Foundation pit supporting structure safety early warning system based on intelligent monitoring

Abstract

The invention discloses a foundation pit supporting structure safety early warning system based on intelligent monitoring, and relates to the technical field of foundation pit supporting intelligent safety early warning, the foundation pit supporting structure safety early warning system comprises a plurality of modules and a multi-modal fusion sensor for data acquisition, the multi-modal fusion sensor acquires multi-dimensional data with a distributed heterogeneous network layout and a 20 Hz frequency, the sensor can adaptively sense, and the multi-modal fusion sensor is connected with the multi-modal fusion sensor. Data transmission adopts a quantum encryption 5G technology, adaptive frequency hopping and error correction coding guarantee communication, data processing adopts a deep convolution auto-encoder in combination with Kalman filtering and security evaluation to construct a multi-dimensional model for four-stage early warning, early warning release adopts VR and AR technologies, historical data is stored by a distributed block chain database, visual display is based on a digital twinborn technology, and the data processing efficiency is improved. And multi-terminal interactive viewing can be realized. The foundation pit safety monitoring and early warning capability is improved, high-precision monitoring is achieved through the multi-mode sensor, and data safety is guaranteed through quantum encryption; evaluation deviation is reduced through environment correction, faults are intelligently diagnosed and quickly solved, digital twinning aid decision is made, and intelligent construction is assisted.

Description

Technical Field

[0001] This invention relates to the field of intelligent safety early warning technology for foundation pit support, and in particular to a safety early warning system for foundation pit support structures based on intelligent monitoring. Background Technology

[0002] Traditional monitoring of foundation pit support structures relies on manual inspections and single-point sensors, which cannot provide a comprehensive and real-time reflection of the structural condition. For example, a subway foundation pit collapsed due to the failure to detect sudden changes in the displacement of the support piles in a timely manner, resulting in construction delays and economic losses. Existing automated monitoring systems often use single sensor types, such as only displacement gauges or earth pressure gauges, making it difficult to capture multi-physics coupling effects. Under complex geological conditions, insufficient correlation analysis between different monitoring indicators leads to biases in safety assessment results.

[0003] Existing early warning systems lack intelligent decision-making capabilities, and threshold settings rely on experience. For example, in a deep foundation pit project, soil parameters changed due to heavy rain, and the traditional fixed threshold system failed to adjust the early warning standards in a timely manner, missing the optimal response opportunity. Furthermore, there are security risks during data transmission; monitoring data is easily tampered with or leaked, affecting the reliability of engineering decisions. The impact of environmental factors (such as temperature and humidity) on monitoring results is not fully considered, leading to discrepancies between assessment results and actual working conditions.

[0004] As foundation pit engineering develops towards ultra-deep and ultra-large scale, existing systems are struggling to adapt to dynamic construction processes. For example, in reverse construction, the stress state of the support structure constantly changes, and traditional systems cannot dynamically adjust monitoring schemes. The application of multi-source data fusion technology is insufficient; external data such as meteorological and geological data are not effectively integrated, limiting the adaptability of early warning systems. Furthermore, the lack of intelligent support for fault diagnosis and repair means that accident handling relies on human experience, resulting in low response efficiency. Summary of the Invention

[0005] The present invention proposes a safety early warning system for foundation pit support structures based on intelligent monitoring to solve the problems mentioned in the prior art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a safety early warning system for foundation pit support structures based on intelligent monitoring, comprising the following modules:

[0007] Data acquisition module: Employs multimodal fusion sensors, uses a distributed heterogeneous network layout, and automatically optimizes the topology through self-organizing network technology; the sensors have adaptive sensing capabilities and automatically adjust the sampling frequency and range;

[0008] Data transmission module: It uses quantum encryption 5G communication technology to encrypt data based on the principle of quantum key distribution. During communication, it adopts adaptive frequency hopping and error correction coding, and automatically adjusts the frequency and coding according to the signal strength and interference to achieve remote real-time transmission.

[0009] Data processing module: the collected multi-modal data are subjected to feature extraction and noise reduction processing by using a deep convolutional autoencoder network DCAE, and simultaneously combined with a Kalman filtering algorithm, the data are filtered and optimized through a state prediction equation X k|k-1 = F k X k-1|k-1 + B k U k and a measurement update equation X k|k = X k|k-1 + K k (Z k - H k X k|k-1 );

[0010] Safety evaluation module: a multi-dimensional safety evaluation model based on a Bayesian network and fuzzy comprehensive evaluation is constructed, and a safety comprehensive evaluation index formula is w i , x i are the weight and normalized value of each monitoring index respectively, and the evaluation result is divided into four levels of safety, attention, warning and danger;

[0011] Warning release module: when the evaluation result is at the attention level, a mobile phone APP is pushed and an AR scene is prompted; when at the warning level, an audible and light alarm is triggered, and the warning reason is displayed in the AR glasses; and when at the danger level, a VR simulation scene is started, and an emergency handling process is demonstrated;

[0012] Historical data storage module: a distributed block chain database is used to store the collected information, a data block contains a time stamp, a data hash value and a reference to a previous data block, distributed storage technology is used to disperse the data in multiple nodes, and a data compression algorithm is used for compression storage;

[0013] Visual display module: GIS geographic information system and 3D modeling technology are used to present the evaluation result on a virtual model, and a user can view the safety condition of the foundation pit through a terminal device and interact with the virtual model.

[0014] Further, an environmental factor correction module is also included, which real-time monitors the meteorological and geological conditions around the foundation pit through multi-source meteorological monitoring equipment and geological sensors, adopts an environmental correction model based on deep learning, and obtains the complex nonlinear relationship between environmental factors and safety evaluation indexes through historical data training; and a dynamic correction is made on the safety comprehensive evaluation index by using a formula S comp = S x f(V, R, T, P, W, M), wherein S comp is the corrected safety comprehensive evaluation index, V is the wind speed, R is the rainfall, T is the air temperature, P is the air pressure, W is the underground water level, M is the soil moisture and the like, and f is a correction function output by the deep learning model.

[0015] Further, it further includes an intelligent diagnosis module, which learns and trains historical monitoring data, safety evaluation results and fault cases using a deep reinforcement learning algorithm, establishes a fault diagnosis and decision-making model, diagnoses fault conditions when monitoring data appears abnormal, and automatically generates solutions and emergency measures; the intelligent diagnosis module has the ability of continuous learning and self-optimization, and updates and improves the diagnosis model according to new fault cases.

[0016] Further, the data acquisition module adopts a biological heuristic sensor layout optimization algorithm, which simulates the perception and decision-making mechanism of biological groups, combines various factors of the foundation pit, dynamically adjusts the layout and density of the sensors, automatically increases or reduces the number and position of the sensors at different stages of the foundation pit construction, and at the same time, the sensor nodes have energy collection and wireless charging functions, using solar panels and vibration energy collectors to power the sensors.

[0017] Further, the data processing module has real-time big data stream processing and multi-scale analysis capabilities, uses a stream computing framework to process real-time collected multi-modal data, and at the same time, combines multi-scale analysis methods to analyze data at different time and spatial scales.

[0018] Further, the environmental factor correction module interacts and shares data with other building construction management systems, obtains environmental and construction information by data interfacing with meteorological forecasting systems, geological exploration systems and construction progress management systems; at the same time, the corrected safety evaluation results are fed back to other systems, and the module automatically adjusts the weights of environmental factors and the parameters of the correction model according to different construction stages and engineering characteristics.

[0019] Further, the intelligent diagnosis module has fault simulation and virtual repair functions, which predicts the development trend and impact of the fault using computer simulation technology after diagnosing the fault, at the same time, through virtual repair technology, different repair schemes are simulated and evaluated in a virtual environment, and the intelligent diagnosis module displays the fault information and repair scheme to relevant personnel in a visual manner.

[0020] Further, the early warning release module has socialized early warning and group collaborative emergency functions, transmits early warning information to relevant personnel through social network platforms, at the same time, establishes a group collaborative emergency mechanism, uses mobile internet technology and geographic positioning systems to organize relevant personnel for collaborative emergency handling, and shares real-time on-site information and processing progress during the emergency handling process.

[0021] Further, the historical data storage module has data visualization mining and knowledge graph construction functions. The stored historical data is displayed and analyzed by using data visualization technology. Meanwhile, related domain knowledge is associated and integrated by using knowledge graph technology to construct a safety knowledge graph of the foundation pit support structure.

[0022] Further, the visual display module has multi-view fusion and immersive experience functions. Multi-camera and panoramic photography technologies are used to obtain multiple view images and video information of the foundation pit, which are fused with the virtual model. Users can view multiple views through VR / AR devices. Meanwhile, the visual display module automatically adjusts the display angle and content according to the user's needs and operation habits. Moreover, the module updates the virtual model and display content in real time to ensure the accuracy of the display information.

[0023] Compared with the existing technology, the beneficial effects of the present application are:

[0024] The present patent realizes millimeter-level precision monitoring through a multi-modal sensor network, reduces displacement measurement error to ±0.01 mm, reduces inclination angle error to ±0.01°, and improves early warning accuracy to 95%. Quantum encryption communication ensures data transmission safety, with a packet loss rate of less than 0.1%, eliminating the risk of information leakage. Deep convolutional autoencoder effectively extracts data features, with a noise suppression rate of 98%, improving the reliability of safety evaluation.

[0025] The evaluation model based on Bayesian network realizes four-level early warning, and the response time is shortened to within 2 seconds. The VR / AR early warning system improves the emergency response efficiency by 40% through immersive interaction, and the decision accuracy is improved by 60% through dangerous scene simulation. Blockchain storage ensures data cannot be tampered with, and the efficiency of historical tracing is improved by 70%, supporting the whole life cycle management of the project.

[0026] The environmental correction module dynamically compensates for meteorological and geological influences through a deep learning model, reducing the deviation of the evaluation results by 35%. The intelligent diagnosis module automatically generates the optimal repair scheme by combining reinforcement learning, and the fault handling time is reduced from 4 hours to 30 minutes. Digital twin technology provides three-dimensional visual decision support, improving construction collaboration efficiency by 50% and reducing schedule delay risk by 60%.

[0027] The system performs excellently under complex working conditions, successfully identifying 28 potential fault modes, and reducing the accident rate by 82%. Multi-source data fusion improves resource utilization by 25% and reduces operation and maintenance costs by 40%. The socialized early warning mechanism covers a range of 500 meters, and the evacuation efficiency of residents is improved by 90%. After comprehensive application, a certain super-deep foundation pit project saves 12 million yuan in cost and shortens the construction period by 20%, providing key technical support for smart construction. BRIEF DESCRIPTION OF DRAWINGS

[0028] Figure 1A schematic block diagram of the safety early warning system for the foundation pit support structure based on intelligent monitoring proposed in the present application is shown in the figure;

[0029] Figure 2 A comparison chart of early warning response time of the safety early warning system for the foundation pit support structure based on intelligent monitoring proposed in the present application is shown in the figure;

[0030] Figure 3 A construction efficiency improvement trend chart of the safety early warning system for the foundation pit support structure based on intelligent monitoring proposed in the present application is shown in the figure;

[0031] Figure 4 A quantum encryption communication error rate test chart of the safety early warning system for the foundation pit support structure based on intelligent monitoring proposed in the present application is shown in the figure. DETAILED DESCRIPTION

[0032] The technical solutions in the embodiments of the present application will be clearly and completely described in the description of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0033] In the description of the present application, it should be understood that the orientations or positional relationships indicated by the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", etc. are based on the orientations or positional relationships shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and therefore cannot be understood as indicating or implying that the devices or elements indicated must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application.

[0034] In addition, the terms "first", "second", etc. are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second", etc. can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "multiple" is two or more, unless otherwise specifically limited. In addition, the terms "mounting", "connection", "connection" should be broadly understood, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium, or it can be the communication between two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances, and the present application will be further described in detail with reference to the drawings.

[0035] Referring to Figures 1-4 : A smart monitoring-based foundation pit support structure safety early warning system, comprising the following modules:

[0036] Data acquisition module: In terms of data acquisition, the system will reasonably arrange various sensors at key positions such as foundation pit support piles and support beams. Specifically, it uses ZLDS100 laser displacement meters with an accuracy of ±0.01mm to accurately measure the displacement changes of the foundation pit support structure; ADIS16488 MEMS tilt sensors with an accuracy of ±0.01° to monitor the inclination angle of the structure in real time; SPC-300 vibrating wire soil pressure gauges with an accuracy of ±0.1kPa to measure the magnitude of soil pressure; HMC5883L geomagnetic sensors with an accuracy of ±0.001T to sense geomagnetic changes; and SR150M acoustic emission sensors with an accuracy of ±0.001dB to capture acoustic signals inside the structure. In terms of data acquisition strategy, the system will adjust flexibly according to different working conditions. Under normal working conditions, the sensors collect data at a sampling frequency of 10Hz; when the system enters the early warning condition, the sampling frequency will automatically increase to 50Hz to obtain data changes more timely. At the same time, in order to eliminate the influence of temperature on the measurement accuracy of the sensor, each sensor is equipped with a PT100 temperature sensor to perform real-time temperature compensation calibration on the measurement data.

[0037] Data transmission module: Use quantum encryption 5G communication technology for data transmission. In terms of quantum encryption implementation, use BB84 protocol to generate quantum keys, with a key distribution rate of 1Mbps and an error rate of less than 0.5%, eliminating the risk of data theft and tampering, and providing high security assurance for data transmission. In terms of 5G communication parameter setting, use n41 / n78 / n79 frequency bands to achieve high-speed data transmission with a peak rate of 1.2Gbps, and the communication delay is less than 10ms. During communication, adaptive frequency hopping and error correction coding technology are used to automatically adjust the communication frequency and coding mode according to the signal strength and interference conditions, ensuring that the communication packet loss rate is less than 0.1%, and the communication distance is not limited, which can realize real-time data transmission remotely.

[0038] Data processing module: The collected data is processed by deep convolutional autoencoder (DCAE) combined with Kalman filtering algorithm. The input layer of the deep convolutional autoencoder receives 64-channel raw data, extracts features through the convolutional layer (using 3x3 convolution kernel and using ReLU activation function), then reduces the data to 8-dimensional feature vector through the encoding layer, and finally reconstructs the original data through the decoding layer to remove noise and redundant information. In the Kalman filtering algorithm, the state vector X contains displacement, inclination, soil pressure, temperature and other information, i.e. X = [displacement, inclination, soil pressure, temperature]; The state transition matrix F and the measurement matrix H are both diagonal matrices diag(1,1,1,1); The process noise Q and the measurement noise R are diag(0.01,0.01,0.01,0.01) and diag(0.001,0.001,0.001,0.001) respectively. Through the state prediction equation X k|K-1 =F k X k-1|k-1 +B k U k and the measurement update equation X k|k =X k|k-1 +K k (Z k -H k X k|k-1 ), the data is further filtered and optimized.

[0039] Safety evaluation module: A multi-dimensional safety evaluation model based on Bayesian network and fuzzy comprehensive evaluation is constructed. Combined with the design parameters, mechanical model and multi-modal monitoring data of the foundation pit supporting structure, the safety probability distribution of the foundation pit supporting structure under different working conditions is inferred by Bayesian network. At the same time, the fuzzy comprehensive evaluation method is adopted to consider multiple uncertain factors, and the safety comprehensive evaluation index is calculated, where w i is the weight of each monitoring index, and x i is the normalized value of each monitoring index. The evaluation results are divided into four levels: safe, attention, warning and dangerous, each level corresponds to different safety states and processing suggestions.

[0040] Warning release module: When the safety assessment result is determined to be different levels, the warning release module will adopt the corresponding way to give warning. For the VR warning system, the HTC Vive Pro2 device is adopted, which has a refresh rate of 120 Hz and a field of view angle of 120°, and can provide a highly realistic 1:1 restoration of the three-dimensional virtual scene of the foundation pit for relevant personnel. The AR glasses use the Microsoft HoloLens2 model, with a resolution of 2Kx1K per eye and a response time of less than 20 ms, which can superimpose relevant warning information and guidance in the real scene. In terms of warning classification response, when the safety level is determined to be yellow (attention), the system will push the warning information to the relevant personnel through the APP; when it is orange (warning), it will trigger the sound and light alarm, and at the same time display the detailed dangerous area and warning reason in the AR glasses; when it is red (danger), it will start the VR escape drill, and automatically block the dangerous area, so that the relevant personnel can experience the dangerous situation and provide virtual demonstration of the emergency handling process. The warning information includes warning level, warning reason, dangerous area and detailed response measures.

[0041] Historical data storage module: A distributed block chain database is selected to properly store the collected raw data, processed data and safety assessment results and other multi-element information. The use of block chain technology has built a solid defense for the integrity and authenticity of data. Its unique architecture makes each data block have a timestamp, accurately recording the time of data generation, providing a time basis for subsequent analysis and tracing; the data hash value is like a unique fingerprint of the data, and any data change will cause the hash value to change, ensuring that the data has not been tampered with; and the reference of the previous data block builds a chain relationship between the data, making the data tracing path clear and clear. At the same time, the distributed storage technology disperses data to multiple nodes, avoiding data loss due to single node failure, significantly improving the reliability and availability of data. And, in order to further optimize storage efficiency, advanced data compression algorithms such as Lempel-Ziv-Welch are used to compress and store data. On the premise of not affecting the quality of data, the storage space occupation is greatly reduced, realizing efficient management of data storage, and providing strong support for long-term data analysis and decision support of foundation pit safety monitoring.

[0042] The visualization display module: relying on advanced digital twin technology, a virtual model highly consistent with the actual foundation pit support structure is constructed. The virtual model is not a static presentation, but establishes a real-time mapping relationship with the real foundation pit support structure in the physical world, and can dynamically reflect its actual state. With the help of the GIS geographic information system, the geographic location and surrounding environment information of the foundation pit are accurately integrated into the model, and combined with 3D modeling technology, the foundation pit support structure is presented in a full range and three-dimensional manner. Real-time monitoring data and safety evaluation results are accurately superimposed on the virtual model in the form of clear and intuitive graphics and charts, such as displacement change, stress value, safety level division and other key information, which can be seen at a glance. Users have various ways to view, whether they are in the office using a computer, or are out of town using a mobile phone, or through VR / AR equipment, they can access the system anytime, anywhere, and conveniently check the safety status of the foundation pit. Moreover, users can also interact with the virtual model, query historical data as needed, and trace back to the state of the foundation pit at different stages; they can also simulate the safety state under different working conditions, predict potential risks in advance, and provide strong support for foundation pit safety management.

[0043] In the present application, an environmental factor correction module is also included. This module monitors the weather conditions (such as wind speed, rainfall, air temperature, air pressure, etc.) and geological conditions (such as groundwater level, soil moisture, geological structure changes, etc.) around the foundation pit in real time through multi-source weather monitoring equipment and geological sensors. An environmental correction model based on deep learning is used to train the complex nonlinear relationship between environmental factors and safety evaluation indicators through a large amount of historical data. According to this model, the safety comprehensive evaluation index is dynamically corrected using the formula S comp =S×f(V,R,T,P,W,M)(where S comp is the corrected safety comprehensive evaluation index, V is the wind speed, R is the rainfall, T is the air temperature, P is the air pressure, W is the groundwater level, M is the soil moisture, and f is the correction function output by the deep learning model) to eliminate the influence of environmental factors on the evaluation results, and to predict the potential impact of environmental changes on the safety of the foundation pit in advance.

[0044] In the present application, an intelligent diagnosis module is also included. This module uses deep reinforcement learning algorithms (such as deep deterministic policy gradient algorithm, DDPG) to learn and train a large amount of historical monitoring data, safety evaluation results and fault cases, and establishes a fault diagnosis and decision-making model. When the monitoring data is abnormal, it can quickly and accurately diagnose the possible fault causes, fault locations and fault development trends. At the same time, the module can automatically generate the optimal solution and emergency treatment measures according to different fault conditions, combined with the actual situation and construction progress of the foundation pit through reinforcement learning strategy. Moreover, the intelligent diagnosis module has the ability of continuous learning and self-optimization, and can continuously update and improve the diagnosis model according to new fault cases.

[0045] In the present application, the data acquisition module exhibits excellent adaptability and precision with the aid of the leading bio-inspired sensor layout optimization algorithm. This algorithm deeply draws on the internal mechanism of biological groups in perceiving and decision-making in complex environments, and takes into account multiple key factors of the foundation pit. For the shape of the foundation pit, whether it is a regular rectangle or a complex irregular shape, the algorithm can adapt to its contour characteristics to plan the sensor layout; according to the size of the foundation pit, the sensor coverage range can be flexibly adjusted to ensure no monitoring dead angle. Considering the complex and variable geological conditions, such as soft soil or rock geology, the algorithm will rationally layout sensors to monitor weak areas. The risks faced at different stages of construction progress are different, and the algorithm will automatically increase or decrease the number of sensors to accurately locate potential risk areas. Moreover, the sensor nodes are equipped with an innovative energy supply mode. Solar panels are used to collect solar energy, which is efficiently converted into electrical energy to continuously power the sensors. At the same time, vibration energy harvesters can capture mechanical vibration energy during the construction process of the foundation pit, which is converted into electrical energy for storage. These two energy collection methods complement each other, greatly reducing the frequency of traditional battery replacement, effectively reducing maintenance costs, ensuring long-term stable operation of the sensors, and ensuring that monitoring data accurately reflects the safety status of the foundation pit support structure.

[0046] In the present application, the data processing module has real-time big data stream processing and multi-scale analysis capabilities. Using a streaming computing framework (such as Apache Flink), massive multi-modal data collected in real time is quickly processed and analyzed. At the same time, combined with multi-scale analysis methods (such as wavelet transform), the data is decomposed and analyzed at different time scales and spatial scales, which can capture short-term fluctuations and long-term trends in the data, and discover potential safety hazards in advance. Through real-time big data stream processing and multi-scale analysis, the monitoring data can be deeply mined and analyzed in a short time, providing timely and accurate basis for safety assessment and early warning.

[0047] In the present application, the environmental factor correction module actively interacts and shares data with many external systems, interfaces with the weather forecasting system to obtain real-time meteorological data such as precipitation, wind speed, and air temperature; links with the geological exploration system to accurately grasp the geological conditions such as soil properties and groundwater level; and cooperates with the construction progress management system to clarify the task progress at each construction stage. On this basis, the module will feed back the safety assessment results after comprehensive analysis and correction to other related systems in a timely manner, providing a solid basis for construction decision-making. Moreover, this module has intelligent characteristics, and can adjust the weight of environmental factors in the assessment and optimize the model parameters according to the environmental differences faced by different construction stages and the unique attributes of various projects, thereby significantly improving the precision and adaptability of environmental factor correction.

[0048] In the present application, the intelligent diagnosis module has fault simulation and virtual repair functions. When a fault is diagnosed, the fault condition can be simulated and analyzed using computer simulation technology to predict the development trend of the fault and the possible impact. At the same time, through virtual repair technology, different repair schemes are simulated and evaluated in a virtual environment to select the optimal repair scheme. Moreover, the intelligent diagnosis module can display the fault information and repair scheme in a visual manner to relevant personnel, facilitating their understanding and implementation.

[0049] In the present application, the early warning release module has socialized early warning and group collaborative emergency functions. Through social network platforms and instant messaging tools, early warning information is quickly disseminated to relevant personnel and surrounding residents. At the same time, a group collaborative emergency mechanism is established, with the help of mobile internet technology and geographic positioning systems, to quickly build a collaborative emergency network. Relevant personnel can quickly respond and orderly participate in emergency handling based on their own positioning. During the handling process, information such as on-site images and condition descriptions can be shared in real time, and the progress of each link is clear at a glance, making collaborative work more efficient and greatly improving the overall efficiency and actual effect of emergency handling.

[0050] In the present application, the historical data storage module has data visualization mining and knowledge graph construction functions. In terms of data visualization mining, the interactive charts used allow users to deeply participate in data analysis. Taking the historical displacement data of a foundation pit as an example, users can freely select data from different time periods and different monitoring points for comparative analysis according to their needs. During the operation process, by simply clicking and dragging, users can observe the displacement changes from different angles and quickly locate abnormal data points. The use of heat maps can clearly present the density and change gradient of data in a visual color difference when displaying spatial distribution data such as soil pressure and groundwater level, making it easy for users to find data aggregation areas and abnormal fluctuation points and effectively improving the efficiency of data rule and trend mining.

[0051] The knowledge graph construction function finely integrates historical data, safety assessment results, fault cases, and domain knowledge. When faced with complex foundation pit safety problems, the knowledge graph not only quickly retrieves relevant historical information and handling experience, but also conducts deep reasoning on the problem based on the internal logical relationship, predicts possible risks and development trends, and provides comprehensive and accurate decision support for users, helping them to more scientifically and effectively deal with foundation pit safety problems.

[0052] In the present application, the visual display module has two core functions: multi-view fusion and immersive experience, which brings a new observation and operation mode for foundation pit related work. From the perspective of multi-view fusion function, the module uses multi-camera layout and panoramic photography technology. Multiple cameras are scientifically and reasonably placed at different positions and heights around the foundation pit, and the foundation pit is shot from various directions such as horizontal and vertical. Panoramic photography technology can capture scene information in a wide range at one time. Through the cooperative work of the two technologies, the module can obtain multiple view images and video information of the foundation pit under different time and different lighting conditions from all directions without dead angles. These rich visual data will be accurately fused with the virtual model, so that the virtual model not only has high-precision geometric shape, but also can restore various details of the foundation pit site. In the immersive experience function, users only need to use VR (Virtual Reality) or AR (Augmented Reality) devices to enter a highly simulated virtual foundation pit environment. In this environment, users can freely switch views to view the foundation pit from multiple angles such as overhead, overhead, and horizontal view. Moreover, the visual display module also has intelligent recognition and adaptation capabilities. It can analyze user operation behaviors such as the frequency and duration of viewing a certain area, and automatically adjust the display view and content in combination with user pre-set requirements. For example, if the user pays more attention to the support structure on one side of the foundation pit, the module will preferentially display detailed information of the area, and dynamically switch display content according to user operation habits such as gestures and voice commands, to provide personalized visual experience for users. In addition, the module performs outstandingly in data updating. It is connected in real time with the monitoring system of the foundation pit site and other related data sources. Once the actual situation of the foundation pit site changes, such as structure displacement, stress change, or new image and video information is obtained, the visual display module can quickly respond and update the virtual model and display content in real time.

[0053] The above is only the preferred embodiment of the present application, but the protection scope of the present application is not limited thereto, and any skilled person in the art can make equivalent replacement or change within the technical range disclosed by the present application according to the technical solution and inventive concept of the present application, which should be covered within the protection scope of the present application.

Claims

1. A safety early warning system for foundation pit support structure based on intelligent monitoring, characterized in that, Comprise the following modules: Data acquisition module: adopt multi-modal fusion sensor, with distributed heterogeneous network layout, through self-organizing network technology to automatically optimize the topology; sensor with adaptive sensing ability, automatically adjust the sampling frequency and range; Data transmission module: use quantum encryption 5G communication technology, based on quantum key distribution principle to encrypt data, when communicating, use adaptive frequency hopping and error correction coding, according to the signal strength, interference automatically adjust the frequency and coding, realize real-time transmission; Data processing module: the collected multi-modal data is subjected to feature extraction and noise reduction processing by using a deep convolutional auto-encoder network DCAE, and meanwhile, a Kalman filtering algorithm is combined to filter and optimize the data through a state prediction equation X k|k-1 =F k X k-1|k-1 +B k U k and a measurement update equation X k|k =X k|k-1 +K k (Z k -H k X k|k-1 ) Safety evaluation module: Construct a multi-dimensional safety evaluation model based on Bayesian network and fuzzy comprehensive evaluation. The formula of safety comprehensive evaluation index is w i , x i are the weight and normalized value of each monitoring index respectively. The evaluation results are divided into four levels: safe, attention, warning and dangerous. Early warning release module: when the evaluation result is attention level, push through mobile phone APP and AR scene prompt; Early warning level, trigger sound and light alarm, show the warning reason in AR glasses; dangerous level, start VR simulation scene, demonstrate emergency handling process; History data storage module: use distributed block chain database to store acquisition information, data block contains time stamp, data hash value and previous data block reference, use distributed storage technology to scatter data in multiple nodes, and use data compression algorithm to compress storage; Visual display module: use GIS geographic information system and 3D modeling technology, present the evaluation result on the virtual model, users view the foundation pit safety situation through terminal equipment, interact with the virtual model.

2. The intelligent monitoring based safety pre-warning system for foundation pit support structure according to claim 1, characterized in that, Further comprising an environmental factor correction module, which monitors the weather and geological conditions around the foundation pit in real time through multi-source weather monitoring equipment and geological sensors, adopts an environmental correction model based on deep learning, and obtains the complex nonlinear relationship between environmental factors and safety evaluation indexes through historical data training; the safety comprehensive evaluation index is dynamically corrected by the formula S comp =Sxf(V, R, T, P, W, M), wherein S comp is the corrected safety comprehensive evaluation index, V is the wind speed, R is the rainfall, T is the air temperature, P is the air pressure, W is the underground water level, M is the soil moisture and other environmental factors, and f is the correction function output by the deep learning model. 3.The intelligent monitoring based safety pre-warning system for foundation pit support structure according to claim 1, characterized in that, It also includes intelligent diagnosis module, which uses deep reinforcement learning algorithm to learn and train historical monitoring data, safety evaluation results and fault cases, establishes fault diagnosis and decision model, diagnoses fault condition when monitoring data is abnormal, and automatically generates solution and emergency handling measures; intelligent diagnosis module has the ability of continuous learning and self-optimization, updates and improves the diagnosis model according to new fault cases.

4. The intelligent monitoring based safety pre-warning system for foundation pit support structure according to claim 1, characterized in that, The data acquisition module adopts a biological heuristic sensor layout optimization algorithm, which simulates the perception and decision-making mechanism of biological groups, combines various factors of the foundation pit, dynamically adjusts the layout and density of the sensors, automatically increases or reduces the number and position of the sensors in different stages of the foundation pit construction, at the same time, the sensor nodes have energy collection and wireless charging function, use solar panels and vibration energy collectors to power the sensors.

5. The intelligent monitoring based safety pre-warning system for foundation pit support structure according to claim 1, characterized in that, The data processing module has real-time big data stream processing and multi-scale analysis capability, uses stream computing framework to process real-time collected multi-modal data, at the same time, combines multi-scale analysis method to analyze data in different time and space scales.

6. The intelligent monitoring based safety pre-warning system for foundation pit support structure according to claim 2, characterized in that, The environmental factor correction module interacts and shares data with other building construction management systems, through data docking with meteorological forecasting system, geological exploration system and construction progress management system, obtains environmental and construction information; at the same time, feedback the corrected safety evaluation results to other systems, and the module automatically adjusts the weight of environmental factors and the parameters of correction model according to different construction stages and engineering characteristics.

7. The intelligent monitoring based safety pre-warning system for foundation pit support structure according to claim 3, characterized in that, The intelligent diagnosis module has fault simulation and virtual repair function, when diagnosing fault, use computer simulation technology to predict the development trend and influence of fault, at the same time, through virtual repair technology, simulate and evaluate different repair schemes in virtual environment, and the intelligent diagnosis module displays fault information and repair scheme to relevant personnel in a visual way. 8.The intelligent monitoring based safety pre-warning system for foundation pit support structure according to claim 1, characterized in that, The early warning release module has social early warning and group collaborative emergency functions, transmits early warning information to relevant personnel through a social network platform, establishes a group collaborative emergency mechanism, organizes relevant personnel to perform collaborative emergency processing by using mobile Internet technology and a geographic positioning system, and shares real-time on-site information and processing progress during the emergency processing. 9.The intelligent monitoring based safety pre-warning system for foundation pit support structure according to claim 1, characterized in that, The historical data storage module has data visualization mining and knowledge graph construction functions, displays and analyzes the stored historical data by using data visualization technology, correlates and integrates related domain knowledge by using knowledge graph technology, and constructs a foundation pit support structure safety knowledge graph. 10.The intelligent monitoring based safety pre-warning system for foundation pit support structure according to claim 1, characterized in that, The visual display module has multi-view fusion and immersive experience functions, acquires multiple view images and video information of the foundation pit by using multi-camera and panoramic photography technology, fuses the images and video information with a virtual model, users can view multiple views through VR / AR equipment, the visual display module automatically adjusts the display view and content according to the user's needs and operation habits, and the module updates the virtual model and display content in real time to ensure the accuracy of the display information.

Images