Anchor cable construction dust monitoring system and method based on multi-source data fusion

Abstract

The application relates to the technical field of construction environment management, and discloses an anchor cable construction dust monitoring system and method based on multi-source data fusion, which comprises a collection module, a transmission module, a fusion module, an analysis module, a feedback module and a traceability module; the collection module collects dust particle parameters, construction process related data and environment supporting data to form a three-dimensional original data set; the transmission module determines the priority of dust inducing data and performs differential precision preprocessing; the fusion module forms a unified dimension key data set through feature alignment, redundancy elimination and attention weight fusion; the analysis module analyzes dust concentration levels, diffusion trends and influence ranges; the feedback module dynamically adjusts the early warning threshold and outputs graded early warning; and the traceability module realizes full-process data traceability. The application realizes accurate monitoring, trend prediction and graded early warning of anchor cable construction dust, improves the refinement and intelligent level of construction dust control, and reduces the influence of dust pollution on the surrounding environment.

Description

Technical Field

[0001] This invention relates to the field of construction environment management technology, specifically to a dust monitoring system and method for anchor cable construction based on multi-source data fusion. Background Technology

[0002] Anchor cable construction is a core technology in geotechnical engineering projects such as slope reinforcement and foundation pit support. Through drilling, lowering steel strand anchor cables, grouting for wall reinforcement, and tensioning and locking, the anchor cables are anchored into deep, stable rock or soil layers. This transfers the load of the superstructure to the stable strata, effectively inhibiting deformation and slippage of the rock and soil, and ensuring the stability of the structure. However, the drilling and grouting processes during anchor cable construction generate a large amount of dust particles, which not only worsens the working environment at the construction site and endangers the occupational health of construction workers, but also spreads to the surrounding area, causing air pollution.

[0003] Currently, technologies related to construction dust monitoring have made some progress. For example, Chinese invention patent application number 2025109388059 discloses a construction site dust monitoring system based on big data. This system integrates multi-source data such as meteorological parameters, construction activity records, and image information, solving the problem that traditional monitoring systems only focus on particulate matter concentration and have a single data dimension. However, the aforementioned existing technologies are aimed at overall dust monitoring of general construction sites and have not been adapted and optimized for the technological characteristics of anchor cable construction.

[0004] Anchor cable construction is characterized by dispersed work locations, dynamic movement of dust-generating points during drilling operations, and a strong correlation between dust generation intensity and construction parameters such as drilling speed, drill rod advance speed, and grout flow rate. Existing general-purpose dust monitoring systems have the following technical shortcomings:

[0005] The data collection was not targeted enough, and the construction process parameters directly related to the dust generation during anchor cable construction were not collected simultaneously. It was impossible to establish the correlation between construction processes and dust generation, making it difficult to trace the source of dust and achieve precise control.

[0006] Data processing efficiency and accuracy cannot be balanced. Using a uniform preprocessing and fusion method for all collected data makes it impossible to distinguish the contribution of different data to dust monitoring. This results in the loss of details of high-value core data, while low-value redundant data consumes a lot of computing resources, reducing the real-time performance of the monitoring system.

[0007] The fusion effect of multi-source heterogeneous data is poor. For heterogeneous data with different dimensions and sampling frequencies, such as dust particle data, construction process data, and environmental data, there is a lack of suitable feature alignment and fusion mechanisms, resulting in insufficient data fusion accuracy and inability to accurately predict the diffusion trend and impact range of dust.

[0008] The early warning mechanism has poor adaptability. It uses a fixed threshold for early warning and does not take into account the impact of real-time working conditions such as construction process intensity, ambient wind speed, and humidity on dust diffusion. This can easily lead to problems such as delayed early warning, false alarms, or missed alarms, and it cannot provide timely and effective guidance for on-site dust control.

[0009] Therefore, developing a dust monitoring system tailored to the construction scenarios of anchor cables and adapted to their technological characteristics, and achieving efficient fusion of multi-source data, accurate monitoring and trend prediction of dust conditions, and hierarchical dynamic early warning, has become an urgent technical problem to be solved in this field. Summary of the Invention

[0010] The purpose of this invention is to overcome the above-mentioned defects of the prior art and provide a dust monitoring system and method for anchor cable construction based on multi-source data fusion. It addresses the technical problems of insufficient data acquisition targeting, inability to balance data processing efficiency and accuracy, poor fusion effect of multi-source heterogeneous data, and poor adaptability of early warning mechanism in the prior art, and realizes accurate monitoring, intelligent analysis and dynamic early warning of dust in anchor cable construction throughout the entire process.

[0011] The present invention is achieved through the following technical solution.

[0012] The anchor cable construction dust monitoring system based on multi-source data fusion of the present invention includes a data acquisition module, a transmission module, a fusion module, an analysis module, a feedback module, and a source tracing module;

[0013] The acquisition module is used to collect dust particle parameters, construction process related data and environmental supporting data in the anchor cable construction area to build a three-dimensional raw dataset for dust monitoring.

[0014] The transmission module is communicatively connected to the acquisition module and is used to receive the three-dimensional raw dataset, determine the data priority based on the causes of dust pollution during anchor cable construction and the correlation between the data, perform differentiated precision preprocessing on the data based on the data priority, and forward the processed data to the fusion module according to the priority.

[0015] The fusion module is communicatively connected to the transmission module and is used to receive data forwarded by the transmission module, construct a fusion model to perform feature alignment and redundancy removal operations on the data and complete the fusion to form a unified-dimensional key dataset for dust monitoring.

[0016] The analysis module is communicatively connected to the fusion module and is used to traverse key datasets to analyze dust concentration levels, diffusion trends, and impact ranges.

[0017] The feedback module is communicatively connected to the analysis module and is used to set dynamic early warning thresholds, generate graded early warning signals and locate early warning positions, and synchronously feed back to the preset construction terminal.

[0018] The tracing module is connected to the feedback module and is used to collect the operating status data and data flow trajectory of each module in the system in real time, and generate data index and monitoring history archive.

[0019] Furthermore, the acquisition module includes a dust particle sensor group, a construction process sensor group, and an environmental sensor group;

[0020] The dust particle parameters include one or more of the following: particle size distribution, mass concentration, and number concentration; the construction process related data includes one or more of the following: drilling speed, anchor cable tension pressure, drill rod advance speed, and grout flow rate; the environmental supporting data includes one or more of the following: ambient wind speed, wind direction, air humidity, and atmospheric pressure.

[0021] Each sensor group collects data synchronously at a preset sampling frequency, and the collection time is uniformly marked by a timestamp.

[0022] Furthermore, in the data priority determination stage of the transmission module, the priority coefficient is obtained in the following way: based on three core dust influencing factors—dust particle parameters, construction process related data, and environmental support data—corresponding causal weight coefficients are set for each of the three types of data; the ratio of the real-time measured value of each type of data to the theoretical / rated maximum value of that type of data in the anchor cable construction scenario is taken as the influence ratio of that type of data; the influence ratios of the three types of data are multiplied by the corresponding causal weight coefficients and then summed to finally obtain the priority coefficient of that type of data.

[0023] Specifically, the data prioritization evaluation model is as follows:

[0024] ;

[0025] In the formula: Let i be the priority coefficient for the i-th data type; , , The inducing weight coefficients for dust particle parameters, construction process-related data, and environmental support data; For the i-th type of data, the measured value of dust particle concentration, the quantitative value of construction process intensity, and the quantitative value of environmental impact are: These represent the theoretical maximum value of dust particle concentration, the rated maximum value of construction process intensity, and the limit quantification value of environmental impact under anchor cable construction scenarios.

[0026] Based on the calculated priority coefficient The data is divided into three priorities—high, medium, and low—based on the preset range. Differentiated preprocessing operations are performed on data of different priorities, with higher priority data receiving higher preprocessing precision.

[0027] Furthermore, the feature alignment and redundancy removal of the fusion model includes the following steps:

[0028] Feature alignment operation stage: Calculate the feature similarity of different data sources, and quantify the linear correlation of each feature dimension of the two data sources based on the Pearson correlation coefficient calculation logic;

[0029] Three types of feature alignment operations are performed based on the feature similarity results: when the similarity is higher than the preset high similarity threshold, the two data sources are determined to be related data with the same source features, and are directly aligned by dimension index; when the similarity is in the medium range, cross-dimensional alignment is completed through the feature mapping matrix; when the similarity is lower than the preset low similarity threshold, they are determined to be heterogeneous feature data, and their independent feature dimensions are retained and the feature types are marked to complete the differential alignment.

[0030] Redundancy removal operation stage: Redundancy removal is performed based on feature similarity: When the feature similarity is higher than the preset redundancy judgment threshold, the information integrity of the two types of data sources is compared, the data with higher information integrity is retained, and redundant data is removed;

[0031] The fusion model derives the attention weights of each data source based on feature similarity and the information contribution of each data source. The feature vectors of each data source are multiplied by their corresponding attention weights and then summed to obtain a unified dimensional feature vector after fusion. The attention weights of all data sources are greater than 0, and the sum of all attention weights equals 1. The magnitude of the attention weights is positively correlated with the feature similarity and information contribution of the corresponding data source.

[0032] Specifically, the formula for calculating the feature similarity between different data sources is as follows:

[0033] ;

[0034] In the formula: The feature similarity between the m-th data source and the n-th data source; This represents the total number of feature dimensions. This represents the value of the k-th feature dimension of the m-th data source. Let be the mean of all feature dimensions of the m-th data source; This represents the value of the k-th feature dimension of the n-th data source. The mean of all feature dimensions of the nth type of data source; This is the time decay coefficient; , For the collection timestamps of the m-th and n-th data sources;

[0035] The fusion model is based on feature similarity. The attention weight of each data source is calculated based on its information contribution and the data contribution of each data source. Completed the integration, among which The fused feature vector is a unified dimension. This represents the total number of data sources after alignment and redundancy removal. The attention weights are the feature vectors aligned to the m-th data source. All values ​​are greater than zero and follow the rules. .

[0036] Furthermore, based on feature similarity and the information contribution of each data source, the attention weight stage of each data source is calculated. The information contribution of each data source is determined in the following way: the information contribution is calculated by comprehensively considering the following core indicators, namely: the Pearson correlation coefficient between the data source features and the core indicators of dust monitoring, the ratio of the information completeness of the data source itself to the maximum information completeness of all data sources, the ratio of the precision level quantification value of the data source to the maximum precision level of all data sources, and the average feature similarity of the data source with all other data sources.

[0037] Correlation weight coefficients are set for the above indicators respectively. The first three indicators are positively correlated with information contribution, while the fourth indicator, average feature similarity, is negatively correlated with information contribution. Finally, the information contribution of the data source is calculated by weighting. The higher the information contribution, the greater the attention weight assigned to the corresponding data source.

[0038] Specifically, the information contribution of each data source is determined by the following formula:

[0039] ;

[0040] In the formula: Contribution to information; This refers to the relevance weighting coefficient. The Pearson correlation coefficient between the characteristics of the m-th type of data source and the core indicators of dust monitoring; The information completeness of the m-th type of data source; This represents the precision level quantization value for the m-th type of data source; This represents the maximum value of information completeness across all data sources and the maximum value of precision level quantization across all data sources. This is the result of averaging the feature similarities between the m-th data source and all other data sources.

[0041] Furthermore, the analysis logic for dust concentration level, diffusion trend, and impact range in the analysis module is as follows:

[0042] Concentration level: The effective value of dust particle mass concentration in the fused key dataset is used as the basis for judgment. It is compared with several preset concentration classification thresholds to divide the dust concentration into several levels, which correspond to different levels of pollution.

[0043] Diffusion Trend: Based on the convection-diffusion partial differential equation, the predicted dust concentration at a certain time and spatial coordinate is calculated. During the calculation, three core effects are comprehensively considered: the molecular diffusion effect brought about by the dust particle diffusion coefficient; the dust convective transport effect brought about by the environmental wind speed vector; and the concentration replenishment effect brought about by the dust source strength at the corresponding spatial location. Finally, the change law of dust concentration with time and space is accurately described, and the dust diffusion trend is predicted.

[0044] Specifically:

[0045] Dust dispersion trend: ,in, This represents the predicted dust concentration at spatial coordinates (x, y, z) at time t. This represents the diffusion coefficient of dust particles. Represents the Laplace operator. Represents the ambient wind speed vector. Represents the concentration gradient vector. This represents the dust source intensity at spatial coordinates (x, y, z) at time t;

[0046] Scope of impact: ,in, This is a quantitative value representing the area affected by dust pollution. To predict the length of the time window; The spatial extent of the monitoring area; This is the baseline value for dust concentration; This is an indicator function that takes the value 1 when the condition inside the parentheses is true, and 0 otherwise.

[0047] Furthermore, this scheme sets 3-6 concentration levels, with 4 levels being the preferred setting.

[0048] Furthermore, the feedback module constructs a dynamic adjustment model for the early warning threshold based on historical dust data, construction process data, and environmental data, generating several levels of dynamic early warning thresholds, with each level of early warning threshold corresponding to the output of an early warning signal.

[0049] Specifically, the early warning threshold dynamic adjustment model:

[0050] ;

[0051] In the formula: Let t be the threshold for the j-th level early warning; This serves as the baseline value for the early warning threshold. , , The threshold influence coefficients for construction process intensity, ambient wind speed, and air humidity; The values ​​of construction process intensity, ambient wind speed, and air humidity at time t are real-time values. The historical average intensity of construction procedures, historical average wind speed, and historical average air humidity are all included.

[0052] The tiered early warning signals include Level 1, Level 2, and Level 3 early warnings, respectively corresponding to... Three levels of warning thresholds, and .

[0053] Furthermore, the traceability module generates a data index containing a unique identifier, categorizes and stores the entire process data archive, monitors the operating status of each module in real time, records abnormal information, and generates an abnormal directory.

[0054] Specifically, the data index generated by the traceability module includes a unique data identifier, a data source module identifier, a collection timestamp, a data type identifier, and a data processing status identifier;

[0055] The monitoring history archive includes raw data archive, preprocessed data archive, fused data archive, analysis result archive, and early warning feedback archive, and each archive is stored in chronological order.

[0056] The traceability module monitors the operating status parameters of each module in real time. When the operating status parameters exceed the preset normal range, it automatically records the abnormal information and generates an abnormal information directory simultaneously.

[0057] Furthermore, the acquisition module is interconnected with a transmission module via a wireless network, the transmission module is interconnected with a fusion module via a wireless network, the fusion module is interconnected with an analysis module via a wireless network, the analysis module is interconnected with a feedback module via a wireless network, and the feedback module is interconnected with a tracing module via a wireless network.

[0058] Optionally, the wireless network may employ any one or more combinations of LoRa, WiFi, 4G, or 5G.

[0059] Alternatively, in addition to the wireless connection method described above, the modules can also be connected via a wired connection.

[0060] The method for monitoring dust pollution during anchor cable construction based on multi-source data fusion includes the following steps:

[0061] S1: Collect dust particle parameters, construction process related data and environmental supporting data in the anchor cable construction area to build a three-dimensional raw dataset for dust monitoring;

[0062] S2: Construct a data priority evaluation model based on the weight of dust pollution causes, calculate the priority coefficient of various types of data, perform differentiated precision preprocessing on the data based on the priority coefficient, and forward the data synchronously according to priority;

[0063] S3: Calculate the feature similarity of different data sources, complete feature alignment through direct alignment, cross-dimensional mapping or differential labeling, remove redundant data, calculate the fused data based on attention weights, and form a unified-dimensional key dataset for dust monitoring.

[0064] S4: Classify dust concentration levels according to preset thresholds, analyze dust diffusion trends, quantify the impact range of dust within a preset time window, and output analysis results;

[0065] S5: Based on historical data and real-time parameters, a dynamic adjustment model for early warning thresholds is constructed, three levels of dynamic early warning thresholds are set, graded early warning signals are generated and the early warning location is located, and the data is synchronously fed back to the preset construction terminal.

[0066] S6: Generates a data index with a unique identifier, categorizes and stores raw data, preprocessed data, and fused data, monitors the running status of each module in real time, records abnormal information, and generates an abnormal directory.

[0067] Furthermore, the method is implemented based on the aforementioned anchor cable construction dust monitoring system based on multi-source data fusion.

[0068] The beneficial effects of this invention are:

[0069] This invention addresses the technological characteristics of anchor cable construction by constructing a three-dimensional raw dataset that includes dust particle parameters, construction process-related data, and environmental support data. It simultaneously collects all-dimensional data directly related to dust generation and diffusion, establishing a direct correlation between construction processes and dust generation intensity. This provides a complete data foundation for tracing the source of dust, precise monitoring, and control, and solves the problem of insufficient targeting in existing data collection technologies.

[0070] By constructing a data priority evaluation model, data priority is determined based on the weight of dust pollution causes and data correlation. Differentiated precision preprocessing is performed on data with different priorities. While ensuring the processing accuracy of high-value core data, the computational resource consumption of low-value redundant data is reduced. This balances the accuracy of data processing with the real-time performance of system operation, and solves the problem that existing technologies cannot balance data processing efficiency and accuracy.

[0071] This invention constructs a fusion model adapted to multi-source heterogeneous data. It calculates feature similarity based on an improved Pearson correlation coefficient with a time decay coefficient and adopts differentiated feature alignment methods for data sources with different similarity, effectively solving the alignment problem of heterogeneous data with different dimensions and sampling frequencies. At the same time, it completes data fusion by allocating attention weights based on information contribution, improving the accuracy of multi-source data fusion. It can accurately analyze the concentration level, diffusion trend and impact range of dust, providing accurate data support for dust control.

[0072] This invention constructs a dynamic adjustment model for early warning thresholds, which dynamically adjusts the early warning thresholds based on real-time working conditions such as construction process intensity, ambient wind speed, and air humidity. This adapts to the dust monitoring and early warning needs of different construction stages and environmental conditions, effectively reducing the probability of false alarms and missed alarms. At the same time, through graded early warning and early warning location positioning, it can promptly guide the construction site to take targeted dust control measures, improving the response speed and effectiveness of dust control. Attached Figure Description

[0073] To more clearly illustrate the technical solutions in the embodiments of the invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0074] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0075] Figure 1 This is a schematic diagram of the composition of an anchor cable construction dust monitoring system based on multi-source data fusion.

[0076] Figure 2 This is a flowchart illustrating a method for monitoring dust pollution during anchor cable construction based on multi-source data fusion. Detailed Implementation

[0077] The following is combined with Figures 1-2 The present invention will be described in detail below.

[0078] Example 1:

[0079] This embodiment presents a dust monitoring system for anchor cable construction based on multi-source data fusion, such as... Figure 1 As shown, it includes:

[0080] The data acquisition module is used to collect dust particle parameters, construction process related data, and environmental data in the anchor cable construction area to build a three-dimensional raw dataset for dust monitoring.

[0081] The data acquisition module includes a dust particle sensor group, a construction process sensor group, and an environmental sensor group;

[0082] Dust particle parameters include particle size distribution, mass concentration, and number concentration; construction process related data include drilling speed, anchor cable tension pressure, drill rod advance speed, and grout flow rate; environmental data include ambient wind speed, wind direction, air humidity, and atmospheric pressure; each sensor group collects data synchronously at a preset sampling frequency, and the collection time is uniformly marked by a timestamp.

[0083] The dust particle sensor group adopts a distributed deployment method, with at least three dust particle sensors evenly arranged in a ring around the anchor cable drilling point, and the deployment height of each sensor is consistent with the center height of the borehole opening; the construction process sensor group is integrated into the key execution components of the anchor cable construction equipment, with the drilling speed sensor deployed in the drill rod drive unit, the anchor cable tension pressure sensor deployed in the hydraulic circuit of the tension jack, the drill rod advance speed sensor deployed in the guide rail of the advance mechanism, and the shotcrete flow sensor deployed at the outlet end of the shotcrete pipe; the environmental sensor group is deployed at preset reference positions in the upwind, downwind, and crosswind directions of the construction area, and the deployment height of the environmental sensors is higher than the preset proportion of the maximum dust diffusion height of the construction area;

[0084] The transmission module is used to receive the three-dimensional raw dataset, determine the data priority based on the causes of dust pollution during anchor cable construction and the correlation between the data, perform differentiated precision preprocessing on the data based on the data priority, and forward the processed data to the fusion module according to the priority.

[0085] During the data priority determination phase of the transmission module, a data priority evaluation model is constructed:

[0086] ;

[0087] In the formula: The priority coefficient for the i-th data type; , , The inducing weight coefficients for dust particle parameters, construction process-related data, and environmental support data; For the i-th type of data, the measured value of dust particle concentration, the quantitative value of construction process intensity, and the quantitative value of environmental impact are: These represent the theoretical maximum value of dust particle concentration, the rated maximum value of construction process intensity, and the limit quantification value of environmental impact under anchor cable construction scenarios.

[0088] The above formula combines the different degrees of influence of dust particle parameters, construction process related data, and environmental support data on dust generation and diffusion, and sets weight coefficients and combines the measured values ​​of various data with the corresponding maximum values ​​to achieve quantitative determination of data priority, thereby highlighting the dominant role of the core influencing factors of dust, while taking into account the auxiliary influence of construction processes and environmental conditions.

[0089] Transmission module according to Within the preset range, the data undergoes three levels of differentiated preprocessing: high, medium, and low precision, so that data with higher priority is processed with higher precision.

[0090] in, , , The sum is 1, and ∈[0.4, 0.6], the larger the value is when the dust generated during anchor cable construction is mainly affected by the characteristics of the particles themselves, such as particle concentration and particle size distribution, the smaller the value is when the generation and diffusion of dust are more dominated by the construction process or environmental conditions; ∈[0.2, 0.4], the larger the value is when the intensity of construction operations such as drilling speed and tension pressure is strongly positively correlated with the amount of dust generated, and the smaller the value is when the optimization of construction technology or adjustment of procedures has little impact on dust; ∈[0.1, 0.3], the larger the value is when environmental factors such as wind speed and humidity have a significant impact on the dust diffusion range and settling velocity, the smaller the value is when environmental conditions are stable and have little interference with dust propagation;

[0091] when When the data is in a high-priority range, a refined noise reduction preprocessing based on wavelet transform is used to retain high-frequency features while removing random noise; when... When the data is in the medium priority range, a sliding window mean filter preprocessing is used to balance the data noise reduction effect and processing efficiency; when... When the data is in a low priority range, threshold filtering is used for preprocessing to quickly remove abnormal data that exceeds the reasonable range.

[0092] The fusion module is used to receive data forwarded by the transmission module, build a fusion model to perform feature alignment and redundancy removal operations on the data and complete the fusion to form a unified-dimensional key dataset for dust monitoring.

[0093] In the feature alignment stage, the feature similarity between different data sources is calculated:

[0094] ;

[0095] In the formula: The feature similarity between the m-th data source and the n-th data source; This represents the total number of feature dimensions. This represents the value of the k-th feature dimension of the m-th data source. Let be the mean of all feature dimensions of the m-th data source; This represents the value of the k-th feature dimension of the n-th data source. The mean of all feature dimensions of the nth type of data source; This is the time decay coefficient; , For the collection timestamps of the m-th and n-th data sources;

[0096] The above formula fully considers the feature differences and temporal correlations of different data sources. It calculates the similarity of feature dimensions through the core logic of Pearson correlation coefficient, and introduces a time decay coefficient to weaken the interference of data sources at different collection times. It not only accurately captures the inherent correlation of data from the same source, but also effectively distinguishes the unique attributes of heterogeneous data, providing support for subsequent feature alignment and redundancy removal, and ensuring accurate and efficient fusion of multi-source data.

[0097] based on Perform feature alignment: when When the similarity exceeds a preset high similarity threshold, the two data sources are determined to be related data with the same features, and the feature dimensions of the two data sources are directly aligned according to their corresponding dimension indices; when... When the data is within a preset medium similarity range, the features from the low-dimensional data source are mapped to the feature space of the high-dimensional data source using a feature mapping matrix to achieve cross-dimensional feature alignment; when... When the similarity is less than the preset low similarity threshold, it is determined to be heterogeneous feature data. Each feature dimension is retained and the feature type is marked to complete the differential alignment.

[0098] Redundancy removal is performed based on the aligned feature similarity results: when When the data exceeds the preset redundancy threshold, calculate the information integrity score of the two data sources, retain the data from the data source with higher information integrity, and remove redundant data.

[0099] Fusion models based on feature similarity The attention weight of each data source is calculated based on its information contribution and the data contribution of each data source. Completed the integration, among which The fused feature vector is a unified dimension. This represents the total number of data sources after alignment and redundancy removal. The attention weights are the feature vectors aligned to the m-th data source. All values ​​are greater than zero and follow the rules. ;

[0100] in, ∈[0.01, 0.5], the larger the ambient wind speed in the construction area and the shorter the data collection time interval, the smaller the value; the smaller the ambient wind speed and the longer the data collection time interval, the larger the value.

[0101] Based on feature similarity In the stage of calculating the attention weight of each data source in relation to its information contribution, the information contribution of each data source is determined by the following formula:

[0102] ;

[0103] In the formula: Contribution to information; This refers to the relevance weighting coefficient. The Pearson correlation coefficient between the characteristics of the m-th type of data source and the core indicators of dust monitoring; The information completeness of the m-th type of data source; This represents the precision level quantization value for the m-th type of data source; This represents the maximum value of information completeness across all data sources and the maximum value of precision level quantization across all data sources. This is the result of averaging the feature similarities between the m-th data source and all other data sources;

[0104] The above formula integrates the correlation between the data source and the core indicators of dust monitoring, the completeness of its own information and the accuracy level. It balances the effects of different influencing factors through the correlation weight coefficient, and combines the average feature similarity between the data source and all other data sources to quantify the information contribution. This provides support for the allocation of attention weights in the fusion model, enabling high-value data sources to play a greater role in the fusion process.

[0105] The analysis module is used to traverse key datasets to analyze dust concentration levels, diffusion trends, and impact range.

[0106] The analysis logic for dust concentration levels, diffusion trends, and impact range in the analysis module is as follows:

[0107] Dust concentration level: Where L represents the dust concentration level, with levels 1-4 corresponding to light, moderate, heavy, and severe pollution, respectively, and C represents the effective value of dust particle mass concentration in the fused key dataset. These are the threshold values ​​for different concentration levels, and the threshold values ​​are dynamically calibrated based on environmental standards for anchor cable construction scenarios.

[0108] Dust dispersion trend: ,in, This represents the predicted dust concentration at spatial coordinates (x, y, z) at time t. This represents the diffusion coefficient of dust particles. Represents the Laplace operator. Represents the ambient wind speed vector. Represents the concentration gradient vector. Let represent the dust source intensity at spatial coordinates (x, y, z) at time t, where The value >0 increases with decreasing dust particle size, decreasing air humidity, and increasing environmental turbulence intensity, and decreases with increasing dust particle size, increasing air humidity, and decreasing environmental turbulence intensity. ≥0, which increases with the increase of drilling speed, grout flow rate and drill rod advance speed of the anchor cable construction corresponding to the spatial location, and decreases with the decrease of drilling speed, grout flow rate, drill rod advance speed or no construction activity at the location;

[0109] The above formula combines the physical characteristics of dust particle diffusion, introduces key parameters such as diffusion coefficient and ambient wind speed vector, considers the relationship between dust source strength and construction procedure parameters, and constructs a diffusion model through partial differential equations. It can accurately describe the variation law of dust concentration at different time and space locations. At the same time, it can dynamically adjust the diffusion coefficient according to factors such as particle size and air humidity, so that the diffusion trend prediction is more in line with the actual environmental conditions.

[0110] Dust impact range: ,in, This is a quantitative value representing the area affected by dust pollution. To predict the length of the time window; The spatial extent of the monitoring area; This is the baseline value for dust concentration; This is an indicator function that takes the value 1 when the condition inside the parentheses is true, and 0 otherwise.

[0111] The above formula takes the area where the dust concentration exceeds the benchmark value within the prediction time window as the statistical object. It calculates the quantitative value of the impact range through multi-dimensional integration, which not only clarifies the spatial boundary of the dust impact, but also takes into account the continuous impact in the time dimension. Furthermore, it uses an indicator function to accurately screen the effective statistical area, making the quantitative result of the impact range more valuable for reference and providing support for the synchronous management of construction dust.

[0112] The feedback module is used to set dynamic early warning thresholds, generate graded early warning signals and locate early warning locations, and synchronously feed back to the preset construction terminals;

[0113] When setting dynamic warning thresholds in the feedback module, the following rules apply:

[0114] Based on historical dust concentration data, construction process intensity data, and environmental data from key dust monitoring datasets, a dynamic adjustment model for early warning thresholds is constructed:

[0115] ;

[0116] In the formula: Let t be the threshold for the j-th level early warning; This serves as the baseline value for the early warning threshold. , , The threshold influence coefficients for construction process intensity, ambient wind speed, and air humidity; The values ​​of construction process intensity, ambient wind speed, and air humidity at time t are real-time values. The historical average values ​​of construction process intensity, historical average environmental wind speed, and historical average air humidity are used.

[0117] The above formula determines the baseline value of the warning threshold based on historical data. It combines the ratio of the real-time values ​​of construction process intensity, ambient wind speed, and air humidity to the historical average value. The threshold influence coefficient is used to adjust the degree of influence of different factors on the warning threshold, thereby achieving dynamic optimization of the warning threshold, ensuring the timeliness of the warning, and flexibly adjusting the warning standard according to the actual working conditions on site, thus reducing the probability of false alarms and missed alarms.

[0118] The tiered early warning signals include Level 1, Level 2, and Level 3 warnings, which correspond to... Three levels of warning thresholds, and ;

[0119] in, , , All values ​​are greater than zero; the stronger the inducing effect of the construction process on dust generation, the better. The larger the value, the weaker the inducing effect. The smaller the value, the stronger the effect of ambient wind speed on dust dispersion. The larger the value, the weaker the boosting effect. The smaller the value, the stronger the effect of air humidity on suppressing dust particles. The larger the value, the weaker the inhibitory effect. The smaller the value;

[0120] The traceability module is used to collect real-time data on the operating status of each module in the system and the data flow trajectory, and to generate data indexes and monitoring history archives.

[0121] The data index generated by the traceability module includes a unique data identifier, a data source module identifier, a collection timestamp, a data type identifier, and a data processing status identifier.

[0122] The monitoring history archive includes raw data archive, preprocessed data archive, fused data archive, analysis results archive, and early warning feedback archive, and each archive is stored in chronological order.

[0123] The traceability module monitors the operating status parameters of each module in real time. When the operating status parameters exceed the preset normal range, it automatically records the abnormal information and generates an abnormal information directory simultaneously.

[0124] In this embodiment, the acquisition module is interconnected with the transmission module via a wireless network, the transmission module is interconnected with the fusion module via a wireless network, the fusion module is interconnected with the analysis module via a wireless network, the analysis module is interconnected with the feedback module via a wireless network, and the feedback module is interconnected with the tracing module via a wireless network.

[0125] In other embodiments, the modules may also be connected by wires.

[0126] In this embodiment, the acquisition module collects dust particle parameters, construction process-related data, and environmental data from the anchor cable construction area to build a three-dimensional raw dataset for dust monitoring. The transmission module receives the three-dimensional raw dataset, determines data priority based on the dust causes of anchor cable construction and data correlation, performs differentiated precision preprocessing on the data based on the data priority, and forwards the processed data to the fusion module according to priority. The fusion module then receives the data forwarded by the transmission module, constructs a fusion model, performs feature alignment and redundancy removal operations on the data, and completes the fusion to form a unified-dimensional key dataset for dust monitoring. The analysis module further traverses the key dataset to analyze dust concentration levels, diffusion trends, and impact ranges. The feedback module sets dynamic early warning thresholds, generates graded early warning signals, locates early warning positions, and synchronously feeds them back to the preset construction terminal. Finally, the tracing module collects the operating status data and data flow trajectory of each module in the system in real time, generating a data index and monitoring history archive.

[0127] In the above embodiments, the system is applied to anchor cable construction scenarios. It can accurately collect and efficiently process and integrate various types of construction-related data, monitor dust concentration levels, diffusion trends, and impact ranges in real time, and provide timely feedback on risk locations through dynamic early warnings to help quickly control pollution. At the same time, it can trace and monitor the entire process data and system operation status to ensure that the monitoring is compliant and reliable, effectively reduce the impact of dust on the environment and personnel, and improve the efficiency and precision of environmental protection management during construction.

[0128] Example 2:

[0129] At the implementation level, based on Example 1, this example refers to... Figure 2 The anchor cable construction dust monitoring system based on multi-source data fusion in Example 1 will be further described in detail below:

[0130] A method for monitoring dust pollution during anchor cable construction based on multi-source data fusion includes the following steps:

[0131] Collect dust particle parameters, construction procedure-related data, and environmental data in the anchor cable construction area to build a three-dimensional raw dataset for dust monitoring.

[0132] A data priority evaluation model is constructed based on the weight of dust pollution causes. Priority coefficients for various types of data are calculated. Based on the priority coefficients, the data is preprocessed with differentiated precision and forwarded synchronously according to priority.

[0133] Calculate the feature similarity of different data sources, complete feature alignment through direct alignment, cross-dimensional mapping or differential labeling, remove redundant data, calculate the fused data based on attention weights, and form a unified-dimensional key dataset for dust monitoring.

[0134] The dust concentration levels are classified according to preset thresholds, the dust diffusion trend is analyzed, the influence range of dust within a preset time window is quantified, and the analysis results are output.

[0135] A dynamic adjustment model for early warning thresholds is constructed based on historical data and real-time parameters. Three levels of dynamic early warning thresholds are set, graded early warning signals are generated, and the early warning location is located. The data is then synchronously fed back to the preset construction terminal.

[0136] Generate a data index with a unique identifier, classify and store raw data, preprocessed data, and fused data, monitor the running status of each module in real time, record abnormal information and generate an abnormal directory.

[0137] In summary, the system and method described in the above embodiments, during execution, construct a dust monitoring dataset by collecting construction-related parameters and environmental data. Simultaneously, differentiated processing is implemented through priority determination to effectively eliminate redundancy and noise, improving data quality. Feature alignment and fusion technologies are used to form unified-dimensional key data, accurately analyzing dust concentration levels, diffusion trends, and impact ranges. Furthermore, based on historical and real-time data, early warning thresholds are dynamically adjusted to generate tiered early warning signals and locate warning positions, ensuring environmental compliance during construction. Simultaneously, the system fully records data flow trajectories and the entire monitoring process, enabling data traceability and real-time capture and feedback of anomalies. This provides precise data support for construction dust control, effectively improving the scientific nature and environmental protection level of construction environmental management and reducing the impact of dust pollution on the surrounding environment.

[0138] The above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand and implement the present invention. They should not be construed as limiting the scope of protection of the present invention. All equivalent changes or modifications made in accordance with the spirit and essence of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A dust monitoring system for anchor cable construction based on multi-source data fusion, characterized in that: It includes a data acquisition module, a transmission module, a fusion module, an analysis module, and a feedback module; The acquisition module is used to collect dust particle parameters, construction process related data and environmental data in the anchor cable construction area to build a raw dust monitoring dataset. The transmission module is communicatively connected to the acquisition module and is used to receive the raw dataset, determine the data priority based on the causes of dust pollution during anchor cable construction and the correlation between the data, perform differentiated precision preprocessing on the data based on the data priority, and forward the processed data to the fusion module according to the priority. The fusion module is communicatively connected to the transmission module and is used to receive data forwarded by the transmission module, construct a fusion model to perform feature alignment and redundancy removal operations on the data and complete the fusion to form a unified-dimensional key dataset for dust monitoring. The analysis module is communicatively connected to the fusion module and is used to traverse key datasets to analyze dust concentration levels, diffusion trends, and impact ranges. The feedback module is communicatively connected to the analysis module and is used to set dynamic early warning thresholds, generate graded early warning signals, locate early warning positions, and synchronously feed back to the preset construction terminal.

2. The anchor cable construction dust monitoring system based on multi-source data fusion according to claim 1, characterized in that: The anchor cable construction dust monitoring system based on multi-source data fusion also includes a source tracing module. The source tracing module is communicatively connected to the feedback module and is used to collect the operating status data and data flow trajectory of each module in the system in real time, and generate data indexes and monitoring history archives.

3. The anchor cable construction dust monitoring system based on multi-source data fusion according to claim 1, characterized in that: The data acquisition module includes a dust particle sensor group, a construction process sensor group, and an environmental sensor group. The dust particle parameters include one or more of the following: particle size distribution, mass concentration, and number concentration; the construction process related data includes one or more of the following: drilling speed, anchor cable tension pressure, drill rod advance speed, and grout flow rate; the environmental supporting data includes one or more of the following: ambient wind speed, wind direction, air humidity, and atmospheric pressure. Each sensor group collects data synchronously at a preset sampling frequency, and the collection time is uniformly marked by a timestamp.

4. The anchor cable construction dust monitoring system based on multi-source data fusion according to claim 1, characterized in that: In the data priority determination stage of the transmission module, the priority coefficient is obtained in the following way: based on three core dust influencing factors—dust particle parameters, construction process related data, and environmental support data—corresponding causal weight coefficients are set for each of the three types of data; the ratio of the real-time measured value of each type of data to the theoretical / rated maximum value of that type of data in the anchor cable construction scenario is taken as the influence ratio of that type of data; the influence ratios of the three types of data are multiplied by the corresponding causal weight coefficients and then summed to finally obtain the priority coefficient of that type of data. Based on the preset range of the calculated priority coefficients, the data is divided into several priorities from low to high. Differentiated preprocessing operations with different precision are performed on the data with different priorities. The higher the priority of the data, the higher the preprocessing precision is used.

5. The anchor cable construction dust monitoring system based on multi-source data fusion according to any one of claims 1-4, characterized in that: Feature alignment and redundancy removal in the fusion model includes the following steps: Feature alignment stage: Calculate the feature similarity between different data sources; Three types of feature alignment operations are performed based on the feature similarity results: when the similarity is higher than the preset high similarity threshold, the two data sources are determined to be related data with the same source features, and are directly aligned by dimension index; when the similarity is in the medium range, cross-dimensional alignment is completed through the feature mapping matrix; when the similarity is lower than the preset low similarity threshold, they are determined to be heterogeneous feature data, and their independent feature dimensions are retained and the feature types are marked to complete the differential alignment. Redundancy removal operation stage: Redundancy removal is performed based on feature similarity: When the feature similarity is higher than the preset redundancy judgment threshold, the information integrity of the two types of data sources is compared, the data with higher information integrity is retained, and redundant data is removed; The fusion model derives the attention weights of each data source based on feature similarity and the information contribution of each data source. The feature vectors of each data source are multiplied by their corresponding attention weights and then summed to obtain a unified dimensional feature vector after fusion.

6. The anchor cable construction dust monitoring system based on multi-source data fusion according to claim 5, characterized in that: Based on feature similarity and the information contribution of each data source, the attention weight stage of each data source is calculated. The information contribution of each data source is determined in the following way: the information contribution is calculated by comprehensively considering the following core indicators, namely: the Pearson correlation coefficient between the data source features and the core indicators of dust monitoring, the ratio of the information completeness of the data source itself to the maximum information completeness of all data sources, the ratio of the precision level quantification value of the data source to the maximum precision level of all data sources, and the average feature similarity of the data source with all other data sources. Correlation weight coefficients are set for the above indicators respectively. The first three indicators are positively correlated with information contribution, while the fourth indicator, average feature similarity, is negatively correlated with information contribution. Finally, the information contribution of the data source is calculated by weighting. The higher the information contribution, the greater the attention weight assigned to the corresponding data source.

7. The anchor cable construction dust monitoring system based on multi-source data fusion according to claim 5, characterized in that: The analysis logic for dust concentration level, diffusion trend, and impact range in the analysis module is as follows: Concentration level: The effective value of dust particle mass concentration in the fused key dataset is used as the criterion. It is compared with several preset concentration classification thresholds to divide the dust concentration into several levels, which correspond to different levels of pollution. Diffusion Trend: Based on the convection-diffusion partial differential equation, the dust diffusion trend is predicted. Scope of impact: ,in, This is a quantitative value representing the area affected by dust pollution. To predict the length of the time window; The spatial extent of the monitoring area; This is the baseline value for dust concentration; This is an indicator function that takes the value 1 when the condition inside the parentheses is true, and 0 otherwise.

8. The anchor cable construction dust monitoring system based on multi-source data fusion according to claim 1 or 7, characterized in that: The feedback module constructs a dynamic adjustment model for the early warning threshold based on historical dust data, construction process data, and environmental data, generating several levels of dynamic early warning thresholds. Each level of early warning threshold corresponds to the output of an early warning signal.

9. A method for monitoring dust pollution during anchor cable construction based on multi-source data fusion, characterized in that: Includes the following steps: Collect dust particle parameters, construction procedure-related data, and environmental data in the anchor cable construction area to build a raw dust monitoring dataset. A data priority evaluation model is constructed based on the weight of dust pollution causes. Priority coefficients for various types of data are calculated. Based on the priority coefficients, the data is preprocessed with differentiated precision and forwarded synchronously according to priority. Calculate the feature similarity of different data sources, complete feature alignment through direct alignment, cross-dimensional mapping or differential labeling, remove redundant data, calculate the fused data based on attention weights, and form a unified-dimensional key dataset for dust monitoring. The dust concentration levels are classified according to preset thresholds, the dust diffusion trend is analyzed, the influence range of dust within a preset time window is quantified, and the analysis results are output. A dynamic adjustment model for early warning thresholds is constructed based on historical data and real-time parameters. Three levels of dynamic early warning thresholds are set, graded early warning signals are generated, and the early warning location is located. The data is then synchronously fed back to the preset construction terminal. Generate a data index with a unique identifier, classify and store raw data, preprocessed data, and fused data, monitor the running status of each module in real time, record abnormal information and generate an abnormal directory.

10. The method for monitoring dust during anchor cable construction based on multi-source data fusion according to claim 9, characterized in that: The method is implemented based on the anchor cable construction dust monitoring system based on multi-source data fusion as described in any one of claims 1-8.

Images