Construction Site Dust Reduction and Pollution Prevention and Control Method and System Based on Multimodal Sensing

Abstract

The present invention discloses a method and system for dust reduction and pollution prevention and control in a construction site based on multimodal perception, which relates to the technical field of construction management and includes: real-time collecting PM data at the boundary through boundary monitoring devices preset at the construction site boundary, and at the same time using an unmanned aerial vehicle equipped with monitoring devices to fly and monitor the interior of the construction site according to a preset route to collect PM data in different areas of the site; using an edge computing module to preliminarily analyze the multimodal perception data in real time to identify abnormal dust conditions; transmitting to a data processing center to analyze the current pollution scenario of the construction site, matching the best dust reduction strategy for the current pollution scenario, controlling the surrounding dust reduction devices to perform dust reduction operations, constructing a self-organizing communication network among the dust reduction devices, and using the dust reduction feedback information to adjust the dust reduction operation plan. The present invention integrates the real-time data collected by the boundary monitoring devices and the unmanned aerial vehicle, combines the environmental parameters and the construction working conditions, and realizes the precise regulation of the dust reduction devices.

Description

Technical Field

[0001] The present invention relates to the technical fields of environmental protection and construction management. More specifically, it relates to a construction site dust reduction and pollution prevention and control method and system based on multi-modal perception. Background Art

[0002] In the current construction environment, dust pollution has become a highly concerned issue. The traditional dust reduction methods at construction sites are relatively single and ineffective. Relying solely on occasional manual watering cannot perform precise dust reduction operations in real time according to changes in dust concentration. With the increasing requirements for the construction environment, it is urgent to create a dust-free site. Existing factory boundary spraying systems and sprinkler truck spraying systems, due to the lack of effective monitoring and intelligent control, have situations of excessive spraying resulting in waste of water resources, or insufficient spraying leading to unsatisfactory dust reduction effects. At the same time, within the construction site boundary and inside, there is a lack of comprehensive and real-time dust monitoring means, making it difficult to accurately grasp the dust concentration, thus unable to promptly initiate dust reduction measures, easily leading to secondary pollution and affecting the surrounding environment and the health of construction workers.

[0003] In recent years, with the development of the Internet of Things (IoT) and unmanned aerial vehicle (UAV) technology, multi-modal environmental monitoring means have gradually been applied to construction pollution control. For example, by deploying devices such as particulate matter (PM2.5, PM10) sensors and weather stations at the site boundary, the dust concentration and diffusion trend can be monitored in real time; while UAVs equipped with environmental monitoring modules can dynamically collect data at high altitudes and blind spots, making up for the deficiencies of fixed monitoring points. Therefore, how to construct a management system that can comprehensively utilize the data collected by boundary monitoring devices and UAVs equipped with monitoring devices, analyze the data, and automatically control dust reduction devices to achieve energy-saving and efficient prevention of secondary pollution is an urgent problem to be solved currently. Summary of the Invention

[0004] To solve the above technical problems, the present invention proposes a construction site dust reduction and pollution prevention and control method and system based on multi-modal perception. By integrating the real-time data collected by boundary monitoring devices and UAVs, combining environmental parameters and construction working conditions, a dynamic pollution diffusion model is constructed, and an adaptive algorithm is used to achieve precise regulation of dust reduction devices, realizing efficient dust reduction and prevention of secondary pollution.

[0005] The first aspect of the present invention provides a construction site dust reduction and pollution prevention and control method based on multi-modal perception, including the following steps:

[0006] Through boundary monitoring devices preset at the construction site boundary, PM data at the boundary is collected in real time. At the same time, a UAV equipped with monitoring devices is used to conduct flight monitoring of the interior of the construction site according to a preset flight path, and PM data of different areas of the site is collected;

[0007] Combine the monitored PM data with the environmental-related data obtained by the temperature and humidity sensors to generate multi-modal perception data. Use the edge computing module to preliminarily analyze the multi-modal perception data in real time and identify dust anomalies;

[0008] Mark the multi-modal perception data according to the dust anomalies, and give priority to quickly transmitting it to the data processing center. Construct a dynamic pollution diffusion model to analyze the current pollution scenario of the construction site, and match the best dust reduction strategy for the current pollution scenario;

[0009] Generate a dust reduction equipment control signal based on the best dust reduction strategy, control the surrounding dust reduction equipment to perform dust reduction operations, construct a self-organizing communication network among the dust reduction equipment, and use the dust reduction feedback information of the dust reduction equipment to adjust the dust reduction operation plan.

[0010] In this solution, the PM data of the construction site is collected as follows:

[0011] Deploy boundary monitoring equipment at preset intervals at the boundary of the construction site. Based on the PM index, collect the PM data at the boundary through the deployed boundary monitoring equipment. Deploy drones equipped with monitoring equipment to fly and monitor the interior of the construction site according to the preset route, and collect the PM data and environmental-related data inside the construction site;

[0012] Perform grid processing on the construction site to obtain the construction data inside the construction site. Use the construction data to mark the grid blocks, cluster the grid blocks according to the marking information, and use the clustering results to divide the construction site into regions;

[0013] Identify the construction content and construction stage according to the marking information of the region, use the construction content and construction stage to identify the key construction areas, and add attention feedback to the drones for intensive monitoring and attention for the key construction areas;

[0014] Generate multi-modal perception data corresponding to the PM concentration according to the PM data at the boundary and the PM data inside in different regions, and preprocess the multi-modal perception data.

[0015] In this solution, use the edge computing module to preliminarily analyze the multi-modal perception data in real time and identify dust anomalies, specifically as follows:

[0016] Integrate the edge computing module on the boundary monitoring equipment and drones of the construction site, import the multi-modal perception data into the edge computing module for preliminary analysis, and obtain the PM concentration change sequence corresponding to different regions;

[0017] Retrieve construction examples through similarity calculation based on the construction content and stages in different regions. Obtain historical dust monitoring data from the construction examples, perform data preprocessing and averaging on the historical dust monitoring data to obtain a reference PM concentration change sequence;

[0018] Construct an autoencoder model, train it using the reference PM concentration change sequences corresponding to different regions, perform feature extraction through the random forest algorithm, divide the extracted features into feature subsets using hierarchical clustering, map the feature subsets into the autoencoder for feature reconstruction;

[0019] Calculate the root mean square value of each feature subset, calculate the anomaly score based on the root mean square value, use the test data to test the trained autoencoder model, and import the PM concentration change sequences corresponding to different regions into the trained autoencoder network;

[0020] Calculate the anomaly score based on the root mean square value of feature reconstruction, and judge the dust anomaly situation in different regions through the Softmax function.

[0021] In this solution, mark the multi-modal perception data according to the dust anomaly situation and give priority to quickly transmitting it to the data processing center. Specifically:

[0022] When a dust anomaly situation is recognized, mark the multi-modal perception data corresponding to the PM concentration as "urgent pollution warning data" and give priority to quickly transmitting it to the data processing center;

[0023] Locate according to the dust anomaly situation, select the surrounding dust suppression equipment through the location information, use the feature subset corresponding to the PM concentration change sequence of the dust anomaly situation as the anomaly feature, and calculate the similarity between the anomaly feature and each call instance data of the simple dust suppression strategy stored locally;

[0024] Obtain the call instance data that meets the preset standard for similarity, count the call frequency of the simple dust suppression strategy in the selected call instance data, and independently control the surrounding dust suppression equipment to perform preliminary dust suppression operations according to the simple dust suppression strategy with the highest call frequency.

[0025] In this solution, construct a dynamic pollution diffusion model to analyze the current pollution scenario of the construction site. Specifically:

[0026] Obtain the multi-source data sequences such as the multi-modal perception data corresponding to the PM concentration of the construction site, the operation data of the dust suppression equipment, the construction activity data, the site geological condition data, the surrounding traffic flow data, and the environmental related data;

[0027] Use the dynamic time warping algorithm to quantitatively judge the correlation between different regions based on the multi-source data sequences of different regions of the construction site. Take different regions of the construction site as nodes, and construct an edge structure based on the correlation between different regions to generate an association graph corresponding to the construction area.

[0028] Construct a dynamic pollution diffusion model based on the graph attention network and the gated recurrent unit. Import the multi-source data sequence corresponding to the dust anomaly situation into the dynamic pollution diffusion model to obtain the adjacency matrix of the regional nodes in the dust anomaly situation area, and import the adjacency matrix into the graph attention network for representation learning.

[0029] Use graph convolution for information propagation, introduce the multi-head attention mechanism, output the regional node feature representation according to the weights of the regional nodes in the attention layer, splice the regional node feature representations output by multiple attention heads, and perform splicing integration through an average operation in the last layer to achieve neighbor aggregation of regional nodes and obtain the updated regional node feature representation of the dust anomaly situation.

[0030] Import the regional node feature representation of the dust anomaly situation into the gated recurrent unit to obtain the feature time dependence, output the spatio-temporal features of the PM concentration change in the dust anomaly situation area through a fully connected layer, and obtain the dust generation trend to characterize the current pollution scenario according to the spatio-temporal features of the PM concentration change in the dust anomaly situation area.

[0031] In this solution, match the best dust suppression strategy for the current pollution scenario, specifically:

[0032] Predict the PM concentration distribution based on the current pollution scenario of the construction site, update the dust anomaly situation area in the construction site according to the predicted PM concentration distribution, and obtain the dust suppression equipment within the preset distance according to the updated dust anomaly situation area. The dust suppression equipment includes sprinkler devices, fog guns, and drones equipped with spray systems.

[0033] Read historical dust suppression instances according to the dust suppression equipment category, select the Euclidean distance, Manhattan distance, Frechet distance, and Jaccard distance to calculate the four distances between the historical dust suppression instances and the current working conditions of the dust suppression equipment respectively, and obtain the distance feature vector.

[0034] Introduce the Stacking strategy to construct a dust suppression strategy recommendation model. Train the KNN model separately based on each distance feature vector, output a preset number of strategy candidate sets, and score the candidate strategies in combination with the historical dust suppression effect.

[0035] Splice the strategy candidate sets and scores corresponding to different distances into a meta-feature vector, use the meta-feature vector to train a neural network model, learn the weights of each distance metric, and iteratively train to output the dust suppression strategy recommendation model.

[0036] Obtain the current pollution scenario based on the updated area of dust anomaly, import the current pollution scenario into the dust reduction strategy recommendation model for strategy screening and scoring, and obtain the optimal dust reduction strategy.

[0037] In this solution, a self-organizing communication network for dust reduction equipment rooms is constructed, and the dust reduction operation plan is adjusted using the dust reduction feedback information of the dust reduction equipment. Specifically:

[0038] Construct a self-organizing communication network for dust reduction equipment rooms through the data blockchain. When a dust reduction equipment detects that the dust concentration in its coverage area exceeds the preset concentration threshold, it sends a distress signal to the surrounding dust reduction equipment through the self-organizing communication network, and the surrounding equipment automatically adjusts the operation plan to cooperate in dust reduction;

[0039] During the operation of the dust reduction equipment, the data collected by the dust reduction equipment is stored on the chain, providing a credible data source for pollution diffusion analysis, and real-time monitoring of the working status of the dust reduction equipment. When the dust reduction equipment has an abnormal working status, the dust concentration in the coverage area of the abnormal dust reduction equipment is obtained. When it exceeds the preset concentration threshold, a distress signal is sent to the surrounding dust reduction equipment through the self-organizing communication network, and the surrounding equipment automatically adjusts the operation plan.

[0040] The second aspect of the present invention provides a construction site dust reduction and pollution prevention and control system based on multi-modal perception. The system includes a data collection module, an edge computing module, a data processing center module, and a dust reduction equipment management module;

[0041] The data collection module uses boundary monitoring equipment preset at the construction site boundary to collect PM data at the boundary in real time. At the same time, a drone equipped with monitoring equipment is used to conduct flight monitoring of the interior of the construction site according to a preset route to collect PM data in different areas of the site;

[0042] The edge computing module integrates PM data and environment-related data to generate multi-modal perception data, and preliminarily analyzes the multi-modal perception data to identify dust anomalies and mark the multi-modal perception data;

[0043] The data processing center module analyzes the current pollution scenario of the construction site according to the dust generation characteristics in different construction stages and the multi-modal perception data, and matches the best dust reduction strategy for the current pollution scenario.

[0044] The dust reduction equipment management module generates a dust reduction equipment control signal based on the best dust reduction strategy, controls the surrounding dust reduction equipment to perform dust reduction operations, constructs a self-organizing communication network for dust reduction equipment rooms, and adjusts the dust reduction operation plan using the dust reduction feedback information of the dust reduction equipment.

[0045] Compared with the prior art, the beneficial effects of the present invention are:

[0046] Through the multi-modal data collection of boundary monitoring devices and drones, combined with preliminary edge computing analysis and abnormal data processing, the present invention realizes comprehensive, real-time, and accurate dust concentration monitoring of the construction site, and timely grasps the pollution situation. It realizes intelligent coordination of the spraying system at the factory boundary, the sprinkler spraying system, the fog cannon, and the spraying system carried by the drone, accurately starts and adjusts the dust reduction operation according to the change of dust concentration, and improves the dust reduction efficiency. By using intelligent adaptive spraying technology, drone swarm cooperation, flexible dust reduction strategies, and self-organizing network communication between devices, the dust reduction effect is further optimized, the water resource utilization efficiency is improved, and the dust reduction scope is broadened.

[0047] When the dust concentration exceeds the standard, the dust reduction equipment is started in time and stopped in time after reaching the standard, avoiding over-spraying and equipment idling, achieving the purpose of energy conservation, effectively preventing secondary pollution at the same time, and creating a dust-free construction site. The historical data learning and prediction model realizes preventive dust reduction, adjusts the strategy in combination with weather changes, and further improves the dust reduction effect. The entire dust reduction process is automatically analyzed and decision-making by the data processing center according to the monitoring data and controls the operation of the equipment, reducing manual intervention and improving management efficiency. The multi-system linkage and data sharing optimize the overall management process of the construction site, and the blockchain technology ensures the data security and the standardization of equipment maintenance management. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments or exemplary of the present invention, the following will briefly introduce the drawings required for use in the embodiments or exemplary descriptions. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to the drawings shown.

[0049] Figure 1 The flowchart of the dust reduction and pollution prevention and control method for the construction site based on multi-modal perception is shown;

[0050] Figure 2 The flowchart of the preliminary analysis of multi-modal perception data to identify dust anomalies is shown;

[0051] Figure 3 The flowchart of matching the best dust reduction strategy for the current pollution scenario is shown;

[0052] Figure 4 The block diagram of the dust reduction and pollution prevention and control system for the construction site based on multi-modal perception is shown. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] In order to be able to more clearly understand the above objects, features, and advantages of the present invention, the present invention will be further described in detail below in conjunction with the drawings and specific embodiments. It should be noted that, without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other.

[0054] In the following description, many specific details are set forth in order to provide a thorough understanding of the present invention. However, the present invention may be implemented in other ways different from those described herein. Therefore, the scope of protection of the present invention is not limited by the specific embodiments disclosed below.

[0055] Figure 1 A flowchart of a dust reduction and pollution prevention and control method for a construction site based on multimodal perception is shown.

[0056] As Figure 1 shown, this embodiment provides a dust reduction and pollution prevention and control method for a construction site based on multimodal perception, including:

[0057] S102, through boundary monitoring devices preset at the boundary of the construction site, collect PM data at the boundary in real time. At the same time, use a drone equipped with monitoring devices to fly and monitor the interior of the construction site according to a preset flight path, and collect PM data of different areas of the site;

[0058] S104, combine the monitored PM data with the environmental-related data obtained by the temperature and humidity sensors to generate multimodal perception data, and use an edge computing module to preliminarily analyze the multimodal perception data in real time to identify abnormal dust conditions;

[0059] S106, mark the multimodal perception data according to the abnormal dust conditions, and preferentially and quickly transmit it to the data processing center, construct a dynamic pollution diffusion model to analyze the current pollution scenario of the construction site, and match the best dust reduction strategy for the current pollution scenario;

[0060] S108, generate a dust reduction equipment control signal based on the best dust reduction strategy, control the surrounding dust reduction equipment to perform dust reduction operations, construct a self-organizing communication network among the dust reduction equipment, and use the dust reduction feedback information of the dust reduction equipment to adjust the dust reduction operation plan.

[0061] It should be noted that boundary monitoring devices are deployed at the construction site boundary at preset intervals, equipped with high-precision PM sensors. The deployed boundary monitoring devices collect PM data at the boundary based on PM indicators, including indicators such as PM2.5 and PM10. The deployment interval of the boundary monitoring devices is not greater than the effective monitoring radius of the PM concentration of a single device × the coverage coefficient (the value range of the coverage coefficient is 0.8 - 1.0, which is dynamically adjusted according to the historical concentration fluctuation range of the site dust). Referring to the requirement of "the interval of fixed monitoring devices does not exceed 50 meters" in the "Technical Standard for Construction Dust Monitoring" (JGJ / T 393 - 2016), combined with the distribution density of the dust pollution sources at this site, it is calculated through the formula: interval distance = reference distance × correction coefficient, where the correction coefficient is determined according to parameters such as the aspect ratio of the site and the dominant wind direction frequency. Deploy drones equipped with monitoring devices, conduct flight tests and calibrations on the drones, set the flight route to ensure that they can fly stably and accurately collect data, and conduct flight monitoring on the interior of the construction site according to the preset route to collect PM data inside the construction site. In addition to the PM sensor, the drone monitoring device can also be equipped with temperature and humidity sensors, etc., to obtain environmental-related data for auxiliary analysis. The drone route can be adaptively adjusted according to real-time data. The initial preset route is a grid traversal route covering the entire site (such as an S-shaped path). When the edge computing module detects that the PM concentration in a certain grid block exceeds 1.5 times the threshold, the route optimization algorithm is triggered: encrypt the route density in this area, reduce the route spacing to 60% of the original spacing, increase the number of orbits, and extend the single monitoring time by 20% - 30%.

[0062] The construction site is gridded to obtain the construction data inside the construction site. The grid blocks are marked using the construction data, the grid blocks are clustered according to the marking information, and the construction site is divided into regions using the clustering results; the construction content and construction stage are identified according to the marking information of the region, and the key construction areas are identified using the construction content and construction stage, and attention feedback is added to the drone for intensive monitoring and attention for the key construction areas; in different regions, multi-modal perception data corresponding to the PM concentration is generated based on the PM data at the boundary and the PM data inside, and the multi-modal perception data is preprocessed.

[0063] Figure 2 The flowchart showing the preliminary analysis of multi-modal perception data to identify dust abnormal conditions is shown.

[0064] According to the embodiments of the present invention, an edge computing module is used to preliminarily analyze the multi-modal perception data in real time to identify dust abnormal conditions, specifically as follows:

[0065] S202. Integrate an edge computing module on the boundary monitoring device and the UAV at the construction site, import the multi-modal perception data into the edge computing module for preliminary analysis, and obtain the PM concentration change sequences corresponding to different regions.

[0066] S204. Retrieve construction examples through similarity calculation based on the construction content and construction stages in different regions, obtain historical dust monitoring data from the construction examples, perform data preprocessing and averaging on the historical dust monitoring data, and obtain the reference PM concentration change sequences.

[0067] S206. Construct an autoencoder model, train it using the reference PM concentration change sequences corresponding to different regions, perform feature extraction through the random forest algorithm, divide the extracted features into feature subsets using hierarchical clustering, map the feature subsets into the autoencoder, and perform feature reconstruction.

[0068] S208. Calculate the root mean square value of each feature subset, calculate the anomaly score based on the root mean square value, test the trained autoencoder model using test data, and import the PM concentration change sequences corresponding to different regions into the trained autoencoder network.

[0069] S210. Calculate the anomaly score based on the root mean square value of the feature reconstruction, and judge the dust anomaly situation in different regions through the Softmax function.

[0070] It should be noted that when constructing the autoencoder model, the autoencoder model consists of two autoencoders with the same network structure. The first-layer autoencoder is composed of parallel autoencoders with the same structure and is used to reconstruct the feature subsets. The second-layer autoencoder calculates the anomaly score based on the root mean square value of the feature reconstruction for anomaly detection. Each tree in the random forest randomly selects m features from the reference PM concentration change sequences corresponding to different regions to build a tree, calculates the scores according to the importance of each feature and arranges them in descending order, and selects a preset number of features for hierarchical clustering to obtain feature subsets with strong correlation. Import the PM concentration change sequences corresponding to different regions into the trained autoencoder network, obtain the predicted values of the PM concentration change sequences corresponding to different regions under normal conditions, and judge by comparing the root mean square of the difference between the predicted value and the actual value. If the root mean square value increases sharply and deviates from the normal value, it is judged as abnormal.

[0071] It should be noted that a data processing center is built, and the hardware and software environments required for configuring data analysis algorithms are ensured to enable the data processing center to efficiently process and analyze the received data. At the same time, stable communication links are established between the data processing center, boundary monitoring devices, drones, and dust suppression devices. Blockchain nodes are deployed, and interfaces with other management systems and meteorological systems are integrated. The operating status of the devices and communication links are regularly checked to ensure the stability and accuracy of data collection.

[0072] When abnormal dust conditions are identified, the multi-modal perception data corresponding to the PM concentration is marked as "urgent pollution warning data" and preferentially transmitted to the data processing center quickly. Locate according to the abnormal dust conditions, select the surrounding dust suppression devices based on the location information, use the feature subset corresponding to the PM concentration change sequence of the abnormal dust conditions as abnormal features, and calculate the similarity between the abnormal features and the call instance data of the simple dust suppression strategy stored locally. Obtain the call instance data that meets the preset criteria, count the call frequency of the simple dust suppression strategy among the selected call instance data, and autonomously control the surrounding dust suppression devices to perform preliminary dust suppression operations according to the simple dust suppression strategy with the highest call frequency.

[0073] It should be noted that the data processing center analyzes the collected PM data and environmental data, makes specific optimizations for the complex working conditions of the construction site, and combines time and space factors to judge the changing trend of dust concentration. Set a reasonable preset value for the dust concentration according to the environmental requirements and relevant standards of the construction site. Use historical data learning and prediction models, combined with multi-source data such as construction processes, weather, geology, and traffic, to predict the dust generation trend.

[0074] Obtain the multi-modal perception data corresponding to the PM concentration at the construction site, the operation data of the dust suppression devices, the construction activity data, the site geological condition data, the surrounding traffic flow data, and the environmental-related data (temperature and humidity data, wind direction data, weather data, etc.) as a multi-source data sequence; use the dynamic time warping algorithm to quantitatively judge the correlation between different regions based on the multi-source data sequence in different regions of the construction site. The dynamic time warping algorithm calculates the distance between the aligned sequences, which can intuitively represent the similarity between the multi-source data sequences of the two regions, and is used to represent the correlation between the regions.

[0075] Taking different regions of the construction site as nodes and constructing edge structures for the associations between different regions to generate an association graph corresponding to the construction regions; constructing a dynamic pollution diffusion model based on the graph attention network and the gated recurrent unit, importing the multi-source data sequence corresponding to the dust anomaly situation into the dynamic pollution diffusion model to obtain the adjacency matrix of the regional nodes in the dust anomaly situation area, and importing the adjacency matrix into the graph attention network for representation learning; using graph convolution for information propagation, introducing a multi-head attention mechanism, realizing parallel computing through multiple attention heads to accelerate the computing speed of the model, outputting the regional node feature representation according to the weights of the regional nodes in the attention layer, splicing the regional node feature representations output by multiple attention heads, and performing splicing integration through an average operation in the last layer to realize the neighbor aggregation of the regional nodes and obtain the updated regional node feature representation of the dust anomaly situation area; importing the regional node feature representation of the dust anomaly situation area into the gated recurrent unit, using the hidden state update to obtain the feature time dependence, and the gated recurrent unit overcomes the problems of gradient disappearance and gradient explosion in complex networks, and outputs the spatio-temporal features of the PM concentration change in the dust anomaly situation area through a fully connected layer, and obtains the dust generation trend to characterize the current pollution scenario according to the spatio-temporal features of the PM concentration change in the dust anomaly situation area. Set a preset value for the dust concentration. When the PM data monitored continuously for a period of time (such as 15 minutes) exceeds the preset value, the data processing center determines that dust reduction measures need to be started, and generates corresponding dust reduction instructions according to the area and degree of the dust concentration exceeding the standard.

[0076] Figure 3 The flowchart shows matching the best dust reduction strategy for the current pollution scenario.

[0077] According to the embodiments of the present invention, matching the best dust reduction strategy for the current pollution scenario specifically includes:

[0078] S302, predicting the PM concentration distribution based on the current pollution scenario of the construction site, updating the dust anomaly situation area in the construction site according to the predicted PM concentration distribution, and obtaining the dust reduction equipment within a preset distance according to the updated dust anomaly situation area. The dust reduction equipment includes sprinkler devices, fog guns, and drones equipped with spraying systems;

[0079] S304, reading historical dust reduction instances according to the dust reduction equipment category, calculating four distances between the historical dust reduction instances and the current working conditions of the dust reduction equipment respectively by selecting the Euclidean distance, Manhattan distance, Fréchet distance, and Jaccard distance, and obtaining a distance feature vector;

[0080] S306, introducing a Stacking strategy to construct a dust reduction strategy recommendation model, training a KNN model separately based on each distance feature vector, outputting a preset number of strategy candidate sets, and scoring the candidate strategies in combination with the historical dust reduction effect;

[0081] S308. Concatenate the policy candidate sets and scores corresponding to different distances into a meta-feature vector, and use the meta-feature vector to train a neural network model to learn the weights of various distance metrics, and iteratively train to output a dust suppression policy recommendation model.

[0082] S310. Obtain the current pollution scenario based on the updated dust anomaly area, import the current pollution scenario into the dust suppression policy recommendation model for policy screening and scoring, and obtain the best dust suppression policy.

[0083] It should be noted that historical dust suppression instances usually include multi-dimensional environmental data (such as PM2.5, humidity, wind speed) and equipment operation parameters (such as fog cannon angle, water consumption). Stacking integration is used to improve the robustness of policy selection, which is applicable to intelligent dust suppression decision-making in dynamic construction environments. After implementing the selected best dust suppression policy, record the actual dust suppression effect and update the instance library to form a closed-loop learning. The dust suppression policy recommendation model analyzes the contribution degree of each distance under specific working conditions through a neural network model (for example, the Fréchet distance is more important in strong winds), and dynamically optimizes the input of the KNN network.

[0084] After receiving the dust suppression command, the sprinkler device, fog cannon, and drones equipped with a spraying system start working according to the command requirements. According to different construction stages, switch the operation modes and parameters of the dust suppression equipment, and intelligently adjust the size and shape of the sprayed droplets to accurately match different pollution scenarios. For example, in the foundation construction stage, a large-flow and low-pressure spraying mode is adopted, and in the main body construction stage, a small-flow and high-pressure spraying mode is adopted. The sprinkler device adjusts the spraying flow rate and range according to the environmental perception matrix and intelligent flow regulating device, and evenly sprays the perimeter of the factory. The fog cannon adjusts the angle and intensity according to the command to perform targeted dust suppression on key areas. The drones fly to the designated area, turn on the spraying system, and perform dust suppression operations in the air. At the same time, multiple drones cooperate with each other, automatically plan the trajectory and spraying area according to the pollution area situation, and form a cooperative operation network. The drones in the drone group perform real-time information interaction through self-organizing network communication technology, and dynamically adjust the operation strategy according to the change of dust concentration.

[0085] Introduce the concept of flexible dust suppression, and dynamically adjust the operation modes and parameters of the dust suppression equipment according to the dust generation characteristics and environmental conditions in different construction stages. In the foundation construction stage, the dust suppression equipment adopts a large-flow and low-pressure spraying mode; in the main body construction stage, it switches to a small-flow and high-pressure spraying mode. In addition, add microbial flora to the sprayed liquid, use intelligent dust suppression materials to adsorb and fix dust, and convert organic dust into harmless substances to broaden the application range of the dust suppression system. The intelligent dust suppression materials have the characteristics of hygroscopic expansion, can adsorb dust when the dust concentration is high, and fix dust after the concentration decreases, reducing secondary dust emission.

[0086] During the dust reduction process, the boundary monitoring equipment and drones continuously monitor PM data. The data processing center receives the monitoring data in real time and compares it with the preset values. When the dust concentration index returns to the normal range and remains so for a period of time, the data processing center sends a stop work instruction to the dust reduction equipment. After receiving the instruction, the dust reduction equipment stops running to avoid unnecessary energy consumption and resource waste. Integrate with other management systems at the construction site (such as the safety monitoring system and the construction progress management system) to obtain data from other systems and optimize the dust reduction plan. At the same time, establish a real-time connection with the weather prediction system of the meteorological department and adjust the dust reduction strategy in advance according to weather changes.

[0087] It should be noted that blockchain technology, due to its characteristics such as decentralization, data immutability, and automatic execution of smart contracts, provides a new solution for building a self-organizing communication network among dust reduction devices. By combining blockchain with the Internet of Things (IoT), distributed decision-making and collaborative optimization of dust reduction devices can be achieved. Each dust reduction device serves as a node in the blockchain network, and through consensus mechanisms (such as PoA or PBFT), data synchronization and decision-making collaboration are realized without relying on a central server, improving the robustness and scalability of the system. Set dust reduction strategies based on smart contracts. For example, when the PM10 in a certain area exceeds the standard, nearby devices automatically negotiate the optimal dust reduction plan to avoid energy waste caused by multiple devices running at high power simultaneously. Dynamically adjust the contract logic in combination with environmental data to achieve adaptive control. The encryption mechanism of blockchain (such as asymmetric encryption and hash verification) ensures the communication security among dust reduction devices, preventing malicious nodes from forging data or launching attacks. The identities of dust reduction devices are authenticated through digital certificates to ensure that only legitimate devices are allowed to access the network.

[0088] Build a self-organizing communication network among dust reduction devices through the data blockchain. Preferably, a help signal trigger mechanism based on smart contracts is used to achieve communication among devices. The hash algorithm (such as SHA-256) and storage structure are used to upload data to the blockchain to ensure data integrity. When a dust reduction device detects that the dust concentration in its coverage area exceeds the preset concentration threshold, it sends a help signal to surrounding dust reduction devices through the self-organizing communication network, and the surrounding devices automatically adjust their operation plans to cooperate in dust reduction; during the operation of the dust reduction device, the data collected by the dust reduction device is uploaded to the blockchain for storage, providing a reliable data source for pollution diffusion analysis and real-time monitoring of the working status of the dust reduction device. When a dust reduction device has an abnormal working status, the dust concentration in the coverage area of the abnormal dust reduction device is obtained. When it exceeds the preset concentration threshold, a help signal is sent to surrounding dust reduction devices through the self-organizing communication network, and the surrounding devices automatically adjust their operation plans.

[0089] Regularly maintain and inspect the entire system, including the edge computing module, data processing center, blockchain nodes, self-organizing network communication module, and interfaces with other management systems and meteorological systems, etc., to ensure the continuous and stable operation of the system. Use blockchain smart contracts to manage equipment maintenance, ensuring the timeliness and standardization of equipment maintenance. Record information such as the maintenance cycle, maintenance content, and maintenance responsible party of the equipment. When the operation time of the equipment reaches the maintenance cycle, a maintenance reminder is automatically triggered; after the maintenance is completed, the maintenance record is uploaded to the blockchain for verification and confirmation.

[0090] Figure 4 The block diagram of a construction site dust reduction and pollution prevention and control system based on multi-modal perception is shown.

[0091] The second embodiment of the present invention provides a construction site dust reduction and pollution prevention and control system 4 based on multi-modal perception. The system includes a data acquisition module 401, an edge computing module 402, a data processing center module 403, and a dust reduction equipment management module 404;

[0092] The data acquisition module collects PM data at the boundary in real time through boundary monitoring equipment preset at the construction site boundary. At the same time, use drones equipped with monitoring equipment to conduct flight monitoring of the interior of the construction site according to a preset route, and collect PM data in different areas of the site;

[0093] The edge computing module integrates PM data and environment-related data to generate multi-modal perception data, and preliminarily analyzes the multi-modal perception data, identifies abnormal dust situations, and marks the multi-modal perception data;

[0094] The data processing center module analyzes the current pollution scenario of the construction site according to the dust generation characteristics in different construction stages and multi-modal perception data, and matches the best dust reduction strategy for the current pollution scenario.

[0095] The dust reduction equipment management module generates a dust reduction equipment control signal based on the best dust reduction strategy, controls the surrounding dust reduction equipment to perform dust reduction operations, constructs a self-organizing communication network among the dust reduction equipment, and adjusts the dust reduction operation plan using the dust reduction feedback information of the dust reduction equipment.

[0096] The third aspect of the present invention provides a computer-readable storage medium, which includes a program for a construction site dust reduction and pollution prevention and control method based on multi-modal perception. When the program for the construction site dust reduction and pollution prevention and control method based on multi-modal perception is executed by a processor, it realizes the steps of the construction site dust reduction and pollution prevention and control method based on multi-modal perception.

[0097] In several embodiments provided in the present application, it should be understood that the disclosed methods and systems can be implemented in other ways. The system embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods. For example, multiple modules or components can be combined, or can be integrated into another system, or some features can be ignored, or not executed. In addition, the coupling, direct coupling, or communication connection between the displayed or discussed components can be through some interfaces, indirect coupling or communication connection of devices or modules, and can be electrical, mechanical, or other forms. In addition, in each embodiment of the present invention, the functional modules can all be integrated in a processing module, or each module can be separately used as a module, or two or more modules can be integrated in a module; the above integrated modules can be implemented in the form of hardware, or in the form of hardware plus software functional modules.

[0098] Those of ordinary skill in the art can understand that all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions. The foregoing program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps including the above method embodiments; and the foregoing storage medium includes: mobile storage devices, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disks, or optical disks and other various media that can store program codes.

[0099] Alternatively, if the above integrated modules of the present invention are implemented in the form of software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present invention, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the various embodiments of the present invention. And the foregoing storage medium includes: mobile storage devices, ROM, RAM, magnetic disks, or optical disks and other various media that can store program codes.

[0100] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention.

Claims

1. A dust reduction and pollution prevention and control method for construction sites based on multi-modal perception, characterized in that, It includes the following steps: Through the boundary monitoring equipment preset at the construction site boundary, the PM data at the boundary is collected in real time. At the same time, a drone equipped with monitoring equipment is used to fly and monitor the interior of the construction site according to the preset route to collect PM data in different areas of the site; The monitored PM data is combined with the environmental related data obtained by the temperature and humidity sensor to generate multi-modal perception data. The edge computing module is used to preliminarily analyze the multi-modal perception data in real time to identify dust abnormal situations; Mark the multi-modal perception data according to the dust abnormal situation, and give priority to quickly transmitting it to the data processing center. A dynamic pollution diffusion model is constructed to analyze the current pollution scenario of the construction site, and the best dust reduction strategy is matched for the current pollution scenario; Based on the best dust reduction strategy, a dust reduction equipment control signal is generated to control the surrounding dust reduction equipment to perform dust reduction operations. A self-organizing communication network among the dust reduction equipment is constructed, and the dust reduction feedback information of the dust reduction equipment is used to adjust the dust reduction operation plan; Construct a dynamic pollution diffusion model to analyze the current pollution scenario of the construction site, specifically: Obtain the multi-modal perception data corresponding to the PM concentration of the construction site, the operation data of the dust reduction equipment, the construction activity data, the site geological condition data, the surrounding traffic flow data, and the environmental related data as a multi-source data sequence; Based on the multi-source data sequence of different areas of the construction site, the dynamic time warping algorithm is used to quantitatively judge the correlation between different areas. Different areas of the construction site are used as nodes, and the correlation between different areas is used to construct an edge structure to generate an association graph corresponding to the construction area; According to the graph attention network and the gated recurrent unit, a dynamic pollution diffusion model is constructed. The multi-source data sequence corresponding to the dust abnormal situation is imported into the dynamic pollution diffusion model to obtain the adjacency matrix of the area nodes of the dust abnormal situation, and the adjacency matrix is imported into the graph attention network for representation learning; Use graph convolution for information propagation, introduce the multi-head attention mechanism, output the regional node feature representation according to the weight of the regional node in the attention layer, splice the regional node feature representations output by multiple attention heads, and perform splicing integration through an average operation in the last layer to realize the neighbor aggregation of the regional node and obtain the updated regional node feature representation of the dust abnormal situation; The regional node feature representation of the dust abnormal situation is imported into the gated recurrent unit to obtain the feature time dependence, and the spatio-temporal features of the PM concentration change in the area of the dust abnormal situation are output through the fully connected layer. The current pollution scenario is obtained according to the spatio-temporal features of the PM concentration change in the area of the dust abnormal situation.

2. The dust reduction and pollution prevention and control method for construction sites based on multimodal perception according to claim 1, wherein, Collect the PM data of the construction site, specifically: Deploy boundary monitoring equipment at the construction site boundary at preset intervals. Based on the PM index, the boundary monitoring equipment deployed collects the PM data at the boundary. Deploy a drone equipped with monitoring equipment to fly and monitor the interior of the construction site according to the preset route to collect the PM data and environmental related data inside the construction site; Grid the construction site, obtain the construction data inside the construction site, use the construction data to mark the grid blocks, cluster the grid blocks according to the marking information, and divide the construction site into regions using the clustering results; Identify the construction content and construction stage according to the marking information of the region, use the construction content and construction stage to identify the key construction areas, and add attention feedback to the drone for intensive monitoring and attention for the key construction areas; Generate multi-modal perception data corresponding to the PM concentration according to the PM data at the boundary and the PM data inside in different regions, and preprocess the multi-modal perception data.

3. The construction site dust reduction and pollution prevention and control method based on multi-modal perception according to claim 1, characterized in that Use the edge computing module to perform preliminary analysis on the multi-modal perception data in real time to identify dust anomalies, specifically: Integrate the edge computing module on the boundary monitoring equipment and the drone of the construction site, import the multi-modal perception data into the edge computing module for preliminary analysis, and obtain the PM concentration change sequence corresponding to different regions; Retrieve construction examples through similarity calculation based on the construction content and construction stage of different regions, obtain historical dust monitoring data from the construction examples, and perform data preprocessing and averaging on the historical dust monitoring data to obtain a reference PM concentration change sequence; Construct an autoencoder model, train it using the reference PM concentration change sequences corresponding to different regions, perform feature extraction through the random forest algorithm, divide the extracted features into feature subsets using hierarchical clustering, map the feature subsets into the autoencoder for feature reconstruction; Calculate the root mean square value of each feature subset, calculate the anomaly score according to the root mean square value, use the test data to test the trained autoencoder model, and import the PM concentration change sequences corresponding to different regions into the trained autoencoder network; Calculate the anomaly score according to the root mean square value of the feature reconstruction, and judge the dust anomaly situation in different regions through the Softmax function.

4. The construction site dust reduction and pollution prevention and control method based on multi-modal perception according to claim 1, characterized in that, Mark the multi-modal perception data according to the dust anomaly situation and give priority to quickly transmitting it to the data processing center, specifically: When a dust anomaly situation is identified, mark the multi-modal perception data corresponding to the PM concentration as "urgent pollution warning data" and give priority to quickly transmitting it to the data processing center; Locate according to the dust anomaly situation, select the surrounding dust suppression equipment through the positioning information, use the feature subset corresponding to the PM concentration change sequence of the dust anomaly situation as the anomaly feature, and calculate the similarity between the anomaly feature and each call instance data of the simple dust suppression strategy stored locally; Obtain the call instance data whose similarity meets the preset standard, count the call frequency of the simple dust suppression strategy in the selected call instance data, and independently control the surrounding dust suppression equipment to perform preliminary dust suppression operations according to the simple dust suppression strategy with the highest call frequency.

5. The method for dust reduction and pollution prevention and control at a construction site based on multimodal perception according to claim 1, wherein Match the best dust suppression strategy for the current pollution scenario, specifically: Predict the PM concentration distribution based on the current pollution scenario of the construction site, update the dust abnormal situation area within the construction site according to the predicted PM concentration distribution, and obtain the dust suppression equipment within the preset distance according to the updated dust abnormal situation area. The dust suppression equipment includes a sprinkler device, a fog cannon, and a drone equipped with a spraying system; Read historical dust suppression instances according to the dust suppression equipment category, select the Euclidean distance, Manhattan distance, Frechet distance, and Jaccard distance to calculate four distances between the historical dust suppression instances and the current working conditions of the dust suppression equipment respectively, and obtain the distance feature vector; Introduce the Stacking strategy to construct a dust suppression strategy recommendation model. Train the KNN model separately based on each distance feature vector, output a preset number of strategy candidate sets, and score the candidate strategies in combination with the historical dust suppression effect; Concatenate the strategy candidate sets and scores corresponding to different distances into a meta-feature vector, use the meta-feature vector to train a neural network model, learn the weights of each distance metric, and iteratively train to output a dust suppression strategy recommendation model; Obtain the current pollution scenario according to the updated dust abnormal situation area, import the current pollution scenario into the dust suppression strategy recommendation model for strategy screening and scoring, and obtain the best dust suppression strategy.

6. The dust reduction and pollution prevention and control method for construction sites based on multi-modal perception according to claim 1, wherein, Construct a self-organizing communication network between dust suppression equipment, and use the dust suppression feedback information of the dust suppression equipment to adjust the dust suppression operation plan. Specifically: Construct a self-organizing communication network between dust suppression equipment through a data blockchain. When a dust suppression equipment finds that the dust concentration within its coverage area exceeds the preset concentration threshold, it sends a help signal to the surrounding dust suppression equipment through the self-organizing communication network, and the surrounding equipment automatically adjusts the operation plan to cooperate in dust suppression; During the operation of the dust suppression equipment, the data collected by the dust suppression equipment is stored on the chain, providing a credible data source for pollution diffusion analysis, and real-time monitoring of the working state of the dust suppression equipment. When the dust suppression equipment has an abnormal working state, the dust concentration within the coverage area of the abnormal dust suppression equipment is obtained. When it exceeds the preset concentration threshold, a help signal is sent to the surrounding dust suppression equipment through the self-organizing communication network, and the surrounding equipment automatically adjusts the operation plan.

7. A dust reduction and pollution prevention and control system for construction sites based on multimodal perception, characterized in that, Implement the construction site dust suppression and pollution prevention and control method based on multi-modal perception as described in any one of claims 1-6. The system includes a data collection module, an edge computing module, a data processing center module, and a dust suppression equipment management module; The data collection module collects PM data at the boundary in real time through boundary monitoring equipment preset at the construction site boundary. At the same time, a drone equipped with monitoring equipment is used to fly and monitor the interior of the construction site according to a preset flight path to collect PM data in different areas of the site; The edge computing module integrates PM data and environment-related data to generate multi-modal perception data, and preliminarily analyzes the multi-modal perception data to identify dust abnormal situations and mark the multi-modal perception data; The data processing center module analyzes the current pollution scenario of the construction site according to the dust generation characteristics and multi-modal perception data in different construction stages, and matches the best dust suppression strategy for the current pollution scenario. The dust suppression equipment management module generates a dust suppression equipment control signal based on the optimal dust suppression strategy, controls the surrounding dust suppression equipment to perform dust suppression operations, constructs a self-organizing communication network among the dust suppression equipment, and adjusts the dust suppression operation plan using the dust suppression feedback information of the dust suppression equipment.

Images