A real-time monitoring method for silo engineering driven by the Internet of Things

Abstract

The present invention discloses a real-time monitoring method for silo projects driven by the Internet of Things, which relates to the technical field of silo projects. The method comprises: deploying a local Internet of Things sensing network in the target project space, connecting to a monitoring management center, performing local monitoring and collection, and obtaining a local monitoring data set. Obtaining a dynamic Internet of Things sensing device, connecting to a monitoring management center, performing extended monitoring, and obtaining an extended monitoring data set. Interacting with silo project data sources, constructing a silo project knowledge graph, and initializing and supervising the training of an engineering monitoring discrimination model. Using the local monitoring data set and the extended monitoring data set as input, activating the trained engineering monitoring discrimination model for monitoring and analysis, obtaining engineering monitoring results, including risk discrimination information and engineering progress information, and performing management responses. Thereby achieving the technical effect of improving the level of information linkage and data timeliness, and improving monitoring efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of silo engineering, and in particular to a real-time monitoring method for silo engineering driven by the Internet of Things. Background Art

[0002] Silos are important facilities for storing grain, chemical raw materials and other materials. The safety management and quality management in their construction directly affect the stability of subsequent production and supply. Existing silo engineering monitoring mainly relies on manual inspections and simple instrument measurements, which have technical problems such as low monitoring efficiency, untimely data updates, and lack of information linkage. Summary of the invention

[0003] The present invention provides a real-time monitoring method for silo projects driven by the Internet of Things to solve the technical problems of low monitoring efficiency, untimely data update, and lack of information linkage in the prior art, thereby achieving the technical effect of improving the level of information linkage and data timeliness and enhancing monitoring efficiency.

[0004] The present invention provides an Internet of Things driven silo engineering real-time monitoring method comprising:

[0005] A local Internet of Things perception network is deployed in the target engineering space, and a connection is established between the local Internet of Things perception network and a monitoring management center to conduct local monitoring collection and obtain a local monitoring data set, wherein the monitoring management center is configured with a dynamic interface.

[0006] Acquire a dynamic Internet of Things sensing device, establish a connection between the dynamic Internet of Things sensing device and the monitoring management center, perform extended monitoring, and acquire an extended monitoring data set, wherein the extended monitoring data set includes in-transit monitoring data and quality monitoring data.

[0007] Interact with silo engineering data sources, construct a silo engineering knowledge graph, initialize an engineering monitoring discrimination model based on the silo engineering knowledge graph, and perform supervised training of the engineering monitoring discrimination model.

[0008] Taking the on-site monitoring data set and the expanded monitoring data set as input data, the trained engineering monitoring discrimination model is activated to perform monitoring analysis, obtain engineering monitoring results, and perform management response based on the engineering monitoring results, wherein the engineering monitoring results include risk discrimination information and engineering progress information.

[0009] In a feasible implementation method, a local Internet of Things perception network is deployed in the target engineering space, and a connection is established between the local Internet of Things perception network and a monitoring management center. The local Internet of Things perception network includes multiple sub-networks, and at least includes an engineering subject perception sub-network, a personnel perception sub-network, a material perception sub-network, and an environmental perception sub-network.

[0010] In a feasible implementation, a dynamic Internet of Things sensing device is obtained, a connection between the dynamic Internet of Things sensing device and the monitoring management center is established, and extended monitoring is performed to obtain an extended monitoring data set, including:

[0011] Interact with the engineering management database to obtain engineering task information. Based on the engineering task information, determine the dynamic monitoring object and the IoT sensing device of the dynamic monitoring object, and output it as the dynamic IoT sensing device. Collect the device intrinsic information of the dynamic IoT sensing device, and the device intrinsic information at least includes ID information, logical address, encoding protocol, and sensing performance. According to the device intrinsic information, establish a connection between the dynamic IoT sensing device and the monitoring management center, and the monitoring management center calls the preset extended collection list to perform extended monitoring, wherein the extended collection list includes dynamic monitoring objects and associated stored collection indicators.

[0012] In a feasible implementation, the silo engineering data source is interactive, a silo engineering knowledge graph is constructed, and an engineering monitoring discrimination model is initialized based on the silo engineering knowledge graph, including:

[0013] Interact with silo engineering data sources to obtain silo engineering documents. Parse the silo engineering documents, define engineering entities and engineering relationships, and construct a silo engineering knowledge graph with the engineering entities as graph nodes and engineering relationships as relationship edges. Based on graph embedding technology, embed the silo engineering knowledge graph into a multidimensional vector space to obtain an embedded representation of the silo engineering knowledge graph. Based on a neural network algorithm, construct the engineering monitoring discrimination model, and initialize the parameters of the silo engineering monitoring discrimination model according to the embedded representation.

[0014] In a feasible implementation, before performing the supervised training of the silo engineering monitoring and discrimination model, the method further includes:

[0015] Interact with the engineering management database to obtain the original source monitoring records. Based on the principle of big data, obtain the same source monitoring records, perform commonality analysis on the same source monitoring records, extract the monitoring records of common engineering links, and store them as common monitoring records. Fusion the original source monitoring records and the common monitoring records, output as a monitoring discrimination sample set, and divide the monitoring discrimination sample set into a monitoring discrimination training set and a monitoring discrimination verification set.

[0016] In a feasible implementation, the method further includes:

[0017] Configure monitoring update constraints, wherein the monitoring update constraints include periodic update constraints and trigger update constraints. Based on the monitoring update constraints, perform update discrimination. If the update interval is greater than or equal to the periodic update constraint or the cumulative number of misjudgments of the engineering monitoring results is greater than or equal to the trigger update constraint, update the silo engineering knowledge graph and perform enhanced training on the silo engineering monitoring discrimination model.

[0018] In a feasible implementation, the dynamic interface is configured with a connection protocol for adaptive expansion and contraction, including:

[0019] When the dynamic interface is idle, the connection of the dynamic interface is adaptively and actively downgraded, wherein the active downgrade includes horizontal downgrade and vertical downgrade. When the dynamic interface is busy, the connection of the dynamic interface is adaptively and actively expanded, wherein the active expansion includes horizontal expansion and vertical expansion.

[0020] In a feasible implementation, performing management response based on the engineering monitoring result further includes:

[0021] The risk identification information is parsed to obtain the risk category and risk level. The risk category and risk level are used as index constraints to traverse the preset risk response matrix to call the response plan and perform risk management response. The project progress information is compared with the expected progress to perform project progress management.

[0022] The present invention discloses a real-time monitoring method for silo engineering driven by the Internet of Things, comprising: deploying a ground-based Internet of Things sensing network in the target engineering space, and ensuring that the sensing network is connected to the monitoring management center so as to perform on-site monitoring and collection, thereby obtaining a local monitoring data set. The monitoring management center is configured with a dynamic interface. Dynamic Internet of Things sensing devices are obtained, and connections between these devices and the monitoring management center are established for extended monitoring. The extended monitoring data set includes in-transit monitoring data and quality monitoring data. Interacting with the silo engineering data source to construct a silo engineering knowledge graph. Based on the silo engineering knowledge graph, the engineering monitoring discrimination model is initialized and supervised training is performed. Using the on-site monitoring data set and the extended monitoring data set as input data, the trained engineering monitoring discrimination model is activated for monitoring and analysis. The analysis result is the engineering monitoring result, including risk discrimination information and engineering progress information, so as to perform management response according to the engineering monitoring result. The real-time monitoring method for silo engineering driven by the Internet of Things disclosed in the present invention solves the technical problems of low monitoring efficiency, untimely data update, and lack of information linkage, and achieves the technical effect of improving the level of information linkage and data timeliness, and improving monitoring efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1It is a flow chart of a real-time monitoring method of a silo project driven by the Internet of Things of the present invention;

[0024] Figure 2 The present invention is a flowchart of initializing an engineering monitoring discrimination model in a real-time monitoring method for silo engineering driven by the Internet of Things. DETAILED DESCRIPTION

[0025] The technical solution provided in the embodiments of the present invention is to solve the technical problems of low monitoring efficiency, untimely data update and lack of information linkage in the prior art. The overall idea adopted is as follows:

[0026] First, the IoT sensing network is deployed in the target engineering space, and a connection is established with the monitoring management center to conduct local monitoring and acquisition, and obtain the local monitoring data set. The monitoring management center is equipped with a dynamic interface that can receive and process real-time data. Then, the dynamic IoT sensing device is obtained and its connection with the monitoring management center is established to expand the monitoring and obtain the expanded monitoring data set. The expanded monitoring data set includes in-transit monitoring data and quality monitoring data, which will enrich and improve the monitoring information. Next, the silo engineering data source is interacted to build a silo engineering knowledge graph. The engineering monitoring discrimination model is initialized using the silo engineering knowledge graph, and supervised training is performed. Finally, the local monitoring data set and the expanded monitoring data set are used as input data to activate the trained engineering monitoring discrimination model for monitoring analysis. By analyzing the engineering monitoring results, risk discrimination information and engineering progress information are obtained. Management responses are made based on these monitoring results to help optimize and manage the engineering process and ensure the safety and efficiency of the project.

[0027] The above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods of the specification to better understand the above technical solution. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments used only to explain the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. In addition, it should be noted that, for the convenience of description, only the parts related to the present invention are shown in the drawings, rather than all of them.

[0028] Example

[0029] Figure 1 The present invention is a flowchart of a real-time monitoring method for silo engineering driven by the Internet of Things, wherein the method comprises:

[0030] A local Internet of Things perception network is deployed in the target engineering space, and a connection is established between the local Internet of Things perception network and a monitoring management center to conduct local monitoring collection and obtain a local monitoring data set, wherein the monitoring management center is configured with a dynamic interface.

[0031] Specifically, the target engineering space refers to the silo engineering construction area where the engineering management is carried out. According to the engineering requirements, temperature, humidity, pressure, vibration, distance and other sensors are deployed in the construction area to form a complete perception network. Among them, the deployment of the local IoT perception network includes the transformation or replacement of existing non-IoT devices with IoT devices, as well as the addition of necessary IoT sensors or collection equipment.

[0032] Specifically, wired or wireless network technology is used, and appropriate data transmission protocols (such as MQTT, HTTP, etc.) are selected to connect the local IoT sensing network with the monitoring management center to ensure real-time data transmission and realize centralized data management and analysis. The data management center has the functions of receiving, storing, analyzing and displaying data.

[0033] Optionally, the monitoring management center is configured with multiple dynamic interfaces, which can integrate and interact with different types of devices and systems, thereby ensuring the scalability and flexibility of monitoring management. Specifically, the dynamic interface adopts a standardized interface protocol (such as RESTful API, WebSocket, etc.) to ensure the compatibility and interoperability of monitoring management. Through the dynamic interface, it is convenient to add new sensing devices or disconnect abandoned sensing devices in time, thereby improving the flexibility and scalability of monitoring management.

[0034] Furthermore, through real-time monitoring and data collection on the local IoT sensing network, a local monitoring data set is generated. First, the deployed IoT sensing devices (sensors) are activated to collect data such as temperature, humidity, pressure, vibration, distance, and positioning at the construction site in real time. Then, the collected data is transmitted to the monitoring management center for centralized storage and management. Next, the monitoring management center pre-processes, cleans, and integrates the collected raw data to form a complete local monitoring data set. Among them, the local monitoring data set includes the timestamp and collection source of the associated mark.

[0035] In some embodiments, a local Internet of Things perception network is deployed in the target engineering space, and a connection is established between the local Internet of Things perception network and a monitoring management center. The local Internet of Things perception network includes multiple sub-networks, and at least includes an engineering subject perception sub-network, a personnel perception sub-network, a material perception sub-network, and an environment perception sub-network.

[0036] Specifically, the local IoT sensing network includes multiple sub-networks, including at least: the main project sensing sub-network, which is used to monitor the key parameters of silo construction, such as structural stress, deformation, temperature, positioning size, etc. The IoT sensing devices in the main project sensing sub-network include stress sensors, deformation sensors, temperature sensors, etc. The personnel sensing sub-network is used to monitor the safety status and location of construction personnel. The IoT sensing devices of the main project sensing sub-network include smart wearable devices, positioning tags, vital signs monitoring equipment, etc. The material sensing sub-network is used to monitor the quantity, status and location of construction materials. The IoT sensing devices involved include RFID tags, smart storage equipment, transportation monitoring equipment, etc. The environmental sensing sub-network is used to monitor the key parameters of the construction environment, such as meteorological conditions, air quality, noise, etc. The IoT sensing devices of the environmental sensing sub-network include weather stations, air quality sensors, noise sensors, etc.

[0037] Through the above method and steps, the local IoT sensing network is deployed based on IoT sensing devices, which improves the efficiency of monitoring and collection, realizes real-time monitoring of key parameters of the construction site, helps to discover potential problems in time, and improves the efficiency of construction management. At the same time, the data is centrally stored and managed through the monitoring management center, which improves the utilization and reliability of the data.

[0038] Acquire a dynamic Internet of Things sensing device, establish a connection between the dynamic Internet of Things sensing device and the monitoring management center, perform extended monitoring, and acquire an extended monitoring data set, wherein the extended monitoring data set includes in-transit monitoring data and quality monitoring data.

[0039] Specifically, a dynamic IoT sensing device refers to an IoT sensing device that is located outside the target engineering space but is configured on some mobile objects that have an associated relationship with the target engineering space.

[0040] Optionally, dynamic IoT sensing devices include IoT sensing devices on material transport vehicles, transport ships, engineering machinery or equipment, etc. Exemplarily, IoT devices configured on material transport vehicles can provide on-the-way monitoring data, such as the vehicle's current location, speed, and estimated arrival time. Devices configured on engineering machinery or equipment may provide equipment operating status and performance data, such as equipment temperature, work efficiency, etc., which are very useful for equipment maintenance and fault warning. In addition, the expanded monitoring data set also includes real-time material status data. For example, the concrete temperature and humidity information of a ready-mixed concrete truck can be used to monitor the quality of the concrete.

[0041] Specifically, the data collected by dynamic IoT sensing devices can form an expanded monitoring data set, which can be sent to the monitoring management center for analysis and processing to optimize engineering operations and improve efficiency and quality.

[0042] In some embodiments, obtaining a dynamic Internet of Things sensing device, establishing a connection between the dynamic Internet of Things sensing device and the monitoring management center, performing extended monitoring, and obtaining an extended monitoring data set include:

[0043] Interactive engineering management database to obtain engineering task information.

[0044] Based on the engineering task information, a dynamic monitoring object is determined and an Internet of Things sensing device of the dynamic monitoring object is determined, and the output is the dynamic Internet of Things sensing device.

[0045] The device intrinsic information of the dynamic Internet of Things sensing device is collected, and the device intrinsic information at least includes ID information, logical address, encoding protocol, and sensing performance.

[0046] According to the intrinsic information of the device, a connection is established between the dynamic Internet of Things sensing device and the monitoring management center, and the monitoring management center calls a preset extended collection list to perform extended monitoring, wherein the extended collection list includes dynamic monitoring objects and associated stored collection indicators.

[0047] Specifically, first, the database connection string and authentication information (such as user name and password) are used to establish a connection, and the data in the engineering management database is accessed through the database query language (such as SQL) and API interface, and the engineering task information is extracted from the engineering management database, including the bill of quantities, transportation plan of materials or machinery, etc. Specifically, the information items of the engineering task information include task ID, task name, task description, task start date, task end date, task route, etc. By extracting information related to engineering tasks from the engineering management database, including the bill of quantities, transportation plan of materials or machinery, etc., it is possible to fully understand the various tasks in the project and ensure the accuracy of the monitoring objects and monitoring scope in the subsequent expansion of monitoring.

[0048] Furthermore, based on the engineering task information extracted from the engineering management database, key objects that need to be monitored, such as transport vehicles, engineering machinery, construction equipment, etc., are identified as dynamic monitoring objects, and at the same time, the IoT sensing devices on the dynamic monitoring objects are determined. For example, if the engineering task involves the transportation of concrete, then the ready-mixed concrete truck is a dynamic monitoring object, and the temperature and humidity sensors on the truck are IoT sensing devices.

[0049] Specifically, basic information of IoT sensing devices is collected, such as device ID (such as MAC address or serial number), logical address, encoding protocol (such as MQTT, CoAP, HTTP, etc.) and sensing performance (including sensor type, accuracy, range, sampling frequency, transmission rate, etc.) to establish a connection between the device and the monitoring management center, and to determine the function and performance of the device.

[0050] Exemplarily, use the device management interface or communication protocol command to read the device ID, obtain the device's logical address through a network management tool or a device configuration interface, consult the device's user manual or technical specification, query the device's supported encoding protocols through the device management interface (such as a REST API), and obtain the perception performance parameters of the IoT perception device through the device specification, device management interface (such as a REST API), or a self-describing data packet.

[0051] Optionally, a connection is established with the monitoring management center based on the intrinsic information of the device, including configuring the correct IP address and encoding protocol, and performing configuration verification based on the device ID information. Then, the monitoring management center calls a preset extended collection list for extended monitoring, and the extended collection list includes the data types to be monitored, the collection frequency, etc. For example, for a ready-mixed concrete truck, the location and speed of the vehicle, the temperature and moisture content of the concrete, etc. need to be monitored.

[0052] Through the above process, it is ensured that the monitoring management center can effectively collect and process the required data to support the management and optimization of engineering tasks.

[0053] Interact with silo engineering data sources, construct a silo engineering knowledge graph, initialize an engineering monitoring discrimination model based on the silo engineering knowledge graph, and perform supervised training of the engineering monitoring discrimination model.

[0054] In some embodiments, Figure 2 As shown, the interactive silo engineering data source is used to construct a silo engineering knowledge graph, and the engineering monitoring discrimination model is initialized based on the silo engineering knowledge graph, including:

[0055] Interactive silo engineering data source to obtain silo engineering literature.

[0056] The silo engineering literature is parsed, engineering entities and engineering relationships are defined, and a silo engineering knowledge graph is constructed with the engineering entities as graph nodes and engineering relationships as relationship edges.

[0057] Based on graph embedding technology, the silo engineering knowledge graph is embedded into a multi-dimensional vector space to obtain an embedded representation of the silo engineering knowledge graph.

[0058] Based on the neural network algorithm, the engineering monitoring discrimination model is constructed, and the parameters of the silo engineering monitoring discrimination model are initialized according to the embedded representation.

[0059] Specifically, the silo engineering data source refers to the data platform related to silo engineering, including relevant forums, journals, databases, document libraries, etc., through interactive silo engineering data sources to obtain relevant engineering documents. Specifically, silo engineering documents include relevant information on silo design, construction, operation and maintenance.

[0060] Specifically, the acquired silo engineering literature is parsed through natural language processing (NLP) technology, named entity recognition (NER) technology, text mining algorithms, relationship extraction algorithms, etc., to define the engineering entities in the literature (such as equipment, materials, processes) and the relationships between entities (such as use, connection, dependency, etc.). Then, based on the determined entities and relationships, a graph database (such as Neo4j) and a graph processing framework (such as NetworkX) are used to build and store the silo engineering knowledge graph. In the silo engineering knowledge graph, engineering entities are nodes of the graph, and the relationships between entities are edges of the graph.

[0061] Furthermore, graph embedding technology is used to embed nodes and edges in the knowledge graph into a multidimensional vector space using graph embedding algorithms (such as Node2Vec, DeepWalk, and GraphSAGE) to generate an embedded representation of the knowledge graph. Specifically, first, an initial random vector is generated for each node (entity) in the graph. Then, these vectors are updated using the structural information of the graph (such as the neighbors of the node) and the characteristic information of the node. This process is repeated many times until the vector stabilizes, forming a vector representation for each node that can capture the node's position, role, and contextual information in the graph.

[0062] Optionally, the embedded representation is used as input to construct an engineering monitoring discrimination model. The engineering monitoring discrimination model is constructed based on a neural network, including a deep feedforward network, a long short-term memory network, or a recurrent neural network (RNN). Specifically, first, the structure of the neural network is defined, including the number of layers, the number of nodes in each layer, the activation function, etc. Then, the embedded representation of the knowledge graph is used as input to initialize the parameters of the neural network and generate a silo engineering monitoring discrimination model. The initialization process helps the neural network to have a preliminary understanding of the silo engineering before starting training, thereby enabling the silo engineering monitoring discrimination model to have preliminary discrimination analysis capabilities, thereby shortening the convergence time of subsequent training.

[0063] In some implementations, before performing the supervised training of the silo engineering monitoring discrimination model, the method further includes:

[0064] Interact with the engineering management database to obtain source monitoring records.

[0065] Based on the principle of big data, homologous monitoring records are obtained, and commonality analysis is performed on the homologous monitoring records to extract monitoring records of common engineering links and store them as common monitoring records.

[0066] The original monitoring records and the common monitoring records are integrated to output a monitoring discrimination sample set, and the monitoring discrimination sample set is divided into a monitoring discrimination training set and a monitoring discrimination verification set.

[0067] Specifically, we first obtain the original monitoring records from the engineering management database. The original monitoring records are engineering monitoring records collected directly from the target silo engineering site. Then, based on the principle of big data, we use data mining techniques such as cluster analysis and frequent pattern mining to obtain homologous monitoring records from multiple similar engineering projects, and perform commonality analysis on these records to extract monitoring records of common engineering links. Common monitoring records reflect the similarities between different silo projects, which helps to improve the generalization ability of the model.

[0068] Specifically, the original monitoring records and common monitoring records are fused to generate a monitoring discrimination sample set. This sample set contains a large amount of monitoring data, which is used to train and verify the silo engineering monitoring discrimination model. The fusion process includes cleaning and preprocessing of the monitoring records to ensure data quality and consistency.

[0069] Furthermore, the generated monitoring discrimination sample set is divided into a training set and a validation set for model training and validation, wherein data partitioning methods include random partitioning, cross-validation and other methods. Exemplarily, the data is usually divided in a ratio of 8:2 or 7:3 to ensure the data volume and representativeness of the training set and the validation set.

[0070] Through the above steps, original source monitoring records and homologous source monitoring records are obtained from multi-source data, ensuring the comprehensiveness and diversity of the training data. By conducting commonality analysis on homologous source monitoring records and extracting monitoring data of common engineering links, the representativeness and versatility of the data are improved.

[0071] Optionally, supervised training of the silo engineering monitoring and discrimination model is performed based on the monitoring and discrimination training set and the monitoring and discrimination verification set, including calculating the gradient of the loss function through a back-propagation algorithm, and adjusting the model parameters through a gradient descent method, so that the prediction results of the model on the training set are as close as possible to the actual monitoring results, and then, the model performance is verified on the monitoring and discrimination verification set, exemplarily including evaluating the generalization performance of the model using a k-fold cross-validation method, and using evaluation indicators such as accuracy, precision, recall, and F1 value to evaluate the model performance.

[0072] In some embodiments, performing management response based on the engineering monitoring result further includes:

[0073] The risk identification information is analyzed to obtain the risk category and risk level.

[0074] Taking the risk category and the risk level as index constraints, the preset risk response matrix is ​​traversed to call the response plan and perform risk management response.

[0075] Compare the project progress information with the expected progress and perform project progress management.

[0076] Optionally, first, parse the risk discrimination information output by the monitoring discrimination model to extract the risk category and risk level. Then, according to the extracted risk category and risk level, use them as indexes to traverse the preset risk response matrix, match and select appropriate response plans, and execute risk management responses. Among them, the pre-constructed risk response matrix contains response plans corresponding to different risk categories and risk levels. For example, for high-risk situations, the response plans include stopping the project, conducting safety inspections, etc.; for low-risk situations, the response plans may include ordering adjustments, increasing the monitoring frequency, and raising the alert level.

[0077] Optionally, use project management software or a project progress tracking system to obtain and record project progress information in real time, and then compare the real-time project progress information with the expected progress, identify progress deviations, and take corresponding measures to manage the project progress. Examples include resource deployment (such as adding staff), project plan revision, and construction period adjustment.

[0078] Through the above method steps, the response plan in the preset risk response matrix is ​​automatically called based on the risk category and risk level, which helps to improve the efficiency and response speed of risk management. By comparing the project progress information with the expected progress in real time, the progress deviation can be identified and adjusted in time to ensure that the project proceeds as planned.

[0079] Furthermore, the method further comprises:

[0080] Configure monitoring update constraints, wherein the monitoring update constraints include periodic update constraints and triggered update constraints.

[0081] Based on the monitoring update constraint, update judgment is performed. If the update interval is greater than or equal to the periodic update constraint or the cumulative number of misjudgments of the engineering monitoring results is greater than or equal to the trigger update constraint, the silo engineering knowledge graph is updated and the silo engineering monitoring judgment model is enhanced and trained.

[0082] Specifically, define and configure monitoring update constraints, including periodic update constraints and trigger update constraints. The periodic update constraints specify the time interval for regular updates, such as monthly or quarterly updates. The trigger update constraints set the conditions for triggering updates, such as triggering updates when the cumulative number of misjudgments of engineering monitoring results reaches a certain threshold.

[0083] Furthermore, according to the configured monitoring update constraints, the monitoring data is analyzed and judged to determine whether it needs to be updated, including: using timestamp and timer functions to monitor the time interval to determine whether the time for periodic update has been reached; counting the number of false positives in the engineering monitoring results, and when the cumulative number of false positives exceeds the trigger update threshold, it is determined that an update is needed.

[0084] Specifically, when the discrimination results meet the update conditions, new silo engineering literature and monitoring data are collected to supplement and improve the engineering entities and relationships in the knowledge graph, and the silo engineering knowledge graph is updated to ensure the real-time and accuracy of the graph data. Then, based on the updated knowledge graph data, the neural network model is retrained using the monitoring discrimination training set and validation set, thereby further improving the prediction accuracy and robustness of the silo engineering monitoring discrimination model.

[0085] In some implementations, the dynamic interface is configured with an adaptive expansion and contraction connection protocol, including:

[0086] When the dynamic interface is idle, the connection of the dynamic interface is adaptively and actively downgraded, wherein the active downgrade includes horizontal downgrade and vertical downgrade.

[0087] When the dynamic interface is busy, the connection of the dynamic interface is actively expanded adaptively, wherein the active expansion includes horizontal expansion and vertical expansion.

[0088] Specifically, adaptive scaling policy parameters are defined through configuration files (such as JSON and YAML), including the trigger conditions and operations for downgrade and expansion, so as to adaptively scale the capacity according to the usage of the interface, save system resources, improve system efficiency, and thus improve the efficiency and stability of the monitoring management center.

[0089] Exemplarily, when the interface is idle, the monitoring management center adaptively and actively downgrades the connection of the interface. Active downgrade includes horizontal downgrade and vertical downgrade. Horizontal downgrade refers to reducing the number of concurrent connections or the size of the connection pool (such as downgrading a full-duplex connection to a half-duplex connection, downgrading a half-duplex connection to a simplex connection, suspending the connection or disconnecting the connection when the connection is completely idle and times out). Vertical downgrade refers to reducing the use of system resources (such as reducing CPU usage, memory usage, reducing the bandwidth of the connection or the quality of service (QoS), etc.).

[0090] Exemplarily, when the dynamic interface is in a busy state, the connection is adaptively and actively expanded to meet the high load demand. Active expansion also includes horizontal expansion and vertical expansion, where horizontal expansion may include upgrading from simplex to half-duplex, or from half-duplex to full-duplex; vertical expansion includes increasing the bandwidth or quality of service (QoS) of the connection, using existing connections for multiplexing, increasing CPU usage, memory usage, etc.

[0091] Through the above steps, when the dynamic interface is idle, the active downgrade strategy is used to save system resources of the monitoring management center, improve overall utilization, and reduce performance overhead. When the dynamic interface is busy, the active expansion strategy is used to meet high load requirements, ensuring the stability and high performance of the monitoring management center.

[0092] In summary, the real-time monitoring method of silo engineering driven by the Internet of Things provided by the present invention has the following technical effects:

[0093] By deploying a local IoT sensing network in the target engineering space and establishing a connection between the local IoT sensing network and the monitoring management center, local monitoring collection is carried out to obtain a local monitoring data set, wherein the monitoring management center is configured with a dynamic interface; dynamic IoT sensing devices are obtained, and a connection is established between the dynamic IoT sensing devices and the monitoring management center, and extended monitoring is carried out to obtain an extended monitoring data set, which includes in-transit monitoring data and quality monitoring data; the data source of the silo engineering is interacted with, a silo engineering knowledge graph is constructed, and an engineering monitoring discrimination model is initialized based on the silo engineering knowledge graph, and supervised training of the engineering monitoring discrimination model is carried out; the local monitoring data set and the extended monitoring data set are used as input data, and the trained engineering monitoring discrimination model is activated for monitoring analysis, and engineering monitoring results are obtained, and management responses are carried out based on the engineering monitoring results, and the engineering monitoring results include risk discrimination information and engineering progress information. Thus, the technical effect of improving the level of information linkage and data timeliness and improving monitoring efficiency is achieved.

[0094] It should be understood that the embodiments disclosed in the present invention and the above description can enable those skilled in the art to use the present invention to implement the present invention. At the same time, the present invention is not limited to the above-mentioned embodiments. It should be understood that those skilled in the art can still modify the technical solutions recorded in the above-mentioned embodiments, or replace some of the technical features therein by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. A real-time monitoring method for silo engineering driven by the Internet of Things, characterized in that: The method comprises: Deploy a local IoT sensing network in the target engineering space, and establish a connection between the local IoT sensing network and a monitoring management center to perform local monitoring collection and obtain a local monitoring data set, wherein the monitoring management center is configured with a dynamic interface; Acquire a dynamic Internet of Things sensing device, establish a connection between the dynamic Internet of Things sensing device and the monitoring management center, perform extended monitoring, and acquire an extended monitoring data set, wherein the extended monitoring data set includes in-transit monitoring data and quality monitoring data; Interact with silo engineering data sources, construct a silo engineering knowledge graph, initialize an engineering monitoring discrimination model based on the silo engineering knowledge graph, and perform supervised training of the engineering monitoring discrimination model; Taking the on-site monitoring data set and the expanded monitoring data set as input data, the trained engineering monitoring discrimination model is activated to perform monitoring analysis, obtain engineering monitoring results, and perform management response based on the engineering monitoring results, wherein the engineering monitoring results include risk discrimination information and engineering progress information.

2. The method for real-time monitoring of silo engineering driven by the Internet of Things according to claim 1, characterized in that: A local Internet of Things perception network is deployed in the target engineering space, and a connection is established between the local Internet of Things perception network and a monitoring management center. The local Internet of Things perception network includes multiple sub-networks, and at least includes an engineering subject perception sub-network, a personnel perception sub-network, a material perception sub-network, and an environment perception sub-network.

3. The method for real-time monitoring of silo engineering driven by the Internet of Things as claimed in claim 2, characterized in that: Acquire a dynamic Internet of Things sensing device, establish a connection between the dynamic Internet of Things sensing device and the monitoring management center, perform extended monitoring, and acquire an extended monitoring data set, including: Interactive engineering management database to obtain engineering task information; Based on the engineering task information, determine a dynamic monitoring object and an Internet of Things sensing device for the dynamic monitoring object, and output the dynamic Internet of Things sensing device; Collecting device intrinsic information of the dynamic Internet of Things sensing device, wherein the device intrinsic information at least includes ID information, logical address, encoding protocol, and sensing performance; According to the intrinsic information of the device, a connection is established between the dynamic Internet of Things sensing device and the monitoring management center, and the monitoring management center calls a preset extended collection list to perform extended monitoring, wherein the extended collection list includes dynamic monitoring objects and associated stored collection indicators.

4. The method for real-time monitoring of silo engineering driven by the Internet of Things as claimed in claim 3, characterized in that: Interact with the silo engineering data source, build a silo engineering knowledge graph, and initialize the engineering monitoring discrimination model based on the silo engineering knowledge graph, including: Interactive silo engineering data source to obtain silo engineering literature; Parse the silo engineering literature, define engineering entities and engineering relationships, and construct a silo engineering knowledge graph with the engineering entities as graph nodes and engineering relationships as relationship edges; Based on graph embedding technology, the silo engineering knowledge graph is embedded into a multi-dimensional vector space to obtain an embedded representation of the silo engineering knowledge graph; Based on the neural network algorithm, the engineering monitoring discrimination model is constructed, and the parameters of the silo engineering monitoring discrimination model are initialized according to the embedded representation.

5. The method for real-time monitoring of silo engineering driven by the Internet of Things as claimed in claim 4, characterized in that: Before conducting the supervised training of the silo engineering monitoring and discrimination model, it also includes: Interact with the engineering management database to obtain source monitoring records; Based on the principle of big data, homologous monitoring records are obtained, and commonality analysis is performed on the homologous monitoring records, and monitoring records of common engineering links are extracted and stored as common monitoring records; The original monitoring records and the common monitoring records are integrated to output a monitoring discrimination sample set, and the monitoring discrimination sample set is divided into a monitoring discrimination training set and a monitoring discrimination verification set.

6. The method for real-time monitoring of silo engineering driven by the Internet of Things according to claim 1, characterized in that: The method further comprises: Configure monitoring update constraints, wherein the monitoring update constraints include periodic update constraints and trigger update constraints; Based on the monitoring update constraint, update judgment is performed. If the update interval is greater than or equal to the periodic update constraint or the cumulative number of misjudgments of the engineering monitoring results is greater than or equal to the trigger update constraint, the silo engineering knowledge graph is updated and the silo engineering monitoring judgment model is enhanced and trained.

7. The method for real-time monitoring of silo engineering driven by the Internet of Things as claimed in claim 1, characterized in that: The dynamic interface is configured with a connection protocol for adaptive expansion and contraction, including: When the dynamic interface is idle, the connection of the dynamic interface is adaptively and actively downgraded, wherein the active downgrade includes horizontal downgrade and vertical downgrade; When the dynamic interface is busy, the connection of the dynamic interface is actively expanded adaptively, wherein the active expansion includes horizontal expansion and vertical expansion.

8. The method for real-time monitoring of silo engineering driven by the Internet of Things as claimed in claim 1, characterized in that: Performing management response based on the engineering monitoring results also includes: Analyze the risk identification information to obtain risk categories and risk levels; Taking the risk category and the risk level as index constraints, traverse the preset risk response matrix to call the response plan and perform risk management response; Compare the project progress information with the expected progress and perform project progress management.

Images