamic traceability threshold of the construction state is calculated based on the matching degree between the component tracking feature vector of the construction state and the typical vector of the state group to which it belongs and the group traceability threshold.

[0013] Preferably, the specific method for regularly updating the dynamic traceability threshold of each construction status is: Set a fixed update cycle. During each update cycle, collect the deviation data between the actual traceability status indicators and the dynamic traceability thresholds for all construction statuses within that cycle. The number and magnitude of the threshold values ​​exceeded in the statistical deviation data are combined with the changing trend of the component tracking feature vector in the current state group to adjust the dynamic traceability threshold of the construction state; If the typical vector of a state group deviates significantly due to component loss or construction process adjustment, the dynamic traceability threshold of all construction states in the group is recalculated.

[0014] Preferably, the specific method of generating the corresponding tracking adjustment instruction is: Based on the type and degree of the traceability status indicator exceeding the dynamic traceability threshold, the preset adjustment strategy table is called; If the excess type is missing identification, an adjustment instruction for component identification re-recording is generated; if it is positioning deviation, an adjustment instruction for installation coordinate correction is generated; if it is material abnormality, an adjustment instruction for component batch replacement is generated.

[0015] Preferably, the tracking adjustment instruction includes a specific parameter adjustment value and an execution time node.

[0016] Compared with the prior art, the present invention has the following beneficial effects: By collecting IoT-sensed information and construction process information from steel structure construction components, correlating the two, and constructing a component tracking feature set, we can comprehensively and accurately obtain various information about components during the construction process, providing a rich and reliable data foundation for subsequent tracking and tracing. By defining tracking nodes based on the component tracking feature set and defining associated edges and edge weights based on construction process information, we construct a multi-dimensional tracking network. This allows the tracking network to fully reflect the temporal and collaborative associations between nodes during the component construction process, thereby more accurately depicting the evolution of components.

[0017] By using a time-series tracking network model to learn the evolution patterns of tracking nodes in a multi-dimensional tracking network and generate component tracking feature vectors, we can deeply explore the dynamic changes of components during construction, providing strong support for predicting the traceability status of components. A traceability prediction model for predicting the real-time traceability status of components is trained. Using component tracking feature vectors and real-time construction data as input, it can accurately predict the traceability status indicators of components in the current construction phase in real time, providing timely and accurate decision-making for construction management.

[0018] Based on the component tracking feature vector, all construction states are divided into several state groups. The group traceability threshold and the dynamic traceability threshold for each construction state are calculated. This allows the traceability threshold to be dynamically adjusted according to changes in the construction state, improving the accuracy and adaptability of traceability judgments. The traceability status indicator output by the traceability prediction model is compared with the dynamic traceability threshold of the construction state. When the threshold exceeds the threshold, a corresponding tracking adjustment instruction is generated and the IoT terminal execution action is triggered. This can promptly identify and resolve problems that arise during the component construction process, ensuring the construction quality and progress of the component. Regularly updating the dynamic traceability threshold for each construction state enables traceability management to always adapt to various changes in the construction process and maintain the accuracy and effectiveness of traceability judgments.

[0019] Through the synergistic effect of the above links, this method realizes the full-process, multi-dimensional tracking and tracing and dynamic management of steel structure construction components, improves the efficiency and accuracy of construction management, reduces construction costs and risks, and provides strong support for the smooth construction and quality assurance of steel structure projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a working principle diagram of the steel structure construction component tracking and tracing method based on the Internet of Things according to the present invention; Figure 2 Flowchart constructed for multi-dimensional tracking network; Figure 3 Flowchart for training the traceability prediction model; Figure 4 Flowchart for dynamic traceability threshold update. DETAILED DESCRIPTION

[0021] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0022] See also Figure 1-Figure 4 The present invention provides a steel structure construction component tracking and tracing method based on the Internet of Things. The specific implementation steps are as follows: Collect IoT perception information and construction process information of steel structure construction components, associate the perception information with the process information, and build a component tracking feature set; Based on the component tracking feature set, tracking nodes are defined. According to the construction process information, associated edges and edge weights are defined to build a multi-dimensional tracking network. Based on the temporal tracking network model, the evolution law of tracking nodes in the multi-dimensional tracking network is learned to generate component tracking feature vectors. Train a traceability prediction model for predicting the real-time traceability status of components. This model uses component tracking feature vectors and real-time construction data as input data, and component traceability status indicators as output data to predict the traceability status indicators of steel structure construction components at the current construction stage. Based on the component tracking feature vector, all construction states are divided into several state groups. The group traceability threshold is calculated for each state group. Based on the group traceability threshold and the matching degree between the component tracking feature vector of each construction state and the typical vector of the state group to which it belongs, the dynamic traceability threshold of the construction state is calculated. The traceability status indicator output by the traceability prediction model is compared with the dynamic traceability threshold of the construction status. If the traceability status indicator exceeds the dynamic traceability threshold, a corresponding tracking adjustment instruction is generated, and the Internet of Things is triggered to execute terminal actions and regularly update the dynamic traceability threshold of each construction status.

[0023] Example 1

[0024] When constructing a component tracking feature set, it is necessary to obtain sensory information and construction process information from IoT sensor terminals on steel structure construction components. This sensory information includes component unique identification codes, installation coordinate locations, and material inspection data. The component unique identification code is generated by an RFID tag. This tag uses specific coding rules to assign a globally unique identity to each component. For example, the EPC coding system, which includes fields such as a manufacturer identification code, object classification code, and serial number, enables precise differentiation between different components. Installation coordinate locations are obtained using the Beidou satellite navigation system. This system uses signal transmission and positioning from multiple satellites to accurately locate component positions on the construction site to the centimeter level. This is achieved by installing a Beidou positioning module on the component, which receives satellite signals in real time and calculates three-dimensional coordinates (X, Y, and Z). Material inspection data is collected using specialized equipment such as spectrometers. Spectrometers illuminate the component surface with light of a specific wavelength and analyze the spectral characteristics of the reflected light to determine the component's material composition, such as the content of elements such as iron, carbon, and manganese, as well as mechanical properties such as strength and hardness.

[0025] Construction process information includes transportation timestamps, lifting sequence records, and welding quality logs. Transportation timestamps are recorded by the logistics management system. When a component passes through various logistics nodes during transportation, the system will automatically collect and store time information accurate to the second, such as the time the component departs from the warehouse, the time it arrives at the construction site, etc. The lifting sequence records are entered by construction personnel through handheld terminals. When lifting components, construction personnel enter the lifting number and corresponding lifting time of each component in sequence according to the construction plan and actual site conditions, and record the lifting sequence of the components in detail. The welding quality log is automatically generated by the welding quality monitoring equipment. The equipment monitors the welding current, voltage, temperature and other parameters in real time during the welding process, such as the size of the welding current, the fluctuation of the voltage, the temperature change curve of the welding area, etc., and records these parameters in chronological order to form a complete welding quality log.

[0026] Extract characteristic attributes of the perception information and construction process information and associate them. For the unique identification code of the component in the perception information, extract characteristic attributes such as the manufacturer identification code, object classification code, and serial number; extract the specific X, Y, and Z coordinate values ​​from the installation coordinate positioning; extract characteristic attributes such as the content ratio of each element, strength value, and hardness value from the material testing data. Extract the specific year, month, day, hour, minute, and second time information from the transportation timestamp in the construction process information; extract the hoisting number and corresponding hoisting time from the hoisting sequence record; and extract characteristic attributes such as the specific value of the welding current, the specific value of the voltage, and the temperature variation range from the welding quality log.

[0027] These extracted characteristic attributes are associated with each other, with the unique identification code of the component as the core, and are associated with the X, Y, and Z coordinate values ​​of the corresponding installation coordinate positioning to clarify the specific location of the component in the construction site; associated with the characteristic attributes such as the element content, strength, and hardness of the material test data to understand the material characteristics of the component; associated with the specific time information of the transportation timestamp to grasp the transportation time node of the component; associated with the lifting number and lifting time recorded in the lifting sequence to understand the lifting sequence and time of the component; associated with the welding current, voltage, temperature and other parameters in the welding quality log to know the quality of the component during the welding process.

[0028] In this way, a component tracking feature set is formed for each construction state. For example, in the transportation stage, the component tracking feature set of a component contains its unique identification code, the specific time of the transportation timestamp, the coordinate value of the warehouse in the installation coordinate positioning, material detection data and other information; in the hoisting stage, the component tracking feature set of the component contains its unique identification code, the hoisting number and time recorded in the hoisting sequence, the coordinate value of the hoisting position in the installation coordinate positioning, the transportation timestamp and other information. The component tracking feature set of each construction state covers all the key information of the component in that construction stage. This information is interrelated and constitutes a complete feature data set, which provides rich feature data support for the subsequent tracking and tracing of the component, so that in the subsequent tracking process, these feature data can be used to accurately understand the status and information of the component at each construction stage.

[0029] Example 2

[0030] When building a multi-dimensional tracking network, sensor information is used as the fundamental attribute of tracking nodes. Each tracking node consists of its component tracking feature set and corresponding fundamental attributes. For example, the sensor information for a steel structure construction component contains a unique identifier code: "GS-001-20250702-001." This code is generated using specific rules, where "GS" represents the steel structure, "001" is the manufacturer code, "20250702" indicates the production date, and "001" is the serial number, ensuring uniqueness. Installation coordinates are obtained using the Beidou navigation system, resulting in three-dimensional coordinates (100.234, 200.567, 10.345), accurately reflecting the component's location at the construction site. Material inspection data, collected by a spectrometer, indicates that the component is Q355B steel, with a carbon content of 0.2%, a manganese content of 1.4%, a yield strength of 355 MPa, and a tensile strength of 510 MPa. These perception information constitute the basic attributes of the tracking node, and together with the component tracking feature set formed by the construction process information of the component in the transportation, lifting, welding and other stages, constitute a complete tracking node.

[0031] Based on construction process information, the association edges between tracking nodes are constructed. First, the temporal association degree is calculated. Taking tracking nodes p and q as examples, the time series records of tracking node p represent a sequence of information within a continuous construction period. For example, during the transportation phase, from 8:00 to 9:30 on July 2, 2025, the transportation timestamps are 8:00 (departure from the warehouse), 8:30 (passing through node A), and 9:30 (arrival at the site), along with the corresponding transportation vehicle information. The time series records of tracking node q during the corresponding period are 8:15 (another component departs from the warehouse), 9:00 (passing through node A), and 9:45 (arrival at the site). A time series alignment algorithm (such as the dynamic time warping algorithm) is used to align these two time series records, aligning events from different periods in chronological order. The association degree of the time series data in different periods is then adjusted based on a time decay parameter. The time decay parameter can be set so that the weight decreases exponentially with increasing time intervals. For example, the weight is multiplied by 0.8 for every hour from the current time. In this way, the temporal correlation between tracking nodes p and q during the transportation period is calculated. During the specific calculation, the time interval of each event is first determined, and then the weight is adjusted according to the attenuation parameter. Finally, the correlation of all events is combined to obtain the overall temporal correlation.

[0032] The timing correlation is converted into the tracking correlation. The conversion method can use a linear function, such as tracking correlation = timing correlation × 0.7 + 0.3 (where 0.7 and 0.3 are empirical coefficients that can be adjusted according to actual construction conditions), so that the tracking correlation is in the range of [0,1] to facilitate subsequent calculations.

[0033] The collaborative association degree is calculated. Based on the historical collaborative construction records of the component, the set of neighboring nodes for tracking nodes p and q in the collaborative construction network is obtained. The neighboring node set includes other nodes that have collaborated with p and q. For example, during the lifting phase, nodes p and q may have collaborative relationships with the lifting equipment node, the command node, and other nodes. Based on the neighboring node set, the collaborative matching index between the tracking nodes is calculated. The calculation of the collaborative matching index takes into account factors such as the number of common neighbors and the frequency of collaborative construction. For example, if p and q have three common neighbors and have collaborated in six of the past ten construction sessions, the collaborative matching index can be set as (3 / maximum number of common neighbors) × 0.5 + (6 / 10) × 0.5. Different weights are assigned to nodes with different collaborative distances: nodes with close collaborative distances (such as directly adjacent collaborative nodes) are given a weight of 0.8, while nodes with long collaborative distances (such as indirectly collaborative nodes) are given a weight of 0.2. This yields a weighted collaborative matching index. The weighted collaborative matching index is then mapped to the [0,1] interval through a normalization function, for example, using the Min-Max normalization method, to obtain the collaborative correlation between the tracking nodes p and q.

[0034] The edge weights of associated edges are calculated based on the tracking association degree and the collaborative association degree. Edge weights are calculated using a weighted summation method, for example, edge weight = tracking association degree × 0.6 + collaborative association degree × 0.4 (where 0.6 and 0.4 are weight coefficients, which can be adjusted based on the importance of temporal and collaborative associations during the construction process). All tracking nodes, constructed associated edges, and their corresponding edge weights are combined to form a multidimensional tracking network. In this network, each node contains component perception information and construction process characteristics. The edges between nodes reflect the connection strength through the temporal association degree and collaborative association degree, thus comprehensively reflecting the temporal relationship and collaborative operation relationship of the components during the construction process.

[0035] For example, in the hoisting process of steel structure construction, the tracking nodes of multiple components form time-series correlation edges through the time series of the hoisting order. At the same time, they form collaborative correlation edges because they share the same crane. The edge weights are calculated based on their respective correlation degrees. The multi-dimensional tracking network constructed in this way can characterize the mutual relationships during the construction process of components from multiple dimensions, and provide structured data support for the subsequent learning of the evolution laws of tracking nodes based on the time-series tracking network model, so that the model can better understand the state changes and mutual influences of components during the construction process. During the entire construction process, the parameter settings of each step are determined based on the actual construction situation and historical data to ensure that the network structure can accurately reflect the real relationship of the construction process.

[0036] Example 3: When training a temporal tracking network model, a temporal tracking network model must be constructed. This model consists of a node encoding layer, a temporal memory layer, an attention module, an output layer, and a parameter optimization module. The node encoding layer utilizes a multi-layer perceptron architecture, mapping the features of each tracking node and associated edges into a unified feature space. For example, features such as the component unique identifier and installation coordinates of a tracking node are converted into feature vectors of the same dimension through linear transformations and activation functions in the node encoding layer. Building upon the node encoding layer, a temporal memory layer is introduced. This layer utilizes a long short-term memory (LSTM) architecture. This layer takes the multi-dimensional tracking network sequence at each time point as input, stores historical information through memory cells, and thus captures the dynamic evolution of tracking features. It outputs the evolutionary features of the tracking node at each time point, depicting the trajectory of the tracking node's evolutionary features over time. The attention module adaptively assigns influence weights to different neighboring nodes and highlights the role of key neighboring nodes by calculating metrics such as feature similarity between neighboring nodes and the current node. The output layer is used to output the evolution characteristics of each tracking node at each time point, and the parameter optimization module uses the back propagation algorithm to calculate the gradient of the loss function with respect to the model parameters.

[0037] For each tracking node p, it is necessary to construct a positive sample set and a negative sample set based on the tracking correlation and collaborative correlation. First, set the correlation threshold θ1 and the matching threshold θ2. These two thresholds can be determined based on the construction history data and actual needs. Tracking nodes with a tracking correlation greater than θ1 or a collaborative correlation greater than θ2 with the tracking node p are taken as positive samples, and tracking nodes that do not belong to the positive sample set are included in the negative sample set. Then, the positive and negative samples are used as training data, and an unsupervised training method is adopted. The nodes in the multidimensional tracking network are used as input, and the evolutionary characteristics of the nodes are used as output. The time series tracking network model is applied to the multidimensional tracking network, and the evolutionary characteristics of all tracking nodes at each time point t are generated through forward propagation.

[0038] Using the evolutionary features of the tracking node p at time point t, calculate the feature similarity between the tracking node p and the tracking nodes in the positive sample set and the negative sample set. Feature similarity is calculated using cosine similarity, and the formula is: in, Represents the evolution feature vector of the tracking node p and the feature vector of the tracking node q The cosine similarity of is the dot product of two vectors; and are vectors and The L2 norm of .

[0039] Calculate the contrast loss function at the current time point t based on feature similarity. The expression of the contrast loss function is:

[0040] in, is the contrast loss function for tracking node p at time point t; is the total number of samples; is the label, when q is a positive sample ,otherwise ; To track the characteristic distance between nodes p and q, ; The preset distance margin.

[0041] The final loss function is obtained by averaging the comparative loss functions of all tracking nodes and all time points, with minimizing this loss function as the training objective. The gradient of the loss function with respect to the model parameters is calculated using the backpropagation algorithm, and the model parameters are updated using an optimizer (such as stochastic gradient descent).

[0042] Taking a steel structure construction project as an example, assume there are 100 tracking nodes and time points t range from 1 to 10. For each tracking node p, at time point t=1, its positive sample set may contain 20 tracking nodes, and its negative sample set may contain 80 tracking nodes. After obtaining the evolutionary characteristics of each node through forward propagation, the feature similarity of each positive sample and negative sample with p is calculated, and then substituted into the comparative loss function to calculate the loss value. Assuming the loss value calculated at this time is 0.8, after backpropagation and parameter update, the loss value may drop to 0.75 when calculated again at time point t=1. Repeat the above steps of constructing the sample set, forward propagation, calculating the loss function, and backpropagation to update the parameters until the loss function converges. For example, when the change in the loss function is less than 0.001 in 5 consecutive iterations, the model training is considered complete.

[0043] Throughout the training process, the parameters of the node encoding layer, the weights of the temporal memory layer, and the coefficients of the attention module are continuously adjusted with each iteration, enabling the model to better learn the evolutionary patterns of tracking nodes in the multi-dimensional tracking network. For example, the weights of the forget gate in the temporal memory layer are adjusted based on feedback from the loss function to more appropriately forget unimportant historical information while retaining key evolutionary features. The weight coefficients in the attention module are also gradually optimized, allowing the model to more accurately identify adjacent nodes that have a significant impact on the evolution of the current node, thereby improving the model's training effectiveness.

[0044] Example 4: When generating component tracking feature vectors on a multi-dimensional tracking network, it's necessary to leverage a trained temporal tracking network model to learn the evolutionary characteristics of tracking nodes. For example, a steel beam component in a steel structure construction project is divided into multiple tracking nodes during construction, each corresponding to a different construction phase, such as raw material delivery, processing, transportation, hoisting, and welding.

[0045] The multi-dimensional tracking network is fed into a trained temporal tracking network model. Through forward propagation, the evolutionary features of tracking node p at layer n are obtained. Assuming n = 3, the evolutionary features at layer 3 now incorporate multiple layers of feature information about the node in the network, encompassing everything from bottom-level perceptual information to higher-level temporal correlation and collaborative correlation features. For example, for a tracking node p in the transportation phase, its layer 3 evolutionary features might include the temporal features of the transportation timestamp, collaborative correlation features with other transportation component nodes, and abstract representations of basic features such as the component's material and identification.

[0046] An attention module is introduced to calculate the attention weights of adjacent tracking nodes with different collaborative relationships for the tracking node p. Adjacent tracking nodes include other component nodes, transport vehicle nodes, logistics nodes, etc. that have collaborative relationships with the component node during transportation. The attention module determines the weight of each adjacent node based on factors such as the collaborative correlation between these adjacent nodes and the tracking node p, and historical collaborative construction records. For example, the component node q that is in the same transport batch as the tracking node p and has been collaboratively transported multiple times has a high collaborative correlation, and the attention module will assign it a higher attention weight, such as 0.7; while the component node r that is only occasionally collaboratively transported with the tracking node p has a low collaborative correlation and an attention weight of perhaps 0.3. In this way, the model can focus on adjacent nodes that have a greater impact on the evolution of the tracking node p.

[0047] In the temporal memory layer, a gated memory unit aggregation function is defined to fuse the evolutionary features of the adjacent tracking node q at the current time point with the evolutionary features of the tracking node p at the previous time point. Taking time point t as an example, the evolutionary features of the adjacent tracking node q at the current time point contain information about its state at time t, such as the change in transportation status from time point t-1 to time t. The evolutionary features of the tracking node p at the previous time point record the state of p at time t-1. The gated memory unit aggregation function controls the inflow and outflow of information through a gating-like mechanism. For example, the input gate determines how much information from the evolutionary features of the current adjacent node q can flow into the evolutionary features of p at the current time point, the forget gate determines how much information from the evolutionary features of p at the previous time point should be forgotten, and the output gate determines what information should be output in the evolutionary features of p at the current time point. Through this fusion, the evolutionary features of the tracking node p at the current time point t are obtained. These features not only contain the latest information about the current adjacent nodes but also retain key information from the historical evolution process. For example, at time t in the transportation phase, the evolutionary characteristics of the tracking node p integrate the transportation position change information of the adjacent node q at time t and the transportation status information of p at time t-1, thereby forming the complete transportation status evolution characteristics of p at time t.

[0048] The evolutionary characteristics of the tracking node p at time point t are spliced ​​with the perception information of the tracking node. The perception information of the tracking node includes basic attributes such as the component unique identification code, installation coordinate positioning, and material detection data. For example, the evolutionary characteristics of the tracking node p is a 128-dimensional vector that contains the timing and collaborative characteristics of the transportation phase, while the component unique identification code in the perception information can be converted into a 64-dimensional one-hot encoding vector, the installation coordinate positioning is a 3-dimensional coordinate value vector, and the material detection data is a 16-dimensional material composition and mechanical property vector. These vectors are spliced ​​together in sequence to form a 128+64+3+16=211-dimensional component tracking feature vector. This vector combines the evolutionary characteristics of the tracking node p at time point t and its inherent perception information, and can comprehensively reflect all the key characteristics of the component in its current construction state.

[0049] Taking the transportation of a steel beam component from the fabrication plant to the construction site as an example, at time t1, when transportation begins, the component tracking feature vector generated through the above steps includes the component's unique identification code "GL-005-20250705," the initial installation coordinates (coordinates of the fabrication plant warehouse), material testing data (such as the composition and strength of Q355B steel), and evolutionary characteristics at the beginning of transportation (such as collaborative association features with the transport vehicle node and the initial temporal features of the transport timestamp). As the transportation process progresses, at time t2, the evolving characteristics of tracking node p incorporate the position change information and transportation status of other adjacent transport component nodes at time t2, as well as p's transportation status at time t1. These characteristics are then combined with current sensory information (such as the real-time updated installation coordinates, i.e., the position coordinates during transportation) to generate a new component tracking feature vector. In this way, the component tracking feature vector generated at each time point dynamically reflects the component's state changes during construction, providing accurate and comprehensive input features for the subsequent traceability prediction model, enabling the model to predict the component's real-time traceability status based on these feature vectors. Throughout the entire process, the processing of each step is based on actual construction data and model training results, ensuring that the generated component tracking feature vector can truly and effectively represent the construction status of the component.

[0050] Example 5: When calculating the dynamic traceability threshold, we must first divide all construction states into several state groups based on the component tracking feature vector through the density clustering algorithm. Taking a certain steel structure construction project as an example, the project contains various types of components, such as steel beams, steel columns, supports, etc. Each component will generate a large amount of construction status data during the construction process. The component tracking feature vectors corresponding to these data constitute the basis for clustering. The DBSCAN density clustering algorithm is used. This algorithm divides groups according to the density distribution of sample points, and divides sample points with accessible density in the feature space into the same state group. For example, clustering the feature vectors corresponding to the processing and manufacturing status, transportation status, and hoisting status of steel beams may result in processing and manufacturing status groups, transportation status groups, hoisting status groups, etc.

[0051] For each state group, it is necessary to extract the component tracking feature vectors and historical construction records of the construction state. Taking the transportation state group as an example, this group contains the construction status of all components during the transportation phase. Each construction state corresponds to a component tracking feature vector. It also contains the historical construction records of the component during the transportation phase, such as transportation time, transportation route, and environmental data during transportation. These component tracking feature vectors of the construction state and each historical construction data are used as input data and input into the trained traceability prediction model. The traceability prediction model will output the traceability status indicators for each historical construction. These traceability status indicators reflect the traceability of the construction state in the historical construction, such as transportation delays, component damage, etc.

[0052] Statistics are collected on the traceability status indicators of each historical construction of each construction state to obtain the traceability status indicator distribution of the construction state. For example, a component has 100 historical construction records in the transportation stage. Its traceability status indicators may include transportation time deviation, component positioning deviation, etc. Statistical analysis of these indicators may reveal that the transportation time deviation presents a normal distribution with a mean of 15 minutes and a standard deviation of 5 minutes. The statistical characteristics of the traceability status indicator distribution of the construction state are calculated, including the mean and dispersion. The mean can reflect the overall level of the traceability status indicator of the construction state, such as the mean value of the transportation time deviation is 15 minutes; the dispersion can be expressed by the standard deviation, which reflects the degree of fluctuation of the traceability status indicator. For example, a standard deviation of 5 minutes indicates that the transportation time deviation fluctuates more frequently between 10 and 20 minutes.

[0053] Calculate the statistical characteristics of the traceability status indicator distribution for each construction status within the status group, and then combine these statistical characteristics to obtain the group traceability threshold for the status group. The determination of the group traceability threshold needs to consider the statistical characteristics of all construction statuses within the status group. For example, the group traceability threshold can be set to the average value plus a certain multiple of the standard deviation. Assuming that the average transportation time deviation of all construction statuses in the transportation status group is 20 minutes, the standard deviation is 8 minutes, and the multiple is set to 2, the group traceability threshold can be set to 20+2×8=36 minutes. That is, when the transportation time deviation exceeds 36 minutes, it is considered that the traceability status of the construction status may be abnormal.

[0054] For each construction state, the degree of match between its component tracking feature vector and the typical vector of the state group to which it belongs is calculated. The typical vector can be determined by calculating the average value of the component tracking feature vectors of all construction states in the state group, which represents the typical feature set of the state group. The component tracking feature vector represents the feature set of the construction state. By calculating the degree of match between these two vectors, the degree of similarity between the construction state and the typical features of the state group to which it belongs can be reflected. The degree of match can be calculated using the cosine similarity method. For example, the cosine similarity between the component tracking feature vector of a construction state and the typical vector of the state group to which it belongs is 0.85, indicating that the construction state has a high degree of match with the typical features of the group.

[0055] The dynamic traceability threshold for a construction state is calculated based on the degree of match between the component tracking feature vector of the construction state and the representative vector of the state group to which it belongs, as well as the group traceability threshold. For example, a high degree of match indicates that the construction state is similar to the representative characteristics of the group, and a stricter dynamic traceability threshold can be used. A lower degree of match results in a more relaxed dynamic traceability threshold. Specifically, the dynamic traceability threshold can be calculated as: dynamic traceability threshold = group traceability threshold × (1-k × (1-matching degree)), where k is an adjustment factor that can be set based on actual conditions. Assuming the group traceability threshold is 36 minutes, k is 0.5, and the matching degree for a construction state is 0.85, then the dynamic traceability threshold = 36 × (1-0.5 × (1-0.85)) = 36 × (1-0.075) = 36 × 0.925 = 33.3 minutes. In this way, the dynamic traceability threshold for the construction state is adjusted based on its degree of match with the representative characteristics of the group, achieving dynamic threshold setting.

[0056] In actual construction, as construction progresses, factors such as component wear and tear and adjustments to construction techniques can cause significant shifts in the representative vectors of a state group. Therefore, the dynamic traceability threshold for each construction state needs to be regularly updated. A fixed update cycle is set, such as weekly. During each update cycle, data is collected on the deviations between the actual traceability status indicators and the dynamic traceability threshold for all construction states within that cycle. For example, within a week, the actual transportation time deviation for a certain construction state is 35 minutes, while its dynamic traceability threshold is 33.3 minutes, resulting in a deviation of 1.7 minutes. The number and magnitude of these deviations are counted. For example, within a week, the actual traceability status indicator for this construction state exceeded the dynamic traceability threshold twice, by 1.7 minutes and 2.5 minutes, respectively. Based on the changing trends of the component tracking feature vectors within the current state group, if the component tracking feature vectors for multiple construction states show an overall delay in the transportation phase, this may indicate transportation route adjustments or traffic congestion, necessitating adjustment of the dynamic traceability threshold. If the representative vector of a state group shifts significantly due to component wear or construction process adjustments—for example, if component hoisting times are generally shortened after a construction process adjustment, resulting in a significant change in the representative vector of the hoisting state group—then the dynamic traceability threshold for all construction states within that group needs to be recalculated to ensure that the dynamic traceability threshold accurately reflects the current construction status. This regular update allows the dynamic traceability threshold to be dynamically adjusted as the construction process changes, improving the accuracy and reliability of traceability predictions.

[0057] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0058] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A steel structure construction component tracking and tracing method based on the Internet of Things, characterized in that: include: Collect IoT perception information and construction process information of steel structure construction components, associate the perception information with the process information, and build a component tracking feature set; Based on the component tracking feature set, tracking nodes are defined. According to the construction process information, associated edges and edge weights are defined to build a multi-dimensional tracking network. Based on the temporal tracking network model, the evolution law of tracking nodes in the multi-dimensional tracking network is learned to generate component tracking feature vectors. Train a traceability prediction model for predicting the real-time traceability status of components. This model uses component tracking feature vectors and real-time construction data as input data, and component traceability status indicators as output data to predict the traceability status indicators of steel structure construction components at the current construction stage. Based on the component tracking feature vector, all construction states are divided into several state groups. The group traceability threshold is calculated for each state group. Based on the group traceability threshold and the matching degree between the component tracking feature vector of each construction state and the typical vector of the state group to which it belongs, the dynamic traceability threshold of the construction state is calculated. The traceability status indicator output by the traceability prediction model is compared with the dynamic traceability threshold of the construction status. If the traceability status indicator exceeds the dynamic traceability threshold, a corresponding tracking adjustment instruction is generated, and the Internet of Things executes the terminal action, and the dynamic traceability threshold of each construction status is regularly updated.

2. The method for tracing and tracing steel structure construction components based on the Internet of Things according to claim 1, characterized in that: The specific method of collecting IoT perception information and construction process information of steel structure construction components, and performing feature association between the perception information and the process information to construct a component tracking feature set is as follows: Acquire sensory information and construction process information from IoT sensor terminals of steel structure construction components. The sensory information includes: component unique identification code, installation coordinate positioning, and material detection data; the construction process information includes: transportation timestamp, hoisting sequence record, and welding quality log; The characteristic attributes of the perception information and the construction process information are extracted and associated to obtain the component tracking feature set of each construction status.

3. The method for tracing and tracing steel structure construction components based on the Internet of Things according to claim 2, characterized in that: The specific method of defining tracking nodes based on the component tracking feature set, defining associated edges and edge weights according to the construction process information, and constructing a multi-dimensional tracking network is as follows: Taking the perception information as the basic attribute of the tracking node, the tracking node is constructed based on the component tracking feature set and the corresponding basic attributes of each tracking node; The correlation edges between tracking nodes are constructed based on the construction process information, and the temporal correlation between tracking nodes is calculated based on the temporal record data in the construction process information. The temporal record data of tracking node p is the information sequence within the continuous construction period, and the temporal record data of tracking node q is the information sequence within the corresponding period. The time series alignment algorithm is used to assign adjustment weights to the correlation of temporal data in different periods based on the time decay parameter, and the temporal correlation between tracking node p and tracking node q is calculated. Converting the temporal correlation into the tracking correlation; Based on the historical collaborative construction records of the components, the adjacent node set of tracking nodes p and q in the collaborative construction network is obtained. Based on the adjacent node set, the collaborative matching index between tracking nodes is calculated. Different weights are assigned to nodes with different collaborative distances to obtain the weighted collaborative matching index. The weighted collaborative matching index is normalized to the interval [0,1] to obtain the collaborative correlation degree between tracking nodes p and q. The edge weights of the associated edges between tracking nodes p and q are calculated based on the tracking association degree and collaborative association degree, and the tracking nodes, associated edges and their edge weights are used to form a multi-dimensional tracking network.

4. The method for tracing and tracing steel structure construction components based on the Internet of Things according to claim 3, characterized in that: The training method of the time series tracking network model is: Step B1: Construct a time series tracking network model, including a node encoding layer, a time series memory layer, an attention module, an output layer, and a parameter optimization module. The model takes the multi-dimensional tracking network at each time point as input and the evolutionary characteristics of the tracking nodes at each time point as output. The node encoding layer is used to map the features of each tracking node and associated edges into a unified feature space. Based on the node encoding layer, a temporal memory layer is introduced to capture the dynamic evolution pattern of tracking features. It takes the multi-dimensional tracking network sequence at each time point as input, outputs the evolution features of the tracking node at each time point, and describes the trajectory of the tracking node evolution features changing over time. Use the attention module to adaptively assign the influence weights of different adjacent nodes and highlight the role of key adjacent nodes; The output layer outputs the evolution characteristics of each tracking node at each time point; The parameter optimization module is used to calculate the gradient of the loss function with respect to the model parameters using the back-propagation algorithm; Step B2: For each tracking node p, construct a positive sample set and a negative sample set based on the tracking correlation and collaborative correlation, set the correlation threshold and the matching threshold, and take the tracking nodes whose tracking correlation with the tracking node p is greater than the correlation threshold or whose collaborative correlation is greater than the matching threshold as positive samples. The tracking nodes that do not belong to the positive sample set are included in the negative sample set. The positive and negative samples are used as training data. An unsupervised training method is adopted, with the nodes in the multidimensional tracking network as input and the evolution characteristics of the nodes as output. The time series tracking network model is applied to the multidimensional tracking network, and the evolution characteristics of all tracking nodes at each time point t are generated by forward propagation. Step B3: Using the evolutionary features of the tracking node p at time point t, calculate the feature similarity between the tracking node p and the tracking nodes in the positive sample set and the negative sample set. Based on the feature similarity, calculate the contrast loss function at the current time point t. Average the contrast loss functions of all tracking nodes and all time points to obtain the final loss function. Minimizing the loss function is the training goal. Step B4: Calculate the gradient of the loss function with respect to the model parameters through the back-propagation algorithm and use the optimizer to update the model parameters; Step B5: Repeat steps B2 to B4 until the loss function converges.

5. The method for tracing and tracing steel structure construction components based on the Internet of Things according to claim 4, characterized in that: The specific method of learning the evolution law of tracking nodes in the multi-dimensional tracking network based on the time-series tracking network model and generating the component tracking feature vector is as follows: On a multi-dimensional tracking network, a temporal tracking network model is used to learn the evolutionary characteristics of tracking nodes. The evolutionary characteristics of tracking node p at the nth layer are obtained through forward propagation. An attention module is introduced to calculate the attention weights of adjacent tracking nodes with different collaborative relationships for tracking node p. In the temporal memory layer, a gated memory unit aggregation function is defined to fuse the evolutionary features of the adjacent tracking node q at the current time point and the evolutionary features of the tracking node p at the previous time point to obtain the evolutionary features of the tracking node p at the current time point t. The evolution characteristics of the tracking node p at time point t are spliced ​​with the perception information of the tracking node to obtain the component tracking feature vector of the tracking node p.

6. The method for tracing and tracing steel structure construction components based on the Internet of Things according to claim 5, characterized in that: The method for training the traceability prediction model for predicting the real-time traceability status of components is as follows: the component tracking feature vector and real-time construction data are used as input data, and the component traceability status index is used as output data. The specific method for predicting the traceability status index of the steel structure construction component at the current construction stage is as follows: Obtain a batch of historical construction records of steel structure components, extract their component tracking feature vectors, mark the status indicators of abnormal traceability as abnormal traceability labels, construct training samples based on the component tracking feature vectors and abnormal traceability labels, use the component tracking feature vectors and real-time construction data as input data, and the abnormal traceability labels as output data to train the traceability prediction model; For each training sample, the prediction target is to accurately predict the anomaly traceability label. The cross entropy loss function is used as the loss function for the training model. The training target is to minimize the value of the loss function. The training is completed when the loss function converges. Using the trained traceability prediction model, for the current real-time construction status, the component tracking feature vector and real-time construction data of the corresponding tracking node are extracted as input data, and the traceability status indicator of the real-time construction status is output.

7. The method for tracing and tracing steel structure construction components based on the Internet of Things according to claim 6, characterized in that: The method for dividing all construction states into several state groups based on the component tracking feature vector and calculating the group traceability threshold for each state group is as follows: Based on the component tracking feature vector, all construction states are divided into several state groups through density clustering algorithm; For each state group, the component tracking feature vector and historical construction records of the construction state are extracted. The component tracking feature vector and each historical construction data are used as input data to input the traceability prediction model to obtain the traceability state index of each historical construction. Collect statistics on the traceability status indicators of each historical construction of the construction state to obtain the traceability status indicator distribution of the construction state; calculate the statistical characteristics of the traceability status indicator distribution of the construction state, including the average value and dispersion; Calculate the statistical characteristics of the traceability status indicator distribution of each construction status in the status group to obtain the group traceability threshold of the status group; For each construction state, the matching degree between its component tracking feature vector and the typical vector of the state group to which it belongs is calculated, where the component tracking feature vector represents the feature set of the construction state, and the typical vector represents the typical feature set of the state group; The dynamic traceability threshold of the construction state is calculated based on the matching degree between the component tracking feature vector of the construction state and the typical vector of the state group to which it belongs and the group traceability threshold.

8. The method for tracing and tracing steel structure construction components based on the Internet of Things according to claim 1, wherein: The specific method for regularly updating the dynamic traceability threshold of each construction status is: Set a fixed update cycle. During each update cycle, collect the deviation data between the actual traceability status indicators and the dynamic traceability thresholds for all construction statuses within that cycle. The number and magnitude of the threshold values ​​exceeded in the statistical deviation data are combined with the changing trend of the component tracking feature vector in the current state group to adjust the dynamic traceability threshold of the construction state; If the typical vector of a state group deviates significantly due to component loss or construction process adjustment, the dynamic traceability threshold of all construction states in the group is recalculated.

9. The method for tracing and tracing steel structure construction components based on the Internet of Things according to claim 1, wherein: The specific method of generating the corresponding tracking adjustment instruction is: Based on the type and degree of the traceability status indicator exceeding the dynamic traceability threshold, the preset adjustment strategy table is called; If the excess type is missing identification, an adjustment instruction for component identification re-recording is generated; if it is positioning deviation, an adjustment instruction for installation coordinate correction is generated; if it is material abnormality, an adjustment instruction for component batch replacement is generated.

10. The method for tracing and tracing steel structure construction components based on the Internet of Things according to claim 9, characterized in that: The tracking adjustment instruction includes a specific parameter adjustment value and an execution time node.

Images