Intelligent analysis system based on multi-source big data fusion

Abstract

The invention discloses an intelligent analysis system based on multi-source big data fusion, which relates to the field of data analysis and comprises a particle monitoring module, an air humidity monitoring module, a wind speed monitoring module, a ground humidity monitoring module, a test module and a data processing module. The particle monitoring module is used for monitoring the particle concentration in the air of the target area; the air humidity monitoring module is used for monitoring the air humidity of the target area; according to the intelligent analysis system based on multi-source big data fusion, whether particles can be reduced to be lower than a set particle concentration threshold value within a set time length or not is judged according to the current wind speed, air humidity and particle concentration, and the particle concentration can be reduced to be lower than the set particle concentration threshold value at the current wind speed and air humidity. And when the particle concentration cannot be reduced to be lower than the set particle concentration threshold value within the set time length, a worker is reminded, so that the worker can conveniently carry out spraying, watering and other treatment on the corresponding target area in time, and the particle inhibition capability of the target area is improved.

Description

Technical Field

[0001] The present invention relates to the field of data analysis, and particularly to an intelligent analysis system based on multi-source big data fusion. Background Art

[0002] In the process of urban environmental management, in order to reduce the content of particulate matter (such as PM2.5) in urban air, it is necessary to sprinkle water on the particulate matter content area to ensure the air humidity and ground humidity in the area, so that the water in the air can adsorb particulate matter, increase the weight of particulate matter and make the particulate matter settle quickly, or form a water film on the ground to improve the adsorption capacity of ground particulate matter, achieving the effect of reducing the content of particulate matter in the air.

[0003] The prior art with the publication number CN116660463A discloses a big data air quality detection and alarm system, including: a plurality of air detection stations for detecting air quality and saving the detection results; a data analysis device for receiving the detection results of the plurality of air detection stations. The data analysis device includes a data preprocessing module and a data analysis module. The data preprocessing module is used to sort out and summarize the received detection results, and the data analysis module analyzes the sorted and summarized detection results. This kind of big data air quality detection and alarm system can realize information data sharing in air quality detection and alarm through big data technology, give full play to the value of atmospheric environment monitoring data. In the process of big data application, big data and detection data can be combined, and with the help of the constructed big data model, the detection results can be presented intuitively, thereby improving the effectiveness of atmospheric environment monitoring data.

[0004] However, due to the differences in the locations to be monitored, the site environments, such as terrain and building conditions, will also be different, resulting in differences in the wind speeds at the monitoring locations. Wind speed is also an important factor affecting the content of particulate matter in the air. The differences in air humidity, ground humidity and wind speed at different monitoring locations lead to differences in the ability of different monitoring locations to inhibit particulate matter in the air. Therefore, the intervention timing of the solution of setting a threshold to judge whether the particulate matter concentration in the air is greater than the threshold to determine whether artificial intervention is needed for the particulate matter at the monitoring location to reduce the particulate matter concentration at the monitoring location needs to be further optimized. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent analysis system based on multi-source big data fusion to solve the above-mentioned deficiencies in the prior art.

[0006] To achieve the above purpose, the present invention provides the following technical solution: An intelligent analysis system based on multi-source big data fusion, including a particle monitoring module, an air humidity monitoring module, a wind speed monitoring module, a ground humidity monitoring module, a test module, and a data processing module; The particle monitoring module is used to monitor the particle concentration in the target area. Among them, the particle monitoring module can use an air quality monitor to monitor the concentration of particulate matter in the air (such as PM2.5, PM10) in real time to achieve the monitoring of the particle concentration in the air; use a laser particle counter to measure the number and concentration of particulate matter with different diameters in the air based on the principle of laser scattering to achieve the monitoring of the particle concentration in the air; use an optical sensor to monitor the concentration of particulate matter through the principle of light scattering or light absorption to achieve the monitoring of the particle concentration in the air; The air humidity monitoring module is used to monitor the air humidity in the target area, and a humidity sensor can be used to achieve the monitoring of the air humidity; The wind speed monitoring module is used to monitor the wind speed in the target area, and an anemometer can be used to monitor the wind speed; The ground humidity monitoring module is used to monitor the ground humidity in the target area based on machine vision, and the ground humidity includes the height of the ground water volume and the percentage of water content in the ground; The test module is used to test the change rate of the particle concentration in the air in the target area under various air humidity, ground humidity and wind speed conditions in the target area.

[0007] Among them, there are multiple particle monitoring modules, air humidity monitoring modules, wind speed monitoring modules, and ground humidity monitoring modules, which are distributed at various positions in the target area to facilitate the monitoring of the corresponding data at different positions in the target area, so as to improve the monitoring range and accuracy.

[0008] The data processing module is used to obtain the particle concentration monitored by the particle monitoring module, the air humidity monitored by the air humidity monitoring module, the wind speed monitored by the wind speed monitoring module, and the ground humidity monitored by the ground humidity monitoring module; The data processing module is also used to train a machine learning model based on the particle concentration, air humidity and wind speed to obtain an air humidity prediction model. The air humidity prediction model is used to predict the predicted air humidity that makes the particle suppression effect in the target area reach the set particle change rate threshold through the input particle concentration, wind speed, and air humidity; among them, the present invention does not limit the specific machine learning model, and can be selected according to actual training needs to ensure the accuracy of the trained model; Further, the ground humidity monitoring module monitors the ground humidity in the target area based on machine vision, including the following steps: A1. Obtain the historical hyperspectral image of the ground in the target area; A2. Set multiple ground humidity labels, and associate the historical hyperspectral images with the corresponding ground humidity labels respectively; A3. Based on hyperspectral images and ground humidity labels, train a convolutional neural network model to obtain a ground humidity recognition model. The ground humidity recognition model is used to output the corresponding ground humidity label according to the input hyperspectral image. Specifically, during the process of training the convolutional neural network model, the following steps can be carried out: a. Data preprocessing: Process the hyperspectral images, including resizing, normalization, denoising, etc., to ensure data consistency and the training effect of the model. b. Model selection: Based on the actual situation and considering model complexity and computing resources, select a suitable model architecture. Commonly used CNN architectures include VGG, ResNet, EfficientNet, etc. c. Training dataset division: Divide the data (i.e., hyperspectral images and ground humidity labels) into a training set, a validation set, and a test set according to a certain ratio (such as 6:3:1) to evaluate the generalization ability of the model. d. Model training: Use the training set to train the model. The optimization process usually adopts the gradient descent algorithm to adjust hyperparameters such as the learning rate and batch size to improve the model performance. e. Model validation: Evaluate the performance of the model on the validation set and adjust the model parameters to avoid overfitting or underfitting.

[0009] d. Testing and evaluation: Evaluate the final effect of the model on the test set, and use metrics such as the mean squared error (MSE) and root mean squared error (RMSE) to measure the accuracy of the prediction and ensure that the prediction accuracy meets the requirements.

[0010] A4. Obtain the hyperspectral image of the ground in the target area. A5. Input the hyperspectral image into the ground humidity recognition model to obtain the corresponding ground humidity label.

[0011] Furthermore, the testing module is used to test the change speed of the particle concentration in the air in the target area under various air humidity, ground humidity, and wind speed conditions in the target area, and specifically includes the following steps: B1. Prepare a target area where the ground humidity is lower than the set minimum ground humidity threshold and the air humidity is lower than the set minimum air humidity threshold, monitor and record the air humidity and ground humidity in the target area, and the particle concentration in the target area is the set particle concentration threshold. B2. Monitor the particle concentration in the air in the target area under various wind speed conditions to obtain a scatter plot of the change of the particle concentration in the air with time at each wind speed under the corresponding air humidity and ground humidity conditions. B3. Judge whether the ground humidity in the target area is less than or equal to the set maximum ground humidity threshold. B4. If so, sprinkle water evenly on the target area to increase the ground humidity by a first set amount compared to the previously recorded ground humidity, record the ground humidity of the target area at this time, and return to B2. Sprinkling water can moisten the ground and form a water film on the ground surface, thereby making dust particles adhere more tightly to the ground and reducing the probability of their being blown up by the wind, thus reducing particles. B5. If not, perform regression analysis on each scatter plot to obtain the change rate of the particle concentration in the air under various air humidity, ground humidity, and wind speed conditions.

[0012] Furthermore, the testing module is used to test the change speed of the particle concentration in the air of the target area under various air humidity, ground humidity, and wind speed conditions of the target area, and further includes the following steps: C1. Prepare a target area where the ground humidity is lower than the set minimum ground humidity threshold and the air humidity is lower than the set minimum air humidity threshold, monitor and record the air humidity and ground humidity of the target area, and the particle concentration of the target area is the set particle concentration threshold. C2. Monitor the particle concentration in the air of the target area under various wind speed conditions to obtain a scatter plot of the change of the particle concentration in the air with time at each wind speed under the corresponding air humidity and ground humidity conditions. C3. Determine whether the air humidity of the target area is less than or equal to the set maximum air humidity threshold. C4. If so, spray water mist evenly on the target area to increase the air humidity by a second set amount compared to the previously recorded air humidity, record the air humidity of the target area at this time, and return to C2. Increasing the air humidity can increase the moisture in the air, thereby adsorbing more dust in the air, increasing the weight of the dust, and reducing the suspension time of the dust in the air, thus achieving the effect of suppressing particles. C5. If not, perform regression analysis on each scatter plot to obtain the change rate of the particle concentration in the air under various air humidity, ground humidity, and wind speed conditions.

[0013] Furthermore, the data processing module is also used to judge whether it is necessary to adjust the air humidity based on the particle concentration, air humidity, and wind speed, including the following steps: D1. Obtain the particle concentration, air humidity, and wind speed at each historical moment to obtain sample data. D2. Mark the sample data in which the particle concentration drops below the set particle concentration threshold within the set time length under various air humidity and wind speed conditions as the first sample data, and mark the sample data in which the particle concentration drops to be greater than or equal to the set particle concentration threshold within the set time length under various air humidity and wind speed conditions as the second sample data. D3. Set normal labels and abnormal labels, associate the normal labels with the first sample data, and associate the abnormal labels with the second sample data. D4. Train a first machine learning model based on the first sample data, normal labels, second sample data, and abnormal labels to obtain a particle processing anomaly judgment model. The particle processing anomaly judgment model is used to output corresponding normal labels or abnormal labels based on the input sample data. D5. Determine whether the output of the particle anomaly judgment processing model is an abnormal label. D6. If so, adjust the air humidity; if not, there is no need to adjust the air humidity.

[0014] Furthermore, the data processing module is also used to train a machine learning model based on the particle concentration, air humidity, and wind speed to obtain an air humidity prediction model. The air humidity prediction model is used to predict the predicted air humidity that can make the particle suppression effect in the target area reach the set particle change rate threshold by inputting the particle concentration, wind speed, and air humidity, including the following steps: E1. Train a second machine learning model based on the first sample data to obtain an air humidity prediction model. The air humidity prediction model is used to output the predicted air humidity that can reduce the particle concentration to below the particle concentration threshold within a set time period based on the input particle concentration and wind speed. E2. If the output of the particle anomaly judgment processing model is an abnormal label, input the particle concentration and wind speed into the air humidity prediction model to obtain the corresponding predicted air humidity.

[0015] Furthermore, the data processing module is also used to send a reminder when the particle suppression effect predicted by the air humidity prediction model is lower than the set particle change rate threshold, so that the staff can spray water on the target area after receiving the reminder.

[0016] 1. Compared with the prior art, the intelligent analysis system based on multi-source big data fusion provided by the present invention can, by setting a particle monitoring module, an air humidity monitoring module, a wind speed monitoring module, a ground humidity monitoring module, a test module, and a data processing module, determine whether the particles can drop to below the set particle concentration threshold within a set time period according to the current wind speed, air humidity, and particle concentration.

[0017] 2. Compared with the prior art, the intelligent analysis system based on multi-source big data fusion provided by the present invention can also remind the staff when the current wind speed and air humidity cannot make the particle concentration drop to below the set particle concentration threshold within a set time period, which is convenient for the staff to spray, sprinkle water, etc. on the corresponding target area in a timely manner, improve the particle suppression ability of the target area, and make the prediction of the timing of spraying and sprinkling water on the target area more accurate, reducing the phenomenon that the ground of the target area is too wet due to spraying and sprinkling water when the monitored particulate matter concentration exceeds the standard. Description of the Drawings

[0018] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments described in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0019] Figure 1 It is the system structure block diagram provided by the embodiment of the present invention; Figure 2 It is the monitoring step diagram of the ground humidity module provided by the embodiment of the present invention; Figure 3 It is the step diagram for judging whether it is necessary to adjust the air humidity provided by the embodiment of the present invention; Figure 4 It is the predicted air humidity step diagram for predicting that the particle suppression effect in the target area reaches the set particle change rate threshold provided by the embodiment of the present invention. Detailed implementation manners

[0020] In order to enable those skilled in the art to better understand the technical solutions of the present invention, the following will further introduce the present invention in detail with reference to the drawings.

[0021] In the description of the present invention, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of the described features. In the description of the present invention, "a plurality" means two or more unless otherwise specifically defined. In addition, the terms "mounted", "connected", and "connected" should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral connection; it may be a mechanical connection or an electrical connection; it may be directly connected or indirectly connected through an intermediate medium, and it may be the internal communication of two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0022] In the following, the exemplary embodiments will be more fully described with reference to the drawings, but the exemplary embodiments can be embodied in different forms and should not be construed as limited to the embodiments set forth herein. On the contrary, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0023] Without conflict, the various embodiments of the present disclosure and the features in the embodiments can be combined with each other.

[0024] As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0025] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. As used herein, the singular forms "a" and "the" are also intended to include the plural forms unless the context clearly indicates otherwise. It will also be understood that when the terms "comprises" and / or "consists of" are used in this specification, the specified features, wholes, steps, operations, elements, and / or components are present, but do not preclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or groups thereof.

[0026] Please refer to Figures 1 - 4 , an intelligent analysis system based on multi-source big data fusion, including a particle monitoring module, an air humidity monitoring module, a wind speed monitoring module, a ground humidity monitoring module, a test module, and a data processing module; The particle monitoring module is used to monitor the particle concentration in the air of the target area. Among them, the particle monitoring module can use an air quality monitor to monitor the concentration of particulate matter (such as PM2.5, PM10) in the air in real time to achieve the monitoring of the particle concentration in the air; use a laser particle counter based on the principle of laser scattering to measure the number and concentration of particulate matter with different diameters in the air to achieve the monitoring of the particle concentration in the air; use an optical sensor to monitor the concentration of particulate matter through the principle of light scattering or light absorption to achieve the monitoring of the particle concentration in the air; The air humidity monitoring module is used to monitor the air humidity of the target area, and a humidity sensor can be used to achieve the monitoring of the air humidity; The wind speed monitoring module is used to monitor the wind speed of the target area, and an anemometer can be used to monitor the wind speed; The ground humidity monitoring module is used to monitor the ground humidity of the target area based on machine vision. The ground humidity includes the height of the ground water volume and the percentage of water content in the ground, and includes the following steps: A1. Obtain the historical hyperspectral image of the ground in the target area; A2. Set multiple ground humidity labels and associate the historical hyperspectral images with the corresponding ground humidity labels respectively; A3. Based on the hyperspectral image and the ground humidity label, train a convolutional neural network model to obtain a ground humidity recognition model. The ground humidity recognition model is used to output the corresponding ground humidity label according to the input hyperspectral image. Specifically, in the process of training the convolutional neural network model, the following steps can be carried out: a. Data preprocessing: Process the hyperspectral image, including adjusting the size, normalizing, denoising, etc., to ensure the consistency of the data and the training effect of the model; b. Model selection: Based on the actual situation and considering model complexity and computing resources, select an appropriate model architecture. Commonly used CNN architectures include VGG, ResNet, EfficientNet, etc.; c. Training dataset division: Divide the data (i.e., hyperspectral images and ground humidity labels) into a training set, a validation set, and a test set according to a certain ratio (such as 6:3:1) to evaluate the generalization ability of the model; d. Model training: Use the training set to train the model. The optimization process usually adopts the gradient descent algorithm to adjust hyperparameters such as the learning rate and batch size to improve the model performance; e. Model validation: Evaluate the performance of the model on the validation set and adjust the model parameters to avoid overfitting or underfitting.

[0027] d. Testing and evaluation: Evaluate the final effect of the model on the test set, and use metrics such as the mean squared error (MSE) and root mean squared error (RMSE) to measure the accuracy of the prediction, ensuring that the prediction accuracy meets the requirements.

[0028] A4. Obtain the hyperspectral image of the ground in the target area; A5. Input the hyperspectral image into the ground humidity recognition model to obtain the corresponding ground humidity label.

[0029] The test module is used to test the change rate of the particle concentration in the air in the target area under various air humidity, ground humidity, and wind speed conditions in the target area.

[0030] Among them, there are multiple particle monitoring modules, air humidity monitoring modules, wind speed monitoring modules, and ground humidity monitoring modules, which are distributed at various positions in the target area to facilitate the monitoring of the corresponding data at different positions in the target area, so as to improve the monitoring range and accuracy.

[0031] The data processing module is used to obtain the particle concentration monitored by the particle monitoring module, the air humidity monitored by the air humidity monitoring module, the wind speed monitored by the wind speed monitoring module, and the ground humidity monitored by the ground humidity monitoring module; The data processing module is also used to train a machine learning model based on the particle concentration, air humidity, and wind speed to obtain an air humidity prediction model. The air humidity prediction model is used to predict the predicted air humidity that makes the particle suppression effect in the target area reach the set particle change rate threshold by inputting the particle concentration, wind speed, and air humidity; among them, the present invention does not limit the specific machine learning model, which can be selected according to actual training needs to ensure the accuracy of the trained model; In one embodiment, the test module is used to test the change rate of the particle concentration in the air in the target area under various air humidity, ground humidity, and wind speed conditions in the target area, specifically including the following steps: B1. Prepare a target area where the ground humidity is lower than the set minimum ground humidity threshold and the air humidity is lower than the set minimum air humidity threshold. Monitor and record the air humidity and ground humidity of the target area, and the particle concentration in the target area is the set particle concentration threshold; B2. Monitor the particle concentration in the air of the target area under various wind speeds, and obtain a scatter plot of the change of the particle concentration in the air over time at each wind speed under the corresponding air humidity and ground humidity conditions; B3. Determine whether the ground humidity in the target area is less than or equal to the set maximum ground humidity threshold; B4. If so, sprinkle water evenly on the target area to increase the ground humidity by a first set amount compared to the previously recorded ground humidity, and record the ground humidity of the target area at this time. Return to B2, where sprinkling water can moisten the ground and form a water film on the ground surface, thereby making dust particles adhere more tightly to the ground and reducing the probability of their being blown up by the wind, thus reducing particles; B5. If not, perform a regression analysis on each scatter plot to obtain the change rate of the particle concentration in the air under various air humidity, ground humidity, and wind speed conditions.

[0032] In another embodiment, the test module is used to test the change speed of the particle concentration in the air of the target area under various air humidity, ground humidity, and wind speed conditions of the target area, including the following steps: C1. Prepare a target area where the ground humidity is lower than the set minimum ground humidity threshold and the air humidity is lower than the set minimum air humidity threshold. Monitor and record the air humidity and ground humidity of the target area, and the particle concentration in the target area is the set particle concentration threshold; C2. Monitor the particle concentration in the air of the target area under various wind speeds, and obtain a scatter plot of the change of the particle concentration in the air over time at each wind speed under the corresponding air humidity and ground humidity conditions; C3. Determine whether the air humidity in the target area is less than or equal to the set maximum air humidity threshold; C4. If so, spray water mist evenly on the target area to increase the air humidity by a second set amount compared to the previously recorded air humidity, and record the air humidity of the target area at this time. Return to C2, where an increase in air humidity can increase the moisture in the air, thereby adsorbing more dust in the air and increasing the weight of the dust, reducing the suspension time of the dust in the air, and thus achieving the effect of suppressing particles; C5. If not, perform a regression analysis on each scatter plot to obtain the change rate of the particle concentration in the air under various air humidity, ground humidity, and wind speed conditions.

[0033] The data processing module is also used to determine whether it is necessary to adjust the air humidity based on the particle concentration, air humidity, and wind speed, including the following steps: D1. Obtain the particle concentration, air humidity, and wind speed at each historical moment to obtain sample data; D2. Mark the sample data where the particle concentration drops below the set particle concentration threshold within the set time duration under various air humidity and wind speed conditions as the first sample data, and mark the sample data where the particle concentration drops to be greater than or equal to the set particle concentration threshold within the set time duration under various air humidity and wind speed conditions as the second sample data; D3. Set normal labels and abnormal labels, associate the normal labels with the first sample data, and associate the abnormal labels with the second sample data; D4. Train a first machine learning model based on the first sample data, normal labels, second sample data, and abnormal labels to obtain a particle processing anomaly judgment model. The particle processing anomaly judgment model is used to output the corresponding normal label or abnormal label based on the input sample data; D5. Determine whether the output of the particle anomaly judgment processing model is an abnormal label; D6. If so, it is necessary to adjust the air humidity; if not, there is no need to adjust the air humidity.

[0034] The data processing module is also used to train a machine learning model based on the particle concentration, air humidity, and wind speed to obtain an air humidity prediction model. The air humidity prediction model is used to predict the predicted air humidity that enables the particle suppression effect in the target area to reach the set particle change rate threshold by inputting the particle concentration, wind speed, and air humidity, including the following steps: E1. Train a second machine learning model based on the first sample data to obtain an air humidity prediction model. The air humidity prediction model is used to output the predicted air humidity that enables the particle concentration to drop below the particle concentration threshold within the set time duration based on the input particle concentration and wind speed; E2. If the output of the particle anomaly judgment processing model is an abnormal label, input the particle concentration and wind speed into the air humidity prediction model to obtain the corresponding predicted air humidity.

[0035] The data processing module is also used to issue a reminder when the particle suppression effect predicted by the air humidity prediction model is lower than the set particle change rate threshold, and is used to sprinkle water on the target area after the staff receives the reminder.

[0036] Only some exemplary embodiments of the present invention have been described by way of illustration above. Undoubtedly, for those of ordinary skill in the art, without departing from the spirit and scope of the present invention, the described embodiments can be modified in various different ways. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the protection scope of the claims of the present invention.

Claims

1. An intelligent analysis system based on multi-source big data fusion, characterized by: It includes particle monitoring module, air humidity monitoring module, wind speed monitoring module, ground humidity monitoring module, testing module and data processing module; The particle monitoring module is used to monitor the concentration of particles in the air of the target area; The air humidity monitoring module is used to monitor the air humidity in the target area; The wind speed monitoring module is used to monitor the wind speed in the target area; The ground humidity monitoring module is used to monitor the ground humidity of the target area based on machine vision, and the ground humidity includes the ground water height and the ground water percentage; The test module is used to test the change rate of the particle concentration in the air of the target area under various conditions of air humidity, ground humidity and wind speed in the target area; The data processing module is used to obtain the particle concentration monitored by the particle monitoring module, the air humidity monitored by the air humidity monitoring module, the wind speed monitored by the wind speed monitoring module, and the ground humidity monitored by the ground humidity monitoring module; The data processing module is also used to train a machine learning model based on the particle concentration, air humidity and wind speed to obtain an air humidity prediction model. The air humidity prediction model is used to predict the predicted air humidity that will enable the particle suppression effect in the target area to reach a set particle change rate threshold through the input particle concentration, wind speed and air humidity.

2. The intelligent analysis system based on multi-source big data fusion according to claim 1 is characterized by: The ground humidity monitoring module monitors the ground humidity of the target area based on machine vision, including the following steps: A1. Obtain historical hyperspectral images of the ground in the target area; A2. Setting a plurality of ground humidity labels, and associating the historical hyperspectral images with corresponding ground humidity labels respectively; A3. Based on the hyperspectral image and the ground humidity label, a convolutional neural network model is trained to obtain a ground humidity recognition model, wherein the ground humidity recognition model is used to output a corresponding ground humidity label according to the input hyperspectral image; A4. Obtain a hyperspectral image of the ground in the target area; A5. Input the hyperspectral image into the ground humidity recognition model to obtain a corresponding ground humidity label.

3. The intelligent analysis system based on multi-source big data fusion according to claim 1 is characterized in that: The test module is used to test the change rate of the particle concentration in the air of the target area under various conditions of air humidity, ground humidity and wind speed in the target area, and specifically includes the following steps: A1. Prepare a target area where the ground humidity is lower than the set minimum ground humidity threshold and the air humidity is lower than the set minimum air humidity threshold, monitor and record the air humidity and ground humidity of the target area, and the particle concentration of the target area is the set particle concentration threshold; A2. Monitor the particle concentration in the air of the target area under various wind speed conditions, and obtain a scatter plot of the particle concentration in the air at various wind speeds versus time under corresponding air humidity and ground humidity conditions; A3, determining whether the ground humidity of the target area is less than or equal to a set maximum ground humidity threshold; A4. If yes, water the target area evenly to increase the ground humidity by a first setting amount compared to the ground humidity recorded last time, and record the ground humidity of the target area at this time, and return to A2, wherein watering can make the ground moist and form a water film on the ground surface, thereby making the dust particles adhere more tightly to the ground, reducing the probability of them being blown by the wind, thereby reducing the particles; A5. If not, perform regression analysis on each scatter plot to obtain the change rate of particle concentration in the air under different air humidity, ground humidity and wind speed conditions.

4. The intelligent analysis system based on multi-source big data fusion according to claim 1 is characterized in that: The test module is used to test the change rate of the particle concentration in the air of the target area under various conditions of air humidity, ground humidity and wind speed in the target area, and also includes the following steps: B1. Prepare a target area where the ground humidity is lower than the set minimum ground humidity threshold and the air humidity is lower than the set minimum air humidity threshold, monitor and record the air humidity and ground humidity of the target area, and the particle concentration of the target area is the set particle concentration threshold; B2. Monitor the particle concentration in the air of the target area under various wind speed conditions, and obtain a scatter plot of the particle concentration in the air at various wind speeds versus time under corresponding air humidity and ground humidity conditions; B3, determining whether the air humidity in the target area is less than or equal to a set maximum air humidity threshold; B4, if yes, spray water mist evenly on the target area to increase the air humidity by the second setting amount compared with the air humidity recorded last time, and record the air humidity of the target area at this time, and return to B2, wherein the increase in air humidity can increase the amount of water in the air, thereby absorbing dust in the air to increase the weight of the dust and reduce the suspension time of the dust in the air, thereby achieving the effect of suppressing particles; B5. If not, perform regression analysis on each scatter plot to obtain the change rate of particle concentration in the air under different air humidity, ground humidity and wind speed conditions.

5. The intelligent analysis system based on multi-source big data fusion according to claim 1 is characterized by: The data processing module is also used to determine whether the air humidity needs to be adjusted based on the particle concentration, air humidity and wind speed, including the following steps: B1. Obtain the particle concentration, air humidity and wind speed at each historical moment to obtain sample data; B2. Under various conditions of air humidity and wind speed, the sample data whose particle concentration drops below the set particle concentration threshold within the set time period is marked as the first sample data, and under various conditions of air humidity and wind speed, the sample data whose particle concentration drops to greater than or equal to the set particle concentration threshold within the set time period is marked as the second sample data; B3. Set a normal label and an abnormal label, associate the normal label with the first sample data, and associate the abnormal label with the second sample data; B4. Based on the first sample data, the normal label, the second sample data and the abnormal label, a first machine learning model is trained to obtain a particle processing abnormality judgment model, wherein the particle processing abnormality judgment model is used to output a corresponding normal label or an abnormal label based on the input sample data; B5. Determine whether the output of the particle abnormality determination processing model is an abnormal label; B6. If yes, the air humidity needs to be adjusted; if no, the air humidity does not need to be adjusted.

6. The intelligent analysis system based on multi-source big data fusion according to claim 5 is characterized by: The data processing module is also used to train a machine learning model based on the particle concentration, air humidity and wind speed to obtain an air humidity prediction model, wherein the air humidity prediction model is used to predict the predicted air humidity that enables the particle suppression effect in the target area to reach a set particle change rate threshold through the input particle concentration, wind speed and air humidity, and comprises the following steps: C1. Based on the first sample data, a second machine learning model is trained to obtain an air humidity prediction model, wherein the air humidity prediction model is used to output a predicted air humidity that causes the particle concentration to drop below the particle concentration threshold within a set time period based on the input particle concentration and wind speed; C2. If the particle anomaly judgment processing model outputs an abnormal label, the particle concentration and wind speed are input into the air humidity prediction model to obtain the corresponding predicted air humidity.

7. The intelligent analysis system based on multi-source big data fusion according to claim 6 is characterized by: The data processing module is also used to issue a reminder when the particle suppression effect predicted by the air humidity prediction model is lower than the set particle change rate threshold.

Images