Foundation pit collapse prevention intelligent support data analysis system and method

Abstract

The invention relates to the technical field of foundation pit engineering, and discloses a foundation pit anti-collapse intelligent support data analysis system and method.The method comprises the steps that multi-source data are collected through a sensor on a foundation pit construction site, normalization processing is conducted on cleaned data, and preprocessed data are obtained; fusing the preprocessed data by adopting a wavelet transform fusion algorithm to obtain fused data, and performing feature extraction on the fused data to obtain a feature vector reflecting the safety state of the foundation pit; based on the characteristic vector, an IWO algorithm is adopted to optimize a support vector machine to establish a foundation pit safety evaluation model, data collected in real time is input into the foundation pit safety evaluation model, a current safety state evaluation value of the foundation pit is calculated, and a foundation pit safety state evaluation result and an early warning level are output; according to the foundation pit safety state evaluation result and the early warning level, a self-adaptive Kalman filtering algorithm is adopted to adjust control parameters of the intelligent supporting system; according to the invention, intelligent management of collapse prevention of the foundation pit is realized.

Description

Technical Field

[0001] This invention relates to the field of foundation pit engineering technology, specifically to a data analysis system and method for intelligent support to prevent foundation pit collapse. Background Technology

[0002] With the rapid development of urban construction, deep foundation pit projects are becoming increasingly common. Foundation pit collapse accidents not only cause huge economic losses but also threaten the lives of construction workers and the stability of surrounding buildings and infrastructure. Currently, safety monitoring of foundation pit support mainly relies on manual inspections and traditional sensor monitoring. Manual inspections suffer from low efficiency, strong subjectivity, and poor real-time performance. Although traditional sensor monitoring can acquire some data, this data is often isolated and lacks in-depth analysis and effective integration, making it difficult to accurately predict the safety status of the foundation pit and take timely and effective support measures to prevent collapse accidents. At the same time, existing data analysis methods suffer from low calculation accuracy and poor adaptability when dealing with complex and ever-changing foundation pit environmental data, and cannot meet the needs of intelligent support for foundation pit collapse prevention. Summary of the Invention

[0003] The purpose of this invention is to solve the above-mentioned problems by designing an intelligent support data analysis system and method for preventing foundation pit collapse.

[0004] The first aspect of this invention provides a data analysis method for intelligent support for foundation pit collapse prevention, the method comprising the following steps:

[0005] Multi-source data is collected by sensors at the foundation pit construction site. The collected data is cleaned to remove outliers and noise. The cleaned data is then normalized to obtain pre-processed data.

[0006] The preprocessed data was fused using a wavelet transform fusion algorithm to obtain fused data. Then, the feature vector reflecting the safety status of the foundation pit was obtained by using principal component analysis to extract features from the fused data.

[0007] Based on feature vectors, the IWO algorithm is used to optimize the support vector machine to establish a foundation pit safety assessment model. Real-time collected data, after preprocessing and fusion, is input into the foundation pit safety assessment model to calculate the current safety status assessment value of the foundation pit and output the foundation pit safety status assessment result and warning level.

[0008] Based on the safety status assessment results and early warning levels of the foundation pit, the control parameters of the intelligent support system are adjusted using an adaptive Kalman filter algorithm.

[0009] Optionally, in a first implementation of the first aspect of the present invention, the step of fusing the preprocessed data using a wavelet transform fusion algorithm to obtain fused data includes:

[0010] The preprocessed data is decomposed into low-frequency approximation coefficients and high-frequency detail coefficients at different scales using Haar wavelets;

[0011] The low-frequency approximation coefficients of the preprocessed data are weighted and summed, and the low-frequency approximation coefficient with the highest energy is selected.

[0012] The high-frequency detail coefficients of the preprocessed data are fused using the absolute value maximization method, and the weights are dynamically adjusted.

[0013] The fused low-frequency approximation coefficients and high-frequency detail coefficients are combined to form a coefficient matrix, which is then reconstructed into fused data using inverse wavelet transform.

[0014] Optionally, in a second implementation of the first aspect of the present invention, the step of using principal component analysis to extract features from the fused data to obtain a feature vector reflecting the safety status of the foundation pit includes:

[0015] The covariance matrix is ​​calculated from the fused data, and the covariance matrix is ​​decomposed into eigenvalues ​​and corresponding eigenvectors.

[0016] Arrange the eigenvalues ​​in descending order, select the first k principal components, and combine the eigenvectors corresponding to the selected principal components to construct an eigenvector reflecting the safety status of the foundation pit.

[0017] Optionally, in a third implementation of the first aspect of the present invention, the step of establishing a foundation pit safety assessment model based on feature vectors using the IWO algorithm to optimize the support vector machine includes:

[0018] The feature vectors extracted by principal component analysis are divided into training and test sets. The feature vectors are normalized, and the RBF function is selected as the kernel function of the support vector machine.

[0019] Initialize the initial parameters of the IWO algorithm, including population size, maximum number of iterations, and elitist reverse learning ratio;

[0020] An initial weed population is randomly generated, with each weed individual representing a set of SVM parameter combinations. An initial fitness value is assigned to each weed, and the performance of the parameter combinations is evaluated through SVM cross-validation on the training set.

[0021] For each individual weed, the SVM parameter combination is cross-validated on the training set, and the evaluation metric is used as the fitness value.

[0022] In each iteration, weed individuals are sorted according to their fitness values, and boundary treatment and mutation operations are performed on newly generated seeds.

[0023] Based on the ratio of elite reverse learning, the weed individuals with the highest fitness are selected. After each iteration, the parent weeds and offspring seeds are merged into a temporary population. The temporary population is sorted according to the fitness value, the individual with the highest fitness is retained, and other individuals are eliminated to form a new weed population.

[0024] When the number of iterations reaches the preset maximum number of iterations, the weed individual with the highest fitness is determined, and the optimal parameters of the support vector machine are obtained to construct the foundation pit safety assessment model.

[0025] Optionally, in a fourth implementation of the first aspect of the present invention, the step of selecting the weed individual with the highest fitness based on the elite reverse learning ratio includes:

[0026] The parameters of the selected weed individuals are subjected to reverse learning to generate corresponding reverse solutions. The fitness value of the reverse solutions is calculated. If the fitness of a reverse solution is better than that of the original weed individual, the original weed individual is replaced by the reverse solution.

[0027] Optionally, in the fifth implementation of the first aspect of the present invention, the step of inputting the real-time collected and pre-processed and fused data into the foundation pit safety assessment model, calculating the current safety status assessment value of the foundation pit, and outputting the foundation pit safety status assessment result and early warning level includes:

[0028] The data collected in real time, after preprocessing and fusion, is input into the foundation pit safety assessment model to calculate the distance between the input features and the decision boundary of the safety status.

[0029] Based on the output of the support vector machine, the confidence weighting method is used to convert the classification results into a quantitative security status assessment value.

[0030] The safety status assessment value is compared with the preset multi-level thresholds to determine the current warning level of the foundation pit, and the foundation pit safety status assessment result and warning level are output. The multi-level thresholds include safety threshold, attention threshold, warning threshold and danger threshold.

[0031] Optionally, in a sixth implementation of the first aspect of the present invention, adjusting the control parameters of the intelligent support system using an adaptive Kalman filter algorithm based on the pit safety status assessment results and early warning levels includes:

[0032] Obtain the current safety status assessment value and corresponding early warning level of the foundation pit from the foundation pit safety assessment model to determine the current risk status of the foundation pit;

[0033] Initialize the parameters of the adaptive Kalman filter algorithm, construct the state space model, predict the prior state estimate at the current time based on the state space model and the state estimate at the previous time step, and calculate the prior estimate of the state estimation error covariance matrix according to the covariance propagation.

[0034] The relevant data collected during the actual operation of the intelligent support system are used as observation data. Based on the prior estimate of the state estimation error covariance matrix and the observation noise covariance matrix, the Kalman gain is calculated.

[0035] The state at the current moment is updated based on the calculated Kalman gain, and the adjusted control parameters of the intelligent support system are obtained.

[0036] The control parameters of the intelligent support system, adjusted by the adaptive Kalman filter algorithm, are output and transmitted to the actuator of the intelligent support system.

[0037] A second aspect of the present invention provides a data analysis system for intelligent support of foundation pits to prevent collapse, the system comprising:

[0038] The data processing module is used to collect multi-source data through sensors at the foundation pit construction site, clean the collected data to remove outliers and noise, and normalize the cleaned data to obtain pre-processed data.

[0039] The feature extraction module is used to fuse the preprocessed data using a wavelet transform fusion algorithm to obtain fused data, and then to extract features from the fused data using principal component analysis to obtain feature vectors reflecting the safety status of the foundation pit.

[0040] The safety assessment module is used to establish a foundation pit safety assessment model based on feature vectors and the IWO algorithm to optimize the support vector machine. It inputs real-time collected, preprocessed and fused data into the foundation pit safety assessment model, calculates the current safety status assessment value of the foundation pit, and outputs the foundation pit safety status assessment result and warning level.

[0041] The parameter adjustment module is used to adjust the control parameters of the intelligent support system based on the foundation pit safety status assessment results and early warning levels using an adaptive Kalman filter algorithm.

[0042] A third aspect of the present invention provides a data analysis device for intelligent support of foundation pits to prevent collapse. The data analysis device includes a memory and at least one processor. The memory stores instructions. The at least one processor calls the instructions in the memory to cause the data analysis device to perform the various steps of the intelligent support data analysis method for foundation pits to prevent collapse as described in any of the preceding claims.

[0043] A fourth aspect of the present invention provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the steps of the intelligent support data analysis method for foundation pit collapse prevention as described in any of the preceding claims.

[0044] In the technical solution provided by this invention, multi-source data is collected through sensors at the foundation pit construction site. The collected data is cleaned to remove outliers and noise, and then normalized to obtain preprocessed data. A wavelet transform fusion algorithm is used to fuse the preprocessed data, resulting in fused data. Principal component analysis is then used to extract features from the fused data, yielding feature vectors reflecting the safety status of the foundation pit. Based on these feature vectors, an IWO algorithm is used to optimize a support vector machine and establish a foundation pit safety assessment model. The real-time collected, preprocessed, and fused data is input into the foundation pit safety assessment model to calculate the current safety status assessment value of the foundation pit, outputting the foundation pit safety status assessment result and warning level. Based on the foundation pit safety status assessment result and warning level, [further details are needed]. The adaptive Kalman filter algorithm is used to adjust the control parameters of the intelligent support system. This invention fuses multi-source sensor data, fully considering the measurement accuracy and historical data stability of each sensor, which can more accurately reflect the actual state of the foundation pit and improve the reliability and effectiveness of the data. A foundation pit safety assessment model is established, which improves the classification accuracy and generalization ability of the model and can more accurately assess the safety status of the foundation pit. The adaptive Kalman filter algorithm is used for intelligent support strategy adjustment, which can automatically adjust the filter gain according to the dynamic changes of the foundation pit environment, realize the precise adjustment of the control parameters of the intelligent support system, enable the support system to better adapt to the changes in the foundation pit, improve the safety of the foundation pit, realize intelligent management of foundation pit collapse prevention, and effectively reduce the probability of foundation pit collapse accidents. Attached Figure Description

[0045] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention.

[0046] Figure 1 A flowchart of the intelligent support data analysis method for preventing foundation pit collapse provided in an embodiment of the present invention;

[0047] Figure 2 This is a schematic diagram of the structure of the intelligent support data analysis system for foundation pit collapse prevention provided in an embodiment of the present invention;

[0048] Figure 3 A schematic diagram of the structure of the intelligent support data analysis device for foundation pit collapse prevention provided in an embodiment of the present invention. Detailed Implementation

[0049] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” or “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, apparatus, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0050] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 A flowchart of the intelligent support data analysis method for foundation pit collapse prevention provided in this embodiment of the invention is shown. The method specifically includes the following steps:

[0051] Step 101: Collect multi-source data through sensors at the foundation pit construction site. Clean the collected data to remove outliers and noise. Normalize the cleaned data to obtain pre-processed data.

[0052] In this embodiment, various types of sensors are deployed at the foundation pit construction site to form a comprehensive monitoring network. These sensors include, but are not limited to, displacement sensors, stress sensors, groundwater level sensors, tilt sensors, and pore water pressure sensors. Displacement sensors monitor the displacement changes of the foundation pit slope and support structure, stress sensors sense the stress state of the support structure in real time, groundwater level sensors monitor the impact of groundwater level changes on the stability of the foundation pit, tilt sensors detect the inclination of the foundation pit slope, and pore water pressure sensors measure the pore water pressure in the soil. All sensors collect data synchronously at preset time intervals to ensure that dynamic changes during the foundation pit construction process can be captured. At the same time, to ensure the integrity and reliability of the data, the sensors are regularly calibrated and maintained to avoid data loss or anomalies caused by equipment failure.

[0053] The collected raw data may contain outliers and noise, which can affect the accuracy of subsequent analysis. First, the data is initially screened to identify and remove obvious erroneous data caused by sensor malfunctions or transmission errors, such as negative displacement or stress values ​​exceeding the measurement range. Then, statistical analysis methods are used to detect outliers in the data, marking data points that deviate from the mean by more than three standard deviations as outliers. For these outliers, appropriate processing methods are selected based on their distribution: if the number of outliers is small and they exist in isolation, they are deleted directly; if outliers show a clustered distribution and are related to a special construction stage of the foundation pit, such as stress mutations during blasting operations, they are retained and individually labeled. In addition, the continuity of the data is checked. For short-term data gaps caused by communication interruptions, linear interpolation is used to fill in the gaps; for long-term data gaps, the data for that period is discarded.

[0054] Noise data mainly originates from environmental interference such as electromagnetic interference, mechanical vibration, and measurement errors of the sensor itself, which can cause irregular fluctuations in the data. To remove noise, the moving average filtering method is first applied to smooth the data. Random fluctuations are eliminated by calculating the average value of the data within a certain time window. Then, the wavelet transform method is used to further decompose the different frequency components of the data, separating high-frequency noise components from low-frequency effective signals. Noise components are filtered out by setting appropriate thresholds. During the processing, the filtering parameters are adjusted according to the characteristics of different sensor data, such as the slow change characteristics of displacement data and the abrupt change characteristics of stress data, to ensure that the true characteristics of the data are preserved while effectively removing noise.

[0055] After cleaning and denoising, the data collected by different sensors may have different dimensions and numerical ranges, which can affect the training effect of subsequent models and the accuracy of feature extraction. Therefore, it is necessary to normalize the cleaned data to map all data to the interval [0,1]. In practice, the maximum-minimum normalization method is used. For the data series of each sensor, the maximum and minimum values ​​are calculated. The minimum value is subtracted from each data point and then divided by the difference between the maximum and minimum values ​​to obtain the normalized data. This process not only eliminates the influence of differences in dimensions but also makes the data distribution more uniform, which is beneficial to subsequent feature extraction and model training. In addition, to avoid information loss during the normalization process, the statistical characteristics of the original data, such as the mean and variance, are retained so that inverse normalization can be performed when needed.

[0056] Step 102: The preprocessed data is fused using a wavelet transform fusion algorithm to obtain fused data. Principal component analysis is then used to extract features from the fused data to obtain feature vectors reflecting the safety status of the foundation pit.

[0057] In this embodiment, Haar wavelet is used to decompose the preprocessed data into low-frequency approximation coefficients and high-frequency detail coefficients at different scales; the low-frequency approximation coefficients of the preprocessed data are weighted and summed, and the low-frequency approximation coefficient with the largest energy is selected; the high-frequency detail coefficients of the preprocessed data are fused using the absolute value maximization method, and the weights are dynamically adjusted; the fused low-frequency approximation coefficients and high-frequency detail coefficients are combined to form a coefficient matrix, and the coefficient matrix is ​​reconstructed into the fused data through inverse wavelet transform.

[0058] In this embodiment, the covariance matrix is ​​calculated on the fused data, and the covariance matrix is ​​decomposed to obtain a set of eigenvalues ​​and corresponding eigenvectors. The eigenvalues ​​are arranged in descending order, the first k principal components are selected, and the eigenvectors corresponding to the selected principal components are combined to construct a eigenvector reflecting the safety status of the foundation pit.

[0059] In this embodiment, the fused data contains multi-dimensional information from multiple sensors. Different dimensions of data may exhibit varying degrees of correlation. The purpose of calculating the covariance matrix is ​​to quantify the linear relationship between these dimensions. For each pair of data dimensions, their covariance values ​​are calculated to reflect the degree of correlation in their changing trends: a positive covariance value indicates that their changing trends are roughly the same; a negative covariance value indicates opposite trends; and a covariance value close to 0 indicates a weak linear relationship. The covariance values ​​calculated pairwise for all dimensions are arranged in dimensional order to form a square matrix, which is the covariance matrix. This matrix comprehensively displays the interrelationships between the variables in each dimension of the fused data, providing an important basis for subsequent analysis.

[0060] The covariance matrix contains the inherent structure and variation patterns of data. By performing eigenvalue decomposition, this hidden information can be revealed. Eigenvalue decomposition breaks down the covariance matrix into a set of eigenvalues ​​and corresponding eigenvectors. Specifically, it seeks vectors (i.e., eigenvectors) that satisfy specific conditions, such that multiplying the covariance matrix by this vector equals multiplying the vector by a scalar (i.e., an eigenvalue). Each eigenvector represents a direction of data variation, while the eigenvalue represents the degree of dispersion of the data in that direction, i.e., the magnitude of the data variation. Through eigenvalue decomposition, the originally complex covariance matrix is ​​transformed into a simple combination of eigenvalues ​​and eigenvectors, facilitating further analysis and processing.

[0061] After completing the eigenvalue decomposition of the covariance matrix, a set of eigenvalues ​​and eigenvectors is obtained. Each eigenvalue corresponds to a principal component, and its magnitude reflects the explanatory power of the principal component for the original data. To extract the most critical information, all eigenvalues ​​are arranged in descending order. Principal components corresponding to larger eigenvalues ​​contain more important information from the original data and can more significantly reflect the data's changing trends and internal structure; while principal components corresponding to smaller eigenvalues ​​carry relatively less information. Based on actual needs and data characteristics, the first k principal components are selected. The value of k can be determined based on the proportion of original data information retained, such as retaining more than 90% of the information, or based on experience and experiments, ensuring that while effectively reducing data dimensionality, the key information related to the safety status of the foundation pit is retained to the greatest extent.

[0062] After determining the first k principal components, the feature vectors corresponding to these k principal components are extracted and combined in descending order of feature values. These feature vectors form a new vector group, which is the feature vector reflecting the safety status of the foundation pit. This feature vector integrates and refines the key information that best reflects the safety status of the foundation pit in the original fused data, removes redundant and secondary information, and presents the data in a more concise and representative form.

[0063] Step 103: Based on feature vectors, the IWO algorithm is used to optimize the support vector machine to establish a foundation pit safety assessment model. The real-time collected data, after preprocessing and fusion, is input into the foundation pit safety assessment model to calculate the current safety status assessment value of the foundation pit and output the foundation pit safety status assessment result and warning level.

[0064] In this embodiment, the feature vectors extracted through principal component analysis are divided into training and testing sets. The feature vectors are normalized, and the RBF function is selected as the kernel function of the support vector machine. Initial parameters of the IWO algorithm are initialized, including population size, maximum number of iterations, and elite back-learning ratio. An initial weed population is randomly generated, with each weed individual representing a set of SVM parameter combinations. An initial fitness value is assigned to each weed, and the performance of the parameter combinations is evaluated through cross-validation of the SVM on the training set. Cross-validation is then performed on the training set for the SVM parameter combinations corresponding to each weed individual, using the evaluation... The fitness value is estimated as an index. In each iteration, weed individuals are ranked according to their fitness values. Boundary processing and mutation operations are performed on newly generated seeds. Based on the elitist back-learning ratio, the weed individual with the highest fitness is selected. After each iteration, the parent weeds and offspring seeds are merged into a temporary population. The temporary population is ranked according to the fitness value. The individual with the highest fitness is retained, and other individuals are eliminated to form a new weed population. When the number of iterations reaches the preset maximum number of iterations, the weed individual with the highest fitness is determined, and the optimal parameters of the support vector machine are obtained to construct the foundation pit safety assessment model.

[0065] In this embodiment, the parameters of the selected weed individuals are subjected to reverse learning to generate corresponding reverse solutions. The fitness value of the reverse solutions is calculated. If the fitness of a reverse solution is better than that of the original weed individual, the original weed individual is replaced by the reverse solution.

[0066] In this embodiment, real-time collected data, after preprocessing and fusion, is input into the foundation pit safety assessment model to calculate the distance between the input features and the decision boundary of the safety status. Based on the output of the support vector machine, a confidence weighting method is used to convert the classification results into a quantified safety status assessment value. The safety status assessment value is compared with a preset multi-level threshold to determine the current warning level of the foundation pit, and the foundation pit safety status assessment result and warning level are output. The multi-level threshold includes a safety threshold, a attention threshold, a warning threshold, and a danger threshold.

[0067] Step 104: Based on the safety status assessment results and early warning level of the foundation pit, the control parameters of the intelligent support system are adjusted using an adaptive Kalman filter algorithm.

[0068] In this embodiment, the safety status assessment value and corresponding early warning level of the current foundation pit are obtained from the foundation pit safety assessment model to determine the current risk status of the foundation pit; the parameters of the adaptive Kalman filter algorithm are initialized, a state space model is constructed, and the prior estimate of the current state is predicted based on the state space model and the state estimate value of the previous moment, and the prior estimate of the state estimation error covariance matrix is ​​calculated according to the covariance propagation; relevant data collected during the actual operation of the intelligent support system are used as observation data, and the Kalman gain is calculated based on the prior estimate of the state estimation error covariance matrix and the observation noise covariance matrix; the current state is updated based on the calculated Kalman gain to obtain the adjusted intelligent support system control parameters; the intelligent support system control parameters adjusted by the adaptive Kalman filter algorithm are output and transmitted to the actuator of the intelligent support system.

[0069] Please see Figure 2 A schematic diagram of the structure of the intelligent support data analysis system for foundation pit collapse prevention provided in this embodiment of the invention. The system includes:

[0070] The data processing module is used to collect multi-source data through sensors at the foundation pit construction site, clean the collected data to remove outliers and noise, and normalize the cleaned data to obtain pre-processed data.

[0071] The feature extraction module is used to fuse the preprocessed data using a wavelet transform fusion algorithm to obtain fused data, and then to extract features from the fused data using principal component analysis to obtain feature vectors reflecting the safety status of the foundation pit.

[0072] The safety assessment module is used to establish a foundation pit safety assessment model based on feature vectors and the IWO algorithm to optimize the support vector machine. It inputs real-time collected, preprocessed and fused data into the foundation pit safety assessment model, calculates the current safety status assessment value of the foundation pit, and outputs the foundation pit safety status assessment result and warning level.

[0073] The parameter adjustment module is used to adjust the control parameters of the intelligent support system based on the foundation pit safety status assessment results and early warning levels using an adaptive Kalman filter algorithm.

[0074] Figure 3This is a structural schematic diagram of an intelligent support data analysis device for foundation pit collapse prevention provided in an embodiment of the present invention. The intelligent support data analysis device 300 for foundation pit collapse prevention can vary significantly due to different configurations or performance. It may include one or more central processing units (CPUs) 310 (e.g., one or more processors) and a memory 320, and one or more storage media 330 (e.g., one or more mass storage devices) for storing application programs 333 or data 332. The memory 320 and storage media 330 can be temporary or persistent storage. The program stored in the storage media 330 may include one or more modules (not shown in the diagram), each module may include a series of instruction operations on the intelligent support data analysis device 300 for foundation pit collapse prevention. Furthermore, the processor 310 may be configured to communicate with the storage media 330 and execute the series of instruction operations in the storage media 330 on the intelligent support data analysis device 300 for foundation pit collapse prevention to implement the method provided in the above embodiment.

[0075] The intelligent support data analysis device 300 for foundation pit collapse prevention may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input / output interfaces 360, and / or one or more operating devices 331, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 3 The structure of the intelligent support data analysis device for foundation pit collapse prevention shown does not constitute a limitation on the computer equipment provided by the present invention. It may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0076] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the various steps of the intelligent support data analysis method for foundation pit collapse prevention provided in the above embodiments.

[0077] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described equipment or apparatus / unit can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0078] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0079] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for analyzing data of an intelligent supporting system for preventing collapse of a foundation pit, characterized in that, The method comprises the following steps: Collecting multi-source data through sensors at the foundation pit construction site, cleaning the collected data to remove abnormal values and noise data, normalizing the cleaned data to obtain preprocessed data; Fusing the preprocessed data using a wavelet transform fusion algorithm to obtain fused data, and extracting features from the fused data using a principal component analysis method to obtain a feature vector reflecting the safety state of the foundation pit; Based on the feature vector, an IWO algorithm is used to optimize a support vector machine to establish a foundation pit safety evaluation model, real-time collected and preprocessed and fused data are input into the foundation pit safety evaluation model, the current safety state evaluation value of the foundation pit is calculated, and the foundation pit safety state evaluation result and warning level are output; According to the foundation pit safety state evaluation result and the warning level, an adaptive Kalman filter algorithm is used to adjust the control parameters of the intelligent support system.

2. The data analysis method for the anti-collapse intelligent support of foundation pits according to claim 1, characterized in that, The wavelet transform fusion algorithm is used to fuse the preprocessed data to obtain fused data, which comprises: The preprocessed data is decomposed into low-frequency approximation coefficients and high-frequency detail coefficients of different scales using Haar wavelet; The low-frequency approximation coefficients of the preprocessed data are summed by weighting, and the low-frequency approximation coefficient with the maximum energy is selected; The high-frequency detail coefficients of the preprocessed data are fused using the maximum absolute value method, and the weights are dynamically adjusted; The fused low-frequency approximation coefficients and high-frequency detail coefficients are combined to form a coefficient matrix, and the coefficient matrix is reconstructed into fused data through inverse wavelet transform. 3.The data analysis method for the anti-collapse intelligent support of foundation pits according to claim 1, wherein, The principal component analysis method is used to extract features from the fused data to obtain a feature vector reflecting the safety state of the foundation pit, which comprises: The covariance matrix of the fused data is calculated, and the covariance matrix is decomposed to obtain a set of eigenvalues and corresponding eigenvectors; The eigenvalues are arranged in descending order, the first k principal components are selected, and the selected principal components are combined to construct a feature vector reflecting the safety state of the foundation pit.

4. The data analysis method for the anti-collapse intelligent support of foundation pits according to claim 1, characterized in that, The IWO algorithm is used to optimize the support vector machine based on the feature vector to establish the foundation pit safety evaluation model, which comprises: The feature vectors extracted by the principal component analysis are divided into a training set and a test set, the feature vectors are normalized, and the RBF function is selected as the kernel function of the support vector machine; The initial parameters of the IWO algorithm are initialized, including the population size, the maximum number of iterations, and the elite reverse learning ratio; Randomly generate an initial grass population, each grass individual represents a combination of SVM parameters, assign an initial fitness value to each grass, and evaluate the performance of the parameter combination through cross-validation of SVM on the training set; Cross-validation is performed on the SVM parameter combination corresponding to each grass individual on the training set, and the evaluation index is used as the fitness value; In each iteration, the grass individuals are sorted according to the fitness values, and boundary processing and mutation operations are performed on the newly generated seeds. The weed individual with the highest fitness is selected based on an elite reverse learning ratio, parent weeds and child seeds are combined into a temporary population at the end of each iteration, the temporary population is sorted according to fitness values, the individual with the highest fitness is retained, and other individuals are eliminated to form a new weed population; When the number of iterations reaches a preset maximum number of iterations, the weed individual with the highest fitness is determined, and optimal parameters of the support vector machine are obtained to construct a foundation pit safety evaluation model.

5. The data analysis method for the anti-collapse intelligent support of foundation pits according to claim 4, characterized in that, The weed individual with the highest fitness is selected based on an elite reverse learning ratio, parent weeds and child seeds are combined into a temporary population at the end of each iteration, the temporary population is sorted according to fitness values, the individual with the highest fitness is retained, and other individuals are eliminated to form a new weed population; The parameters of the selected weed individual are subjected to a reverse learning operation to generate a corresponding reverse solution, the fitness value of the reverse solution is calculated, and if the fitness of the reverse solution is better than that of the original weed individual, the reverse solution replaces the original weed individual.

6. The data analysis method for the anti-collapse intelligent support of foundation pits according to claim 1, characterized in that, The data collected in real time and subjected to preprocessing and fusion is input into the foundation pit safety evaluation model, the current safety state evaluation value of the foundation pit is calculated, and the foundation pit safety state evaluation result and the warning level are output, including: The data collected in real time and subjected to preprocessing and fusion is input into the foundation pit safety evaluation model, the current safety state evaluation value of the foundation pit is calculated, and the foundation pit safety state evaluation result and the warning level are output, including: The classification result is converted into a quantitative safety state evaluation value by using a confidence weighting method according to the output result of the support vector machine; The safety state evaluation value is compared with a preset multi-level threshold value to determine the warning level of the current foundation pit, and the foundation pit safety state evaluation result and the warning level are output, wherein the multi-level threshold value includes a safety threshold value, an attention threshold value, a warning threshold value and a danger threshold value.

7. The data analysis method for the anti-collapse intelligent support of foundation pits according to claim 1, characterized in that, The control parameters of the intelligent support system are adjusted by using an adaptive Kalman filtering algorithm according to the foundation pit safety state evaluation result and the warning level, including: The safety state evaluation value and the corresponding warning level of the current foundation pit are obtained from the foundation pit safety evaluation model to determine the risk condition of the current foundation pit; The adaptive Kalman filtering algorithm parameters are initialized, a state space model is constructed, the state priori estimation value at the current time is predicted based on the state space model and the state estimation value at the last time, and the priori estimation of the state estimation error covariance matrix is calculated according to covariance propagation; The relevant data collected in the actual operation process of the intelligent support system are taken as observation data, the priori estimation of the state estimation error covariance matrix and the observation noise covariance matrix are obtained, and the Kalman gain is calculated; The state at the current time is updated based on the calculated Kalman gain to obtain the adjusted control parameters of the intelligent support system; The control parameters of the intelligent support system adjusted by the adaptive Kalman filtering algorithm are output and transmitted to the actuator of the intelligent support system.

8. A data analysis system for intelligent support of foundation pit collapse prevention, characterized in that, The system comprises: A data processing module is configured to collect multi-source data through sensors at a foundation pit construction site, clean the collected data to remove abnormal values and noise data, normalize the cleaned data to obtain preprocessed data, and perform fusion on the preprocessed data by using a wavelet transform fusion algorithm to obtain fused data. A feature extraction module is configured to extract features from the fused data by using a principal component analysis method to obtain a feature vector reflecting the safety state of the foundation pit. The security evaluation module is configured to establish a foundation pit security evaluation model based on a feature vector and an IWO algorithm optimization support vector machine, input data collected in real time and preprocessed and fused into the foundation pit security evaluation model, calculate a current security state evaluation value of the foundation pit, and output a foundation pit security state evaluation result and a warning level. The parameter adjustment module is configured to adjust control parameters of the intelligent support system according to the foundation pit security state evaluation result and the warning level and an adaptive Kalman filtering algorithm.

9. A foundation pit anti-collapse intelligent support data analysis device, characterized in that, The foundation pit anti-collapse intelligent support data analysis device comprises a memory and at least one processor, and the memory stores instructions; the at least one processor invokes the instructions in the memory, so that the foundation pit anti-collapse intelligent support data analysis device executes the steps of the foundation pit anti-collapse intelligent support data analysis method according to any one of claims 1-7.

10. A computer-readable storage medium having stored thereon instructions, the instructions comprising, The instructions are executed by the processor to implement the steps of the foundation pit anti-collapse intelligent support data analysis method according to any one of claims 1-7.

Images