Video GNSS micro-core pile intelligent monitoring method and system

Abstract

The invention provides a video GNSS micro-core pile intelligent monitoring method and system, and relates to the technical field of safety monitoring, and the method comprises the steps: collecting real-time position data, real-time inclination angle data and real-time vibration data of a micro-core pile; the current displacement, the current angle change value and the current vibration amplitude of the micro-core pile are determined; calculating displacement, inclination and vibration danger indexes based on the current displacement, the current angle change value and the current vibration amplitude of the micro-core pile; determining dynamic adjustment weights of displacement, inclination and vibration danger indexes; according to the displacement, inclination and vibration danger indexes, the corresponding dynamic adjustment weights are combined, and the comprehensive danger index of the micro-core pile is calculated; judging whether the comprehensive danger index is greater than a preset danger index; if yes, a camera on the micro-core pile is awakened to collect an image, the image is input into a risk prediction model based on the deep convolutional neural network, and a risk prediction result is output; otherwise, monitoring is continued.

Description

Technical Field

[0001] The present invention relates to the field of safety monitoring technology, and in particular to a video GNSS micro-core pile intelligent monitoring method and system. Background Art

[0002] With the rapid development of Global Navigation Satellite System (GNSS) technology and its application in engineering monitoring, GNSS-based monitoring methods have gradually become a mainstream technology in civil engineering, geological disaster monitoring, and infrastructure management. In micro-core pile monitoring, combining video monitoring with GNSS technology enables high-precision, real-time dynamic monitoring, which is particularly important for monitoring engineering structures in complex and challenging environments. This innovative application of technology not only improves monitoring accuracy but also achieves stable monitoring results under diverse environmental conditions.

[0003] Currently, GNSS-based monitoring technology has been widely used in civil engineering, primarily in areas such as structural deformation monitoring, settlement monitoring, and displacement monitoring. Traditional GNSS monitoring methods rely on deploying multiple base stations and receivers, combined with high-precision GNSS satellite signals, to achieve high-precision positioning of the target location. Furthermore, with the development of video monitoring technology, the integration of video monitoring and GNSS is gaining popularity. Using video images to monitor the target area in real time, combined with GNSS data analysis, can more accurately capture even the smallest structural deformations or displacements.

[0004] However, in existing methods, video cameras, as high-power sensing devices, rely on batteries for power supply, which consumes too much energy, resulting in shortened working hours and even interruption of monitoring due to power exhaustion, affecting the safety monitoring effect; when no disaster occurs, the invalid data captured by the camera wastes a lot of storage space and bandwidth, increasing system operating costs. Summary of the Invention

[0005] To address the challenges of existing multi-signal fusion methods, which, while able to compensate for the shortcomings of a single signal, still face complex signal processing and computational complexity, making them difficult to meet real-time requirements. Furthermore, due to the nonlinearity and uncertainty of tool wear, existing methods suffer from low monitoring accuracy in the early stages of wear, when signal changes are insignificant. The present invention provides a video GNSS micro-core pile intelligent monitoring method and system.

[0006] The technical solutions provided by the embodiments of the present invention are as follows:

[0007] First aspect:

[0008] An embodiment of the present invention provides a video GNSS micro-core pile intelligent monitoring method, comprising:

[0009] S1: collecting real-time position data, real-time inclination angle data and real-time vibration data of the micro-core pile;

[0010] S2: determining the current displacement, current angle change value and current vibration amplitude of the micro-core pile according to the real-time position data, the real-time tilt angle data and the real-time vibration data;

[0011] S3: Calculating a displacement risk index, a tilt risk index, and a vibration risk index based on the current displacement, the current angle change value, and the current vibration amplitude of the micro-core pile;

[0012] S4: Determining dynamic adjustment weights of the displacement risk index, the tilt risk index, and the vibration risk index;

[0013] S5: Calculating a comprehensive risk index of the micro-core pile according to the displacement risk index, the tilt risk index, and the vibration risk index in combination with corresponding dynamic adjustment weights;

[0014] S6: Determine whether the comprehensive risk index is greater than a preset risk index; if so, wake up the camera on the micro-core pile to collect multiple images of the area to be monitored and proceed to the next step; otherwise, continue monitoring;

[0015] S7: Input the image of the area to be monitored into a risk prediction model based on a deep convolutional neural network, and output a risk prediction result.

[0016] Second aspect:

[0017] An embodiment of the present invention provides a video GNSS micro-core pile intelligent monitoring system, comprising:

[0018] processor;

[0019] A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the video GNSS micro-core pile intelligent monitoring method as described in the first aspect is implemented.

[0020] The third aspect:

[0021] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the video GNSS micro-core pile intelligent monitoring method as described in the first aspect is implemented.

[0022] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0023] In an embodiment of the present invention, the comprehensive risk index of the micro-core pile is calculated, and the relationship between the comprehensive risk index and the preset risk index is analyzed to determine whether to wake up the camera. The camera is activated only when an actual danger occurs, avoiding continuous filming when there is no disaster, thereby effectively reducing energy consumption and extending the camera's operating time. At the same time, image acquisition and input into the risk prediction model for analysis are only performed when actually needed, avoiding the generation of invalid video data, reducing storage and bandwidth waste, and improving the system's resource utilization efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0025] Figure 1 A schematic diagram of a flow chart of a video GNSS micro-core pile intelligent monitoring method provided in an embodiment of the present invention;

[0026] Figure 2 A schematic structural diagram of a video GNSS micro-core pile intelligent monitoring system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0028] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as an "exemplary" in the present invention should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of the word "exemplary" is intended to present concepts in a concrete manner. Furthermore, in the embodiments of the present invention, "and / or" can mean both or either of the two.

[0029] In the embodiments of the present invention, the terms "image" and "picture" may be used interchangeably. It should be noted that, when the distinction between them is not emphasized, their intended meanings are the same. The terms "of," "corresponding," and "corresponding" may be used interchangeably. It should be noted that, when the distinction between them is not emphasized, their intended meanings are the same.

[0030] In the embodiments of the present invention, sometimes a subscript such as W1 may be written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are the same.

[0031] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0032] Reference Manual Figure 1 , which shows a flow chart of a video GNSS micro-core pile intelligent monitoring method provided by an embodiment of the present invention.

[0033] The embodiment of the present invention provides a video GNSS micro-core pile intelligent monitoring method, which can be implemented by a video GNSS micro-core pile intelligent monitoring device, which can be a terminal or a server. The processing flow of the video GNSS micro-core pile intelligent monitoring method may include the following steps:

[0034] S1: Collect real-time position data, real-time inclination angle data and real-time vibration data of the micro-core pile.

[0035] In a possible implementation, S1 specifically includes:

[0036] Collect location data through GNSS sensors.

[0037] GNSS sensors calculate the device's location (longitude, latitude, and altitude) by receiving radio signals from multiple satellites. By measuring the signal's travel time and combining it with known satellite positions, the sensor's precise location can be calculated.

[0038] The tilt angle data is collected through the tilt angle sensor.

[0039] Among them, the tilt angle sensor (also known as tilt sensor or tilt sensor) is a device used to measure the tilt angle of an object relative to a horizontal or vertical plane. It provides real-time angle change data by detecting the change in the tilt angle of the object.

[0040] Vibration data is collected through MEMS accelerometers.

[0041] Optionally, the MEMS acceleration sensor is a dual-axis acceleration sensor to accurately obtain vibration data.

[0042] MEMS accelerometers operate based on the principle of inertia. They use tiny mechanical components (such as micro-masses, springs, and capacitive or resistive sensors) to sense changes in acceleration. Accelerometers primarily determine acceleration by measuring the physical changes caused by acceleration (such as the displacement of the micro-masses).

[0043] In the present invention, combining these data can fully grasp the motion state and structural stability of the micro-core pile. Especially for complex monitoring tasks (such as earthquake monitoring, building settlement monitoring, etc.), the combination of multi-dimensional data provides more information and improves the accuracy and real-time performance of monitoring.

[0044] S2: Determine the current displacement, current angle change value, and current vibration amplitude of the micro-core pile based on the real-time position data, real-time tilt angle data, and real-time vibration data.

[0045] In a possible implementation, the current displacement, current tilt angle, and current vibration amplitude of the micro-core pile are determined by:

[0046] The position information of the micro-core pile at two adjacent moments is obtained, and the displacement at the two moments is calculated as the current displacement of the micro-core pile using the Euclidean distance formula.

[0047] Euclidean distance is a mathematical metric used to represent the straight-line distance between two points. It is one of the most common distance metrics and is widely used in fields such as geometry, machine learning, and image processing.

[0048] Obtain the inclination angle data of the micro-core pile at two adjacent moments, and calculate the difference between the inclination angle data at the two adjacent moments by subtraction operation as the current angle change value of the micro-core pile.

[0049] The real-time vibration data is collected by using MEMS acceleration sensor, and the vibration amplitude of the real-time vibration data is extracted through Fourier transform as the current vibration amplitude of the micro-core pile.

[0050] The Fourier transform is a mathematical transformation used to convert a function from the time domain (or spatial domain) to the frequency domain. It is based on Fourier analysis theory, which states that any periodic signal can be represented as a superposition of sine and cosine waves of different frequencies. The Fourier transform provides an effective tool for analyzing the frequency components of non-periodic signals.

[0051] In the present invention, the vibration signal usually contains multiple components of different frequencies, and Fourier transform can help accurately identify and extract the vibration amplitude of each frequency, thereby providing more accurate and reliable data support for vibration monitoring of micro-core piles.

[0052] S3: Based on the current displacement, current angle change value and current vibration amplitude of the micro-core pile, a displacement risk index, a tilt risk index and a vibration risk index are calculated.

[0053] It should be noted that the displacement hazard index reflects the degree of micro-core pile displacement, helping to determine whether excessive displacement has occurred, potentially leading to structural instability or failure. The tilt hazard index measures changes in the micro-core pile's tilt angle, indicating whether the micro-core pile has deviated from its normal vertical position, potentially leading to pile instability or structural risks. The vibration hazard index indicates whether the micro-core pile's vibration amplitude exceeds the normal range. Excessive vibration may cause structural damage or instability.

[0054] In a possible implementation, S3 specifically includes:

[0055] Based on the current displacement, current angle change value and current vibration amplitude of the micro-core pile, the displacement risk index, tilt risk index and vibration risk index are calculated using the following formula:

[0056]

[0057] Among them, V d (t) represents the displacement risk index at time t, V t represents the displacement of the micro-core pile at time t, V safe Represents the safe displacement threshold, V max Indicates the maximum allowable displacement, V θ (t) represents the tilt risk index at time t, θ t represents the angle change of the micro-core pile at time t, θ safe Indicates the safe angle change value, θ max Indicates the maximum allowable value of the angle change, V v (t) represents the vibration hazard index at time t, A t Indicates the vibration amplitude at time t, A safe Indicates the safe vibration amplitude threshold, A max Indicates the maximum allowable value of the vibration amplitude.

[0058] In this invention, displacement, tilt, and vibration are three key factors affecting the safety of micro-core piles. Comprehensively considering these factors can yield a more comprehensive and accurate risk assessment, avoiding the inability of a single indicator to fully reflect the problem and improving monitoring accuracy. Furthermore, by calculating these risk indices in real time, changes in the micro-core pile's status can be quickly reflected, enabling the system to respond quickly to sudden safety threats.

[0059] S4: Determine the dynamic adjustment weights of the displacement risk index, the tilt risk index, and the vibration risk index.

[0060] In a possible implementation, S4 specifically includes:

[0061] S401: Calculate the initial dynamic adjustment weights of the displacement risk index, the tilt risk index, and the vibration risk index using the following formula:

[0062]

[0063]

[0064] Among them, W v (t) represents the initial dynamic adjustment weight of the vibration hazard index at time t, W v (t-1) represents the initial dynamic adjustment weight of the vibration risk index at time t-1, λ represents the weight adjustment coefficient of the control vibration risk index, V v (t) represents the vibration hazard index at time t, V v (t-1) represents the vibration hazard index at time t-1, A max Indicates the maximum allowable value of vibration, W d (t) represents the initial dynamic adjustment weight of the displacement risk index at time t, W d (t-1) represents the dynamic adjustment weight of the displacement risk index at time t-1, μ represents the weight adjustment coefficient of the displacement risk index, V d (t) represents the displacement risk index at time t, V d (t-1) represents the displacement risk index at time t-1, V max Indicates the maximum allowable displacement, W θ (t) represents the dynamic adjustment weight of the tilt risk index at time t, W θ (t-1) represents the dynamic adjustment weight of the tilt risk index at time t-1, γ represents the weight adjustment coefficient of the tilt risk index, V θ (t) represents the tilt risk index at time t, V θ (t-1) represents the tilt risk index at time t-1, θ max Indicates the maximum allowable tilt value.

[0065] Optionally, λ, μ, and γ are first preliminarily set according to the actual application scenario and monitoring requirements. Then, they are appropriately adjusted based on historical data and expert experience. Finally, the setting accuracy is tested and verified. When the accuracy is higher than 0.9, the verification is terminated to ensure that the system can make appropriate responses in different situations.

[0066] In the present invention, by introducing the adjustment coefficient, the system can adaptively adjust the weight of each risk index according to real-time data, making the system more flexible and responsive, ensuring that the system pays more attention to current high-risk factors, and improving the sensitivity and effectiveness of monitoring.

[0067] S402: Normalize each initial dynamic adjustment weight to determine the dynamic adjustment weight corresponding to the displacement risk index, the tilt risk index, and the vibration risk index:

[0068]

[0069] in, Represents the dynamic adjustment weight of the displacement risk index, represents the dynamic adjustment weight of the tilt risk index, Indicates the dynamically adjusted weight of the vibration hazard index.

[0070] In this invention, normalization ensures that the relative importance of the three risk indices—displacement, tilt, and vibration—is properly reflected in the system, preventing a single risk factor from dominating the entire decision-making process. Furthermore, through dynamic weighting adjustments, the system effectively avoids over-responses to low-risk indicators, thereby reducing false alarms and redundant operations.

[0071] S5: Calculate the comprehensive risk index of the micro-core pile based on the displacement risk index, the tilt risk index and the vibration risk index in combination with the corresponding dynamic adjustment weights.

[0072] In one possible implementation, the calculation formula for the comprehensive risk index is specifically:

[0073]

[0074] Among them, R represents the comprehensive risk index, Represents the normalized dynamic adjustment weight of the displacement risk index, V d represents the displacement risk index, Represents the dynamic adjustment weight of the tilt risk index, V θ represents the tilt risk index, Represents the dynamic adjustment weight of the vibration hazard index, V v Indicates the vibration hazard index.

[0075] In this invention, based on real-time weight adjustments, the comprehensive risk index can more sensitively reflect the main risk factors affecting micro-core piles, thereby issuing accurate early warnings. At the same time, it avoids the misjudgment that may result from relying on only one factor (such as displacement or vibration) to assess risk.

[0076] S6: Determine whether the comprehensive risk index is greater than the preset risk index. If so, wake up the camera on the micro-core pile to capture multiple images of the monitored area and proceed to the next step. Otherwise, continue monitoring.

[0077] In this invention, the system activates the camera to capture images only when the risk exceeds a set threshold, thus reducing unnecessary image acquisition and avoiding wasting resources when there is no risk. Furthermore, the system activates the camera for image acquisition only when needed, avoiding the need to run the camera for extended periods and store large amounts of ineffective data. This approach effectively reduces the waste of storage space and bandwidth, improving resource efficiency.

[0078] S7: Input the image of the area to be monitored into the risk prediction model based on deep convolutional neural network and output the risk prediction result.

[0079] Among them, the Deep Convolutional Neural Network (CNN) is a type of artificial neural network particularly well-suited for processing data with a grid-like structure (such as images and videos). It is an extension of the Convolutional Neural Network (CNN), and its hierarchical structure enables the network to learn complex features in data. It is widely used in fields such as computer vision, image classification, object detection, and natural language processing.

[0080] In a possible implementation, S7 specifically includes:

[0081] S701: Adjust the image size of the area to be monitored to a fixed size through image normalization.

[0082] S702: Use a 1x1 convolution kernel to perform a convolution operation on the image of the area to be monitored in the enhanced data set to extract a first feature map.

[0083] S703: Generate a second feature map using depthwise convolution technology.

[0084] S704: Concatenate the first feature map and the second feature map in series along the channel dimension:

[0085] Y=Concat([Y′,Y′×F dp ])

[0086] Among them, Y represents the concatenated feature map, Concat represents the concatenation operation, Y′ represents the first feature map, and F dp Denotes the depth convolution kernel, Y′×F dp Represents the second feature map obtained by the depthwise convolution operation.

[0087] In the present invention, through feature map splicing (such as splicing the first feature map and the second feature map along the channel dimension), the system can fuse multi-dimensional features from different convolutional layers, and information extracted from different scales and directions is combined together, which helps to capture more complex image information.

[0088] S705: Performing a pooling operation on the images of the area to be monitored after being serially stitched:

[0089] P ij =max(x mn )

[0090] Among them, P ij Represents the output value at position (i, j) after the pooling operation, x mn Represents the element in the mth row and nth column of the pooling window, and max represents maximization.

[0091] S706: Activate the image of the area to be monitored after the pooling operation through the ReLU activation function:

[0092]

[0093] Among them, ReLU represents the activation function and x represents the input feature map.

[0094] S707: Use the DFC attention mechanism to process the activated feature map to generate an attention map.

[0095] In a possible implementation, S707 specifically includes:

[0096] S7071: Using the activated feature map as input, the rows and columns of the feature map are processed separately through the fully connected layer of the deep convolutional neural network:

[0097]

[0098] Among them, a′ hw Represents the eigenvalue after weighting in the row direction, h represents the row of the image, w represents the column of the image, H represents the total number of rows, and W represents the total number of columns. Represents the coefficient of row weighting, ⊙ represents element-by-element multiplication, z h,h′w Represents the characteristic value of the image at a certain position, a hw represents the weighted eigenvalue in the column direction, The coefficient representing the column weighting, a′ hw′ Represents the weighted features of a certain position in the image after weighting in the row direction.

[0099] S7072: Generate an attention map by performing weighted summation of each position in the processed feature map with its corresponding horizontal and vertical neighbors.

[0100] S7073: Apply the generated attention map to the original feature map to capture spatial information:

[0101] Z′=Z⊙A

[0102] Among them, Z′ represents the adjusted feature map, Z represents the original feature map, ⊙ represents element-wise multiplication, and A represents the generated attention map.

[0103] In this paper, by weighting different parts of the image (e.g., by weighting rows and columns), the model can focus on more relevant spatial context and avoid missing critical image details. At the same time, the attention mechanism allows the model to assign different weights to each location in the image. The network pays more attention to important areas (such as structural anomalies, cracks, or other risk indicators), thereby improving the accuracy and robustness of predictions.

[0104] S708: Take the attention map as input and predict the risk category through the Softmax function:

[0105]

[0106] z=W·x+b

[0107] Among them, P(y=1|z) represents the predicted probability that the input sample z belongs to the positive category, σ represents the Sigmoid activation function, z represents the linear output of the fully connected layer, W represents the weight matrix of the fully connected layer, and b represents the bias term.

[0108] In one possible implementation, the cross entropy loss function is used as the loss function of the risk prediction model:

[0109]

[0110]

[0111] Among them, L represents the cross entropy loss function value, N represents the total number of training samples, and y i represents the true label of the i-th sample, y i =1 means the i-th sample belongs to the positive class, y i = 0 means that the i-th sample belongs to the negative class, P(y = 1 | z i ) represents the probability that the i-th sample is predicted to be positive, log represents the natural logarithm, z i Represents the linear output result after the i-th sample passes through the network layer.

[0112] Among them, the cross-entropy loss function (also known as logarithmic loss) is a commonly used loss function in machine learning and deep learning. It is widely used in classification problems, especially in binary and multi-classification problems. Its goal is to measure the difference or uncertainty between the predicted category probability distribution and the true category.

[0113] The risk prediction model is optimized by the improved Adam optimizer until the function value of the cross entropy loss function is less than the preset loss function value.

[0114] The Adam optimizer (Adaptive Moment Estimation) is a gradient-based optimization algorithm widely used in the training of deep learning models, especially in large-scale datasets and high-dimensional parameter spaces. Combining the concepts of momentum and RMSProp, the Adam optimizer effectively adjusts the learning rate of each parameter, accelerating convergence and improving training efficiency.

[0115] Optionally, the improved Adam optimizer specifically includes:

[0116] Initialization parameters, including model parameters θ0, learning rate η, batch size N B , decay rates β1, β2 and constant ε.

[0117] Initialize the first-order moment estimate m0=0, the second-order moment estimate v0=0, and

[0118] Get the training dataset.

[0119] Extract a batch of training samples from the training data set and calculate the current parameter θ t-1 Gradient of the objective function:

[0120]

[0121] Among them, g t represents the gradient at time t, x ik Represents the input data of the kth sample, y ik represents the label of the kth sample, θ t-1 represents the model parameters in the previous iteration, L represents the cross entropy loss function, L((x ik ,y ik ),θ t-1 ) represents the output of the model.

[0122] Update the first moment estimate using the current gradient:

[0123] m t =β1·m t-1 +(1-β1)·g t

[0124] Among them, m t represents the first-order moment estimate at time t, m t-1 represents the first-order moment estimate at time t-1, β1 represents the decay rate of the first-order moment, g t represents the gradient at time t.

[0125] Calculate the maximum value of the first-order moment at the previous moment and the current gradient:

[0126] m max =max(|m t-1 |,|g t |)

[0127] Among them, m max It represents the maximum value of the absolute value of the first-order moment at the previous moment and the current gradient. max represents the maximum value and || represents the absolute value.

[0128] Calculate the target ratio based on the updated first-order moment estimate and the maximum value of the first-order moment and the current gradient at the previous moment:

[0129]

[0130] Wherein, Λ(t) represents the target ratio.

[0131] In this invention, by controlling the update amplitude through the target ratio, model parameters can be adjusted more smoothly, avoiding parameter updates that are too large or too small during training. This makes the optimization process more stable and ensures that the model gradually approaches the optimal solution over multiple iterations. In this way, the Adam optimizer can maintain the stability of parameter updates and improve model accuracy.

[0132] Compute the second-order moment estimate by applying the target ratio to control the magnitude of the update:

[0133] v t =β2·v t-1 +(1-β2)·|g t | Λ(t)

[0134] Among them, v t represents the second-order moment estimate at time t, v t-1 It represents the second-order moment estimate at time t-1, and β2 represents the decay rate of the second-order moment.

[0135] When the target ratio is less than 2, switch to the AMSGrad optimizer for optimization, set amsgrad = True, and update the parameter θ t :

[0136]

[0137] Among them, θ t represents the model parameters at time t, θ t-1 represents the model parameters at time t-1, η represents the learning rate, m t represents the first-order moment estimate at time t, ε represents a constant, represents the modified second-order moment estimate in the AMSGrad optimizer, represents the revised second moment estimate of the previous moment.

[0138] Among them, the AMSGrad optimizer is an improved version of the Adam optimizer, which aims to solve the instability problems that may occur in the Adam optimizer, especially the problem of non-convergence of training caused by excessive learning rate or unstable second-order moment estimation.

[0139] In the present invention, the AMSGrad optimizer ensures that the update does not overly rely on historical gradient information by maintaining the maximum value of the second-order moment, thereby preventing instability and overfitting during training.

[0140] When the target ratio is greater than or equal to 2, continue to use the Adam optimizer to update the parameter θ t :

[0141]

[0142] When the iteration reaches the preset number of iterations, the iteration is stopped and the final optimization parameter θ is output. T .

[0143] In the present invention, by optimizing the loss function (such as cross entropy loss), the improved Adam optimizer can make the model fit the data more accurately and provide more accurate risk prediction.

[0144] It should be noted that the risk prediction results specifically include risky categories whose risk prediction probability is greater than a preset probability value and risk-free categories whose risk prediction probability is less than or equal to the preset probability value.

[0145] In a possible implementation manner, after S7, the method further includes:

[0146] Within a preset period, when the risk prediction results output by the risk prediction model are all in the risk-free category, an alarm will be issued to remind managers to correct the data.

[0147] When the risk prediction result output by the risk prediction model is a risky category, the collected image of the area to be monitored will be sent to the management personnel for confirmation.

[0148] In this invention, when the results of the risk prediction model remain in the "no risk" category for a long time, the system can automatically issue an alarm, reminding managers to check and correct the data. This can effectively prevent the model from producing inaccurate prediction results due to data errors or missing data. At the same time, when the risk prediction results indicate "risky", the system can automatically send the relevant images to managers, avoiding the tedious task of manually reviewing all monitored areas. In this way, managers can prioritize high-risk areas and allocate resources more efficiently.

[0149] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0150] By calculating the comprehensive hazard index of the micro-core pile and analyzing its relationship with a preset hazard index, the system determines whether to wake up the camera. This system only activates the camera when a real hazard occurs, avoiding continuous recording when no disaster is occurring. This effectively reduces energy consumption and extends the camera's operating time. Furthermore, image acquisition and analysis into the risk prediction model are performed only when actually needed, avoiding the generation of invalid video data, reducing storage and bandwidth waste, and improving the system's resource utilization efficiency.

[0151] Reference Manual Figure 2 , showing a structural schematic diagram of a video GNSS micro-core pile intelligent monitoring system provided by the present invention.

[0152] The present invention further provides a video GNSS micro-core pile intelligent monitoring system 20, which is applied to the above-mentioned video GNSS micro-core pile intelligent monitoring method, comprising:

[0153] Processor 201.

[0154] The memory 202 stores computer-readable instructions. When the computer-readable instructions are executed by the processor 201, the video GNSS micro-core pile intelligent monitoring method of the method embodiment is implemented.

[0155] The video GNSS micro-core pile intelligent monitoring system 20 provided by the present invention can execute the above-mentioned video GNSS micro-core pile intelligent monitoring method and achieve the same or similar technical effects. To avoid repetition, the present invention will not go into details.

[0156] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0157] In an embodiment of the present invention, the comprehensive risk index of the micro-core pile is calculated, and the relationship between the comprehensive risk index and the preset risk index is analyzed to determine whether to wake up the camera. The camera is activated only when an actual danger occurs, avoiding continuous filming when there is no disaster, thereby effectively reducing energy consumption and extending the camera's operating time. At the same time, image acquisition and input into the risk prediction model for analysis are only performed when actually needed, avoiding the generation of invalid video data, reducing storage and bandwidth waste, and improving the system's resource utilization efficiency.

[0158] It should be understood that the processor in the embodiments of the present invention may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0159] It should also be understood that the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0160] The above embodiments can be implemented in whole or in part through software, hardware (such as circuits), firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the processes or functions described in accordance with the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired method (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains a collection of one or more available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, or magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0161] It should be understood that the term "and / or" as used herein simply describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A alone, A and B together, or B alone. A and B can be singular or plural. Furthermore, the character " / " as used herein generally indicates an "or" relationship between the associated objects, but it may also indicate an "and / or" relationship. For specific understanding, please refer to the context.

[0162] In this disclosure, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, "at least one of a, b, or c" can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or plural.

[0163] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0164] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0165] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0166] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interface, indirect coupling or communication connection of the device or unit, which can be electrical, mechanical or other forms.

[0167] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0168] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0169] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0170] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the video GNSS micro-core pile intelligent monitoring method as described in the method embodiment is implemented.

[0171] The computer-readable storage medium provided by the present invention can implement the steps and effects of the video GNSS micro-core pile intelligent monitoring method of the above method embodiment. To avoid repetition, the present invention will not go into details.

[0172] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0173] In an embodiment of the present invention, the comprehensive risk index of the micro-core pile is calculated, and the relationship between the comprehensive risk index and the preset risk index is analyzed to determine whether to wake up the camera. The camera is activated only when an actual danger occurs, avoiding continuous filming when there is no disaster, thereby effectively reducing energy consumption and extending the camera's operating time. At the same time, image acquisition and input into the risk prediction model for analysis are only performed when actually needed, avoiding the generation of invalid video data, reducing storage and bandwidth waste, and improving the system's resource utilization efficiency.

[0174] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

[0175] There are a few points to note:

[0176] (1) The drawings of the embodiments of the present invention only relate to the structures related to the embodiments of the present invention. Other structures may refer to conventional designs.

[0177] (2) For the sake of clarity, the thickness of layers or regions in the drawings used to describe the embodiments of the present invention are exaggerated or reduced, that is, these drawings are not drawn to scale. It is understood that when an element such as a layer, film, region, or substrate is referred to as being "on" or "under" another element, the element may be "directly" "on" or "under" the other element or intervening elements may be present.

[0178] (3) In the absence of conflict, the embodiments of the present invention and the features therein may be combined with each other to form new embodiments.

[0179] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A video GNSS micro-core pile intelligent monitoring method, characterized in that: include: S1: collecting real-time position data, real-time inclination angle data and real-time vibration data of the micro-core pile; S2: determining the current displacement, current angle change value and current vibration amplitude of the micro-core pile according to the real-time position data, the real-time tilt angle data and the real-time vibration data; S3: Calculating a displacement risk index, a tilt risk index, and a vibration risk index based on the current displacement, the current angle change value, and the current vibration amplitude of the micro-core pile; S4: Determining dynamic adjustment weights of the displacement risk index, the tilt risk index, and the vibration risk index; S5: Calculating a comprehensive risk index of the micro-core pile according to the displacement risk index, the tilt risk index, and the vibration risk index in combination with corresponding dynamic adjustment weights; S6: Determine whether the comprehensive risk index is greater than a preset risk index; if so, wake up the camera on the micro-core pile to collect multiple images of the area to be monitored, and proceed to the next step; Otherwise, continue monitoring; S7: Input the image of the area to be monitored into a risk prediction model based on a deep convolutional neural network, and output a risk prediction result.

2. The video GNSS micro-core pile intelligent monitoring method according to claim 1 is characterized in that: The current displacement, current tilt angle and current vibration amplitude of the micro-core pile are determined in the following manner: Obtaining position information of the micro-core pile at two adjacent moments, and calculating the displacements at the two moments as the current displacement of the micro-core pile using the Euclidean distance formula; Obtaining the inclination angle data of the micro-core pile at two adjacent moments, and calculating the difference between the inclination angle data at the two adjacent moments by subtraction operation as the current angle change value of the micro-core pile; A MEMS acceleration sensor is used to collect real-time vibration data, and the vibration amplitude of the real-time vibration data is extracted through Fourier transform as the current vibration amplitude of the micro-core pile.

3. The video GNSS micro-core pile intelligent monitoring method according to claim 1 is characterized in that: The S3 specifically includes: Based on the current displacement, current angle change value and current vibration amplitude of the micro-core pile, the displacement risk index, the tilt risk index and the vibration risk index are calculated by the following formula: Among them, V d (t) represents the displacement risk index at time t, V t represents the displacement of the micro-core pile at time t, V safe Represents the safe displacement threshold, V max Indicates the maximum allowable displacement, V θ (t) represents the tilt risk index at time t, θ t represents the angle change of the micro-core pile at time t, θ safe Indicates the safe angle change value, θ max Indicates the maximum allowable value of the angle change, V v (t) represents the vibration hazard index at time t, A t Indicates the vibration amplitude at time t, A safe Indicates the safe vibration amplitude threshold, A max Indicates the maximum allowable value of the vibration amplitude.

4. The video GNSS micro-core pile intelligent monitoring method according to claim 1 is characterized in that: The S4 specifically includes: S401: Calculate the initial dynamic adjustment weights of the displacement risk index, the tilt risk index, and the vibration risk index using the following formula: Among them, W v (t) represents the initial dynamic adjustment weight of the vibration hazard index at time t, W v (t-1) represents the initial dynamic adjustment weight of the vibration risk index at time t-1, λ represents the weight adjustment coefficient of the control vibration risk index, V v (t) represents the vibration hazard index at time t, V v (t-1) represents the vibration hazard index at time t-1, A max Indicates the maximum allowable value of vibration, W d (t) represents the initial dynamic adjustment weight of the displacement risk index at time t, W d (t-1) represents the dynamic adjustment weight of the displacement risk index at time t-1, μ represents the weight adjustment coefficient of the displacement risk index, V d (t) represents the displacement risk index at time t, V d (t-1) represents the displacement risk index at time t-1, V max Indicates the maximum allowable displacement, W θ (t) represents the dynamic adjustment weight of the tilt risk index at time t, W θ (t-1) represents the dynamic adjustment weight of the tilt risk index at time t-1, γ represents the weight adjustment coefficient of the tilt risk index, V θ (t) represents the tilt risk index at time t, V θ (t-1) represents the tilt risk index at time t-1, θ max Indicates the maximum allowable value of tilt; S402: performing normalization processing on each of the initial dynamic adjustment weights to determine the dynamic adjustment weights corresponding to the displacement risk index, the tilt risk index, and the vibration risk index.

5. The video GNSS micro-core pile intelligent monitoring method according to claim 4 is characterized in that: The calculation formula of the comprehensive risk index is specifically: Among them, R represents the comprehensive risk index, Represents the dynamic adjustment weight of the displacement risk index, V d represents the displacement risk index, Represents the dynamic adjustment weight of the tilt risk index, V θ represents the tilt risk index, Represents the dynamic adjustment weight of the vibration hazard index, V v Indicates the vibration hazard index.

6. The video GNSS micro-core pile intelligent monitoring method according to claim 1 is characterized in that: The S7 specifically includes: S701: adjusting the image size of the area to be monitored to a fixed size through image normalization; S702: Using a 1x1 convolution kernel, perform a convolution operation on the image of the area to be monitored in the enhanced data set to extract a first feature map; S703: Generate a second feature map using deep convolution technology; S704: Concatenate the first feature map and the second feature map in series along the channel dimension: Y=Concat([Y′,Y′×F dp ]) Among them, Y represents the concatenated feature map, Concat represents the concatenation operation, Y′ represents the first feature map, and F dp Represents the depth convolution kernel, Y′×F dp Represents the second feature map obtained by the depth convolution operation; S705: performing a pooling operation on the serially stitched images of the area to be monitored; S706: Activate the image of the area to be monitored after the pooling operation through the ReLU activation function; S707: Use the DFC attention mechanism to process the activated feature map and generate an attention map; S708: Using the attention map as input, predict the risk category using the Softmax function: z=W·x+b Among them, P(y=1|z) represents the predicted probability that the input sample z belongs to the positive category, σ represents the Sigmoid activation function, z represents the linear output of the fully connected layer, W represents the weight matrix of the fully connected layer, and b represents the bias term.

7. The video GNSS micro-core pile intelligent monitoring method according to claim 6 is characterized in that: The S707 specifically includes: S7071: Using the activated feature map as input, the rows and columns of the feature map are processed separately through the fully connected layer of the deep convolutional neural network: Among them, a′ hw Represents the eigenvalue after weighting in the row direction, h represents the row of the image, h′ represents the weighting of the convolution kernel on the row, w represents the column of the image, w′ represents the weighted processing of the column, H represents the total number of rows, W represents the total number of columns, Represents the coefficient of row weighting, ⊙ represents element-by-element multiplication, z h,h′w Represents the characteristic value of the image at a certain position, a hw represents the weighted eigenvalue in the column direction, The coefficient representing the column weighting, a′ hw′ Represents the weighted features of a certain position in the image after weighting in the row direction; S7072: Generate an attention map by performing a weighted summation of each position in the processed feature map with its corresponding horizontal and vertical neighbors; S7073: Apply the generated attention map to the original feature map to capture spatial information: Z′=Z⊙A Among them, Z′ represents the adjusted feature map, Z represents the original feature map, ⊙ represents element-wise multiplication, and A represents the generated attention map.

8. The video GNSS micro-core pile intelligent monitoring method according to claim 1 is characterized in that: Using the cross entropy loss function as the loss function of the risk prediction model; The risk prediction model is optimized by an improved Adam optimizer until the function value of the cross entropy loss function is less than a preset loss function value.

9. The video GNSS micro-core pile intelligent monitoring method according to claim 1 is characterized in that: The risk prediction result specifically includes a risk category with a risk prediction probability greater than the preset probability value and a risk-free category with a risk prediction probability less than or equal to the preset probability value; After S7, the method further includes: When, within a preset period, the risk prediction results output by the risk prediction model are all in the risk-free category, an alarm is issued to remind management personnel to perform data correction; When the risk prediction result output by the risk prediction model is the risky category, the collected image of the area to be monitored is sent to the management personnel for confirmation.

10. A video GNSS micro-core pile intelligent monitoring system, characterized in that: include: processor; A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the video GNSS micro-core pile intelligent monitoring method as described in any one of claims 1 to 9 is implemented.

Images