Environment noise and chemical pollution joint detection and big data comprehensive analysis device
Through the joint detection device of environmental noise and chemical pollution, the convolutional neural network and long-term memory network are used to analyze environmental noise and chemical pollution data, solving the problem of lack of joint detection and comprehensive analysis in the existing technology, and realizing intelligent management and scientific decision-making support for environmental pollution data.
Patent Information
- Application Number
- CN202510307148.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-16
- Publication Date
- 2025-07-22
Smart Images
Figure CN120354067A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of combined detection of environmental pollution, and specifically relates to a device for combined detection of environmental noise and chemical pollution and comprehensive big data analysis. Background Art
[0002] With the acceleration of urbanization and the development of industrialization, the problems of environmental noise and chemical pollution are becoming increasingly serious, posing a huge threat to human health and the ecological environment. In order to effectively address these problems, a combined detection strategy is needed to combine the detection of environmental noise and chemical pollution to obtain more comprehensive and accurate environmental information. Environmental noise detection is a monitoring activity of the sounds that interfere with people's study, work and life and their sound sources. Chemical pollution detection is a process of detecting and analyzing possible chemical pollutants in the environment, aiming to evaluate their impact on the environment and human health.
[0003] The scope of environmental noise detection is extensive, and the detection items include noise level, noise spectrum analysis, remote noise monitoring, noise source localization, noise source identification, assessment of the impact of noise on the human body, assessment of the effect of noise control, etc. The objects of chemical pollution detection include inorganic pollutants such as heavy metals (lead, mercury, cadmium, chromium, etc.), non-metals (arsenic, fluoride, etc.), inorganic acids, inorganic bases, cyanides, sulfides, etc., organic pollutants such as petroleum, volatile organic compounds (VOCs), semi-volatile organic compounds (SVOCs), polycyclic aromatic hydrocarbons (PAHs), organochlorine pesticides, organophosphorus pesticides, etc., and specific pollutants such as persistent organic pollutants such as DDT, polychlorinated biphenyls (PCBs), and dioxins. These chemical pollutants contaminate soil, water bodies and the atmosphere. Among them, due to the strong fluidity of the atmosphere and its close relationship with people's lives, the detection of atmospheric chemical pollutants has become the top priority of environmental detection and treatment.
[0004] The existing methods for environmental noise pollution detection, chemical pollution detection and data analysis are still blank in terms of linking the two together for big data analysis. In addition, the detection and comprehensive analysis of thermal pollution, which has a strong connection with environmental noise pollution and chemical pollution, are also lacking. The traditional detection methods for the problem of automobile exhaust, which is closely related to people's lives, are often single-factor detection and analysis and evaluation. This problem is particularly suitable for combined detection of environmental noise and chemical pollution and comprehensive big data analysis. Only the results obtained from comprehensive analysis can more comprehensively and deeply evaluate the impact of the number of automobile driving trips on environmental pollution in cities and regions.
[0005] Therefore, it is necessary to propose an environmental noise and chemical pollution joint detection and big data comprehensive analysis device that can simultaneously detect the environmental noise pollution and atmospheric chemical pollution of urban streets, use neural networks to analyze and predict data, deeply reveal the relationship between automobile noise pollution, atmospheric chemical pollution and thermal pollution, and provide data support and assistance for urban managers to make decisions on traffic rule changes and urban automobile ownership assessment. Summary of the Invention
[0006] To solve the above problems, the object of the present invention is to provide an environmental noise and chemical pollution joint detection and big data comprehensive analysis device, which analyzes and distinguishes noise events and sound source types, inputs the sound source types, chemical pollution data and other environmental detection data into a long short-term memory network for training, extracts feature vectors to predict the impact of the occurrence of noise events on atmospheric chemical pollution and changes in other environmental detection data, and visualizes the prediction results, so as to deeply reveal the relationship between automobile noise pollution, atmospheric chemical pollution and thermal pollution, and improve the decision-making efficiency of urban managers.
[0007] To achieve the above object, the technical solution of the present invention is as follows: An environmental noise and chemical pollution joint detection and big data comprehensive analysis device, comprising:
[0008] A data acquisition unit, configured to collect environmental pollution data and image data in a set area by setting points, and use a trained convolutional neural network to identify and confirm pollution events based on the environmental pollution data and image data in time sequence to obtain and store sound source type data;
[0009] The environmental pollution data includes: the noise decibels, noise duration and excessive noise recordings collected by an environmental noise collector; the concentrations of sulfur dioxide, nitrogen dioxide, carbon monoxide, ozone, PM10 and PM2.5 obtained after sampling by an air sampler; the air temperature and humidity data collected by a temperature and humidity sensor; the wind direction and wind speed data collected by a wind sensor;
[0010] An Internet of Things data transmission module, configured to ensure data authenticity through Internet of Things authentication, and upload the data collected and stored by the data acquisition unit within a preset time period to a big data analysis center;
[0011] A big data analysis center, configured to classify and store the data collected by the data acquisition unit within a preset range, and perform subsequent analysis and processing on the data;
[0012] A data preprocessing module, configured to preprocess the collected data and convert the preprocessed data into standardized feature vectors;
[0013] A data analysis module for constructing a long short-term memory network to capture long-distance and short-distance dependencies in the data vector sequence, and optimizing and using feature vectors to train the long short-term memory network;
[0014] A model prediction module for using the trained long short-term memory network to predict the change trend of other environmental pollution data based on the change of a single vector in the environmental pollution data;
[0015] A visualization display module for reading geographical data and visualizing the combined predicted change trend of environmental pollution data, and the obtained chart is used to reveal the potential relationship between the changes in environmental pollution data.
[0016] The principle of the basic solution is as follows: By jointly detecting environmental noise and chemical pollution data, advanced neural networks (CNN and LSTM) are used for data processing and analysis to reveal the potential relationship between environmental pollution data. This includes data collection, preprocessing, feature extraction, model training and prediction, as well as visualization display of the results. The entire process is a closed-loop system. From data collection to result display, each link is closely connected to form a complete data analysis chain.
[0017] The beneficial effects of the basic solution are as follows: 1. Through the Internet of Things authentication technology, it is ensured that each piece of data uploaded to the big data analysis center has undergone strict authenticity verification, effectively avoiding the problems of data fraud and false alarms. The data preprocessing module adopts a standardized processing process to convert the original data into comparable feature vectors, further improving the accuracy and consistency of the data.
[0018] 2. By using advanced neural network models such as convolutional neural network (CNN) and long short-term memory network (LSTM), the potential relationship between environmental noise and chemical pollution data can be deeply explored, revealing complex relationships that are difficult to discover by traditional methods. Through model training and optimization, the system can more accurately predict the change trend of environmental pollution data, providing a more scientific basis for environmental management.
[0019] 3. The visualization display module presents the predicted change trend of environmental pollution data in an intuitive and easy-to-understand manner, enabling urban managers to quickly understand the environmental pollution situation and its change trend. The system also supports various data analysis and reporting functions, providing rich decision-making support information for urban managers and helping them make accurate decisions quickly.
[0020] 4. The entire system adopts intelligent data processing and analysis methods, capable of automatically collecting, analyzing, and predicting environmental pollution data, realizing the intelligence and automation of environmental management. The system also supports multiple data interfaces and extended functions, enabling seamless docking with other environmental management systems to achieve data sharing and collaborative processing, further enhancing the refinement and coordination of environmental management.
[0021] 5. By deeply analyzing the relationships among environmental pollution data, the system can provide a more scientific basis for urban planning, contribute to optimizing the urban layout and traffic planning, and reducing the impact of environmental pollution and noise pollution on urban residents. The system can also provide environmental pollution early warning and emergency response support for urban managers, helping to promptly respond to environmental pollution incidents and safeguard public health and safety.
[0022] Furthermore, the data acquisition unit includes a bracket. On one side of the top of the bracket, a crossbeam is vertically and fixedly connected. An environmental noise collector is fixedly connected to the top of the bracket. A variable-direction camera is fixedly connected below the end of the crossbeam far from the bracket. A sound source locator is fixedly connected above the end of the crossbeam far from the bracket. An air sampler, a temperature and humidity sensor, and an identification and storage center are fixedly connected to the middle of the bracket. The Internet of Things data transmission module is fixedly connected to the bracket. A wind sensor is fixedly connected to the middle of the crossbeam. The environmental noise collector, the variable-direction camera, the sound source locator, the air sampler, the temperature and humidity sensor, and the Internet of Things data transmission module are all signal-connected to the identification and storage center.
[0023] The beneficial effects of the basic solution are as follows: 1. Through the carefully designed structures such as the bracket and the crossbeam in the data acquisition unit, various monitoring devices such as the environmental noise collector, the variable-direction camera, the sound source locator, the air sampler, and the temperature and humidity sensor are organically integrated, achieving the synchronous and efficient acquisition of multiple environmental parameters. This integrated design not only saves space resources but also improves the efficiency and accuracy of data acquisition, providing a solid foundation for subsequent data analysis and environmental management.
[0024] 2. The variable-direction camera can flexibly adjust the shooting direction to achieve full coverage of the monitoring area, helping to capture key images of environmental noise and chemical pollution incidents. The sound source locator can accurately locate the position of the noise source, providing an important basis for the control and treatment of noise pollution. The air sampler and the temperature and humidity sensor can real-time monitor the pollutant concentration and temperature and humidity changes in the atmosphere, providing key data for evaluating environmental quality.
[0025] 3. The identification and storage center can receive and process data from each monitoring device in real time, convert it into a standardized feature vector, and facilitate subsequent data analysis. The Internet of Things data transmission module realizes the real-time upload and sharing of data through the Internet of Things technology, ensuring the timeliness and accuracy of the data.
[0026] Furthermore, the IoT data transmission module uses NB-IoT technology to transmit environmental pollution data to the big data analysis center.
[0027] The beneficial effects of the basic solution are as follows: 1. With its characteristics of low power consumption, wide coverage, and large connection, NB-IoT technology provides a strong guarantee for the stable transmission of environmental pollution data. Even in remote or poorly-signaled areas, it can ensure the accurate and timely upload of data. Compared with traditional communication technologies, NB-IoT has a significant improvement in data transmission efficiency, reducing data transmission latency and loss, and enhancing data integrity and availability.
[0028] 2. NB-IoT technology adopts a low-power design, significantly reducing the energy consumption of the IoT data transmission module, and extending the operation time and service life of the monitoring equipment. This is particularly important for long-term and continuous environmental monitoring work. Lower energy consumption also means less maintenance cost and a more environmentally friendly monitoring method, which conforms to the concept of sustainable development.
[0029] 3. NB-IoT technology incorporates various security mechanisms, such as data encryption and identity authentication, ensuring the security and privacy protection during data transmission. This is particularly important for data transmission containing sensitive environmental information. Through IoT authentication technology, the authenticity and credibility of data transmission are further enhanced, avoiding the risk of data being tampered with or forged.
[0030] Furthermore, the convolutional neural network uses the image data and noise soundprint information captured after environmental noise sound source localization to identify the sound source type in the image data and integrate the sound source type data into the environmental pollution data.
[0031] The beneficial effects of the basic solution are as follows: 1. Through deep learning algorithms, the convolutional neural network can automatically extract key features from image data and noise soundprints, achieving high-precision identification of sound source types. This identification method has higher accuracy and stability compared with traditional manual identification methods.
[0032] 2. The processing speed of the convolutional neural network is very fast, capable of completing the processing and identification of a large amount of image data and noise soundprints in a short time. This helps the environmental monitoring system respond promptly to noise pollution incidents and take corresponding measures for intervention and treatment.
[0033] 3. By integrating the sound source type data into the environmental pollution data, the environmental pollution data becomes more comprehensive and accurate. This integration not only helps to understand the sources and types of noise pollution but also provides a scientific basis for subsequent environmental management and decision-making.
[0034] 4. The integrated environmental pollution data can support multi-dimensional analysis, such as time, space, type, etc. This analysis method helps to deeply explore the potential laws and trends in environmental pollution data, providing more refined support for environmental management.
[0035] Furthermore, the environmental pollution data also includes the sound source type data identified by the convolutional neural network, including the sound source object type, sound source object model, and voiceprint information data.
[0036] The beneficial effects of the basic solution are as follows: 1. The convolutional neural network can accurately identify the sound source type in environmental noise, including the sound source object type (such as vehicles, machinery, human voices, etc.) and model (such as specific vehicle models, machinery models, etc.), which provides more refined data support for environmental noise management. As a supplement to sound source identification, the voiceprint information data can further improve the accuracy and reliability of identification. For example, in a specific scenario, the voiceprint information can help distinguish different types of vehicles or machinery, providing strong support for the precise positioning of the noise source.
[0037] 2. The addition of sound source type data makes the environmental pollution data cover multiple dimensions such as noise, air quality, temperature and humidity, wind force, and sound source identification. These data can be correlated and verified with each other, providing more comprehensive and accurate information support for environmental management and decision-making. The integration of multi-dimensional data not only enhances the comprehensiveness of the data but also improves the value of the data. For example, by combining noise data and sound source type data, the contribution degree of different sound sources to noise pollution can be analyzed, providing a scientific basis for noise control.
[0038] Furthermore, the big data analysis center includes a database, which is used to classify and store the environmental pollution data uploaded by the Internet of Things data transmission module.
[0039] Furthermore, data preprocessing includes cleaning, filtering, and converting the environmental pollution data into standardized feature vectors.
[0040] The beneficial effects of the basic solution are as follows: 1. Data cleaning can eliminate problems such as missing values, outliers, and duplicate values in the original data, thereby improving the accuracy and integrity of the data. This is the basis for subsequent analysis and modeling, ensuring the reliability and effectiveness of the analysis results. During the cleaning process, data from different sources and in different formats can be uniformly processed, making the data consistent in format, unit, etc., facilitating subsequent data analysis and comparison.
[0041] 2. Through data filtering, irrelevant or redundant features can be removed, reducing the dimensionality of the data, thereby decreasing the computational amount and storage requirements, and improving the efficiency of data analysis. After filtering out the irrelevant features, the model can focus more on the features related to the target variable, thus improving the accuracy and generalization ability of the model. This helps reduce the risk of overfitting and enhance the stability and reliability of the model.
[0042] 3. Feature vector standardization can scale the data of different features to the same numerical range, thereby eliminating the data differences caused by the dimensionality differences between different features. This helps the stability and accuracy of subsequent algorithms. When training a model using a gradient-based optimization algorithm, the standardization of the feature vector can accelerate the convergence speed of the optimization algorithm, improving the efficiency and performance of model training. The standardized features have similar scales and distributions, which makes the comparison between features fairer and more accurate. This helps subsequent steps such as feature selection, feature extraction, and model evaluation.
[0043] Furthermore, the construction of the long short-term memory network includes,
[0044] Construction of hidden layer units, which is used to encode the input feature vector to capture the long-distance dependencies in the encoded sequence;
[0045] Construction of an input gate, which is used to select the current input and the hidden layer units of the previous time step to update the hidden layer;
[0046] Construction of a forget gate, which is used to select the hidden layer state of the previous time step to discard invalid information;
[0047] Construction of an output gate, which is used to select the hidden layer to generate the output encoded sequence.
[0048] The beneficial effects of the basic solution are as follows: 1. Through its special hidden layer unit design, LSTM can effectively capture the long-distance dependencies in sequence data. This is difficult for traditional RNNs because RNNs are prone to the problems of vanishing gradients or exploding gradients when processing long sequences, resulting in the inability to learn long-term dependencies. The hidden layer units of LSTM can selectively retain and transmit information through the control of the input gate, forget gate, and output gate. This mechanism enables LSTM to maintain the integrity and coherence of information when processing sequence data, thereby improving the accuracy of data processing.
[0049] 2. Through a refined gating mechanism, LSTM can control the inflow and outflow of information, thereby avoiding the interference of irrelevant information. This helps reduce the risk of overfitting and improve the generalization ability of the model. The gating mechanism of LSTM enables the model to converge to the optimal solution faster during training. At the same time, since LSTM can process long sequence data, it can process more information in one training, thus improving the training efficiency. The fine control of information by LSTM through the gating mechanism enables the model to maintain stability when processing complex sequence data. This helps prevent the model from experiencing performance degradation or collapse when processing long sequences.
[0050] 3. The gating mechanism of LSTM (input gate, forget gate, output gate) has clear physical meanings, making the behavior of the model more interpretable. This helps understand how the model processes input data and generates outputs. Since the gating mechanism of LSTM is explicit, it is easier to locate problems and take targeted measures for improvement when debugging and optimizing the model.
[0051] Furthermore, the optimization of the long short-term memory network is to use the backpropagation algorithm to minimize the loss function of the network.
[0052] The beneficial effects of the basic solution are as follows: 1. The backpropagation algorithm guides the update direction of the weights by calculating the gradient of the loss function with respect to the network weights. In LSTM, this mechanism ensures that the network can efficiently learn the complex mapping relationship between the input sequence and the output sequence, thereby improving the training efficiency and accuracy of the model. Since the backpropagation algorithm can accurately guide the weight update, the LSTM network can converge to the optimal solution faster during training, thus reducing the training time.
[0053] 2. During the training process, by minimizing the loss function, the backpropagation algorithm not only focuses on the performance on the training set but also indirectly controls the generalization ability of the model on unseen data. Through reasonable regularization strategies (such as L2 regularization, Dropout, etc.), overfitting of the LSTM network can be further prevented, improving its performance on the test set. The backpropagation algorithm enables the LSTM network to better learn the inherent laws of the input data, thus showing stronger robustness in the face of anomalies such as noise and missing values.
[0054] Furthermore, the visual display includes creating time series charts, block pollution levels, and diffusion heat maps.
[0055] The beneficial effects of the basic solution are as follows: 1. The time series chart can intuitively display the changing trend of data over time, helping managers quickly capture key information and abnormal fluctuations in the data. Through the time series chart, managers can more easily analyze characteristics such as periodicity and seasonality in the data, providing strong support for decision-making.
[0056] 2. The block pollution level simplifies complex pollution data into easy-to-understand block levels, facilitating managers to quickly understand the pollution status of different regions. Through the display of the block pollution level, managers can intuitively compare the pollution degrees of different regions, providing important references for environmental protection and governance.
[0057] 3. The heat map represents the density or intensity of data through the depth of color, and can intuitively display the spatial distribution characteristics of the data. In environmental pollution monitoring, the diffusion heat map can clearly show the diffusion range and intensity of pollutants, helping managers quickly locate pollution sources and take corresponding treatment measures.
[0058] 4. Visual display converts complex data into intuitive charts and images, providing clear and accurate data support for decision-makers. Decision-makers can make more scientific and reasonable decisions based on this data, improving the accuracy and effectiveness of decision-making. Through the real-time monitoring of the time series chart and the diffusion heat map, users can promptly discover abnormal fluctuations and potential risks in the data. This provides timely warning information for environmental protection and governance, helping to take measures in advance to prevent environmental pollution incidents. BRIEF DESCRIPTION OF THE DRAWINGS
[0059] Figure 1 It is a schematic diagram of the device for joint detection of environmental noise and chemical pollution and comprehensive big data analysis in the embodiment of the present invention.
[0060] Figure 2 It is a schematic diagram of the data acquisition unit in the embodiment of the present invention.
[0061] Figure 3 It is a schematic diagram of the big data analysis center in the embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION
[0062] The following is a more detailed description through specific embodiments:
[0063] Embodiment 1
[0064] Basically as shown in Figure 1 、 Figure 2 and Figure 3 : The device for joint detection of environmental noise and chemical pollution and comprehensive big data analysis includes:
[0065] The data acquisition department is used to collect environmental pollution data and image data in the set points. It uses a trained convolutional neural network to confirm pollution events in chronological order and store them. The convolutional neural network uses the image data and noise soundprint information captured after environmental noise sound source localization to identify the sound source type in the image data and integrate the sound source type data into the environmental pollution data.
[0066] The environmental pollution data includes: the noise decibels, noise duration, and excessive noise recordings collected by the environmental noise collector; the concentrations of sulfur dioxide, nitrogen dioxide, carbon monoxide, ozone, PM10, and PM2.5 obtained after sampling by the air sampler; the air temperature and humidity data collected by the temperature and humidity sensor; the wind direction and wind speed data collected by the wind sensor.
[0067] The Internet of Things data transmission module is used to ensure data authenticity through Internet of Things authentication and upload the data collected and stored by the data acquisition department within a preset time period to the big data analysis center. The Internet of Things data transmission module uses NB-IoT technology to transmit the environmental pollution data to the big data analysis center.
[0068] The big data analysis center is used to classify and store the data collected by the data acquisition department within a preset range and perform subsequent analysis and processing on the data. The big data analysis center includes a database, which is used to classify and store the environmental pollution data uploaded by the Internet of Things data transmission module.
[0069] The data preprocessing module is used to clean and filter the collected environmental pollution data and convert the preprocessed data into standardized feature vectors.
[0070] The data analysis module is used to construct a long short-term memory network to capture long-distance and short-distance dependencies in the data vector sequence, and optimize and use the feature vectors to train the long short-term memory network. The construction of the long short-term memory network includes: the construction of hidden layer units, which is used to encode the input feature vectors to capture long-distance dependencies in the encoded sequence; the construction of the input gate, which is used to select the current input and the hidden layer units at the previous time step to update the hidden layer; the construction of the forget gate, which is used to select the hidden layer state at the previous time step and discard invalid information; the construction of the output gate, which is used to select the hidden layer to generate the output encoded sequence. The optimization of the long short-term memory network is to use the backpropagation algorithm to minimize the loss function of the network.
[0071] The model prediction module is used to use the trained long short-term memory network to predict the change trend of other environmental pollution data based on the change of a single vector in the environmental pollution data.
[0072] A visualization display module is used to read geographical data and, in combination with the predicted trend of environmental pollution data changes, perform visualization displays including creating time series charts, block pollution levels, and diffusion heat maps. The obtained charts are used to reveal the potential connections between changes in environmental pollution data.
[0073] The specific implementation process is as follows: In the urban scenario, the pollution of vehicle exhaust to the air is based on vehicle driving, and noise is often generated during vehicle driving, including mechanical noise and horn noise. In addition, the heat dissipated by vehicle driving also causes thermal pollution to the city, resulting in the urban heat island effect, which in turn affects the climate and environment of the city. The research on the potential statistical connections among these three types of pollution based on neural networks is still relatively lacking.
[0074] When driving a car through a street, whether it is honking or generating relatively loud mechanical noise, it will be collected by the data acquisition department. At the same time, the data acquisition department will collect other environmental pollution data, including the concentration of atmospheric basic chemical pollutants and temperature change data. The convolutional neural network will judge the type of noise source based on the environmental noise pollution data and image data.
[0075] As Figure 3 shown, all these data are stored in the data acquisition department, and the Internet of Things data transmission module uses NB-IoT technology to transmit the data of multiple data acquisition departments to the big data analysis center for storage and subsequent analysis. After the data enters the big data analysis center, it will first be preprocessed by the data preprocessing module. Among them, data cleaning can use the average value or median value to replace the missing values in the original data and eliminate outliers and duplicate values, thereby improving the accuracy and integrity of the data. The data of different sources and different formats are uniformly processed, so that the data are consistent in format, unit, etc., which is convenient for subsequent data analysis and comparison. Data filtering can remove irrelevant or redundant features, reduce the dimension of the data, thereby reducing the computational amount and storage requirements, and improving the efficiency of data analysis. After filtering out the irrelevant features, the model can focus more on the features related to the target variable, thereby improving the accuracy and generalization ability of the model. Feature vector standardization can scale the data of different features to the same numerical range, thereby eliminating the data differences caused by the dimensional difference between different features. When training a model using a gradient-based optimization algorithm, the standardization of feature vectors can accelerate the convergence speed of the optimization algorithm and improve the efficiency and performance of model training.
[0076] The pre - processed environmental pollution data enters the long - short - term memory network for training. The main purpose is to explore the relationships among environmental noise data, atmospheric pollutant concentration data, and temperature change data within a region, as well as the influence of sound source types on atmospheric pollutant concentration data and temperature change data. The trained long - short - term memory network can relatively reliably predict the noise pollution, air pollution, and heat pollution caused by the passing times of regional vehicle types. These data records and prediction results can be visually displayed as time - series charts to show the changing trends of data records over time, helping managers quickly and accurately capture the key information and fluctuations of the data, and summarize the periodic changes of the data, so as to master the pollution cycle rate and formulate prevention and control measures. For example, the atmospheric chemical pollution is severe in winter, and license plate restrictions can be implemented before winter to reduce the emissions of atmospheric chemical pollution and reduce the pollution damage in winter; for example, if the appearance of a certain vehicle's sound source is strongly correlated with the fluctuations of air pollution, managers can restrict this type of vehicle from entering the city to avoid causing greater pollution to the urban environment. It can be made into a block pollution level, simplifying complex pollution data into easy - to - understand block levels, facilitating managers to quickly understand the pollution conditions in different regions and prioritize the rectification of blocks with high pollution levels and pollution types. It can be made into a heat map, using the depth of color to represent the density or intensity of the data, which can intuitively display the spatial distribution characteristics of the data and clearly show the diffusion range and intensity of pollutants, helping managers locally understand the possible diffusion trends and influence ranges of these pollutions. For example, the diffusion heat map can combine geographical and wind factors to simulate the flow direction and diffusion range of atmospheric chemical pollution, so as to make a prevention and control plan in advance and reduce the impact of pollution on the physical health of urban residents.
[0077] Embodiment 2
[0078] The difference from the above - mentioned embodiment is that as shown in the attached Figure 1 、 Figure 2 and Figure 3 : The data acquisition unit includes a bracket. One side at the top of the bracket is vertically and fixedly connected with a cross - beam. An environmental noise collector is fixedly connected to the top of the bracket. A variable - direction camera is fixedly connected below the end of the cross - beam far from the bracket. A sound source locator is fixedly connected above the end of the cross - beam far from the bracket. An air sampler, a temperature and humidity sensor, and an identification and storage center are fixedly connected to the middle of the bracket. The Internet of Things data transmission module is fixedly connected to the bracket. A wind sensor is fixedly connected to the middle of the cross - beam. The environmental noise collector, the variable - direction camera, the sound source locator, the air sampler, the temperature and humidity sensor, and the Internet of Things data transmission module are all signal - connected to the identification and storage center.
[0079] The environmental pollution data also includes sound source type data identified by a convolutional neural network, including sound source object type, sound source object model, and voiceprint information data.
[0080] The specific implementation process is as follows: When the data acquisition department collects data, the main atmospheric chemical pollutants in automobiles are nitrogen dioxide, carbon monoxide, and PM2.5. When the environmental noise collector detects excessive noise, the sound source locator detects the position of the sound source, and then controls the variable-direction camera to turn and aim at the sound source for shooting. The obtained image data is first preprocessed, and the convolutional neural network is used to identify the type of the sound source, including sound source objects such as screaming people, passing trucks, honking cars, and colliding electric vehicles, and the specific types and models of automobiles can be classified, so as to obtain relatively comprehensive and detailed sound source type data, laying a solid data foundation for the subsequent prediction of the correlation between the sound source, atmospheric chemical pollution, and heat pollution.
[0081] As Figure 2 shown, in addition, the data acquisition department will also collect air temperature, air humidity, wind speed, and wind direction. From the temperature change data, the long short-term memory network can analyze and predict the correlation between the sound source type and urban heat pollution, and air humidity, wind speed, and wind direction are important simulation parameters for drawing heat maps in visual displays. With the support of these data, it is ensured that environmental protection and governance can obtain timely early warning information, which helps to take measures in advance to prevent environmental pollution incidents.
[0082] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the term "including", "comprising", or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article, or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article, or device.
[0083] The above are only embodiments of the present invention. Specific structures and common knowledge such as characteristics that are well-known in the art are not described in detail herein. Those of ordinary skill in the art know all the common general technical knowledge in the technical field to which the invention pertains before the filing date or the priority date, are able to obtain all the prior art in this field, and have the ability to apply conventional experimental means before this date. Those of ordinary skill in the art can, under the inspiration given in this application, combine their own abilities to improve and implement this solution. Some typical well-known structures or well-known methods should not become an obstacle for those of ordinary skill in the art to implement this application. It should be noted that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can still be made, and these should also be regarded as the protection scope of the present invention, and these will not affect the implementation effect of the present invention and the practicality of the patent. The protection scope claimed in this application shall be subject to the content of its claims, and the specific implementation manners and the like recorded in the specification can be used to interpret the content of the claims.
Claims
1. An environmental noise and chemical pollution joint detection and big data comprehensive analysis device, characterized in that: Including: A data acquisition unit, which is used to set points to collect environmental pollution data and image data in the area, and use a trained convolutional neural network to identify and confirm pollution events based on the environmental pollution data and image data in chronological order, and obtain and store sound source type data; The environmental pollution data includes: the noise decibel, noise duration, and excessive noise recording collected by the environmental noise collector; the concentrations of sulfur dioxide, nitrogen dioxide, carbon monoxide, ozone, PM10, and PM2.5 obtained after sampling by the air sampler; the air temperature and humidity data collected by the temperature and humidity sensor; the wind direction and wind speed data collected by the wind sensor; An Internet of Things data transmission module, which is used to ensure data authenticity through Internet of Things authentication, and upload the data collected and stored by the data acquisition unit within a preset time period to the big data analysis center; The big data analysis center is used to classify and store the data collected by the data acquisition unit within a preset range, and perform subsequent analysis and processing on the data; A data preprocessing module, which is used to preprocess the collected data and convert the preprocessed data into a standardized feature vector; A data analysis module, which is used to construct a long short-term memory network to capture long-distance and short-distance dependencies in the data vector sequence, and optimize and use the feature vector to train the long short-term memory network; A model prediction module, which is used to use the trained long short-term memory network to predict the change trend of other environmental pollution data based on the change of a single vector in the environmental pollution data; A visualization display module, which is used to read geographical data, combine the predicted change trend of environmental pollution data for visualization display, and the obtained chart is used to reveal the potential connection between the changes in environmental pollution data.
2. The combined detection and big data comprehensive analysis device for environmental noise and chemical pollution according to claim 1, characterized in that: The data acquisition unit includes a bracket. One side of the top of the bracket is vertically fixedly connected with a cross beam. The top of the bracket is fixedly connected with an environmental noise collector. A variable-direction camera is fixedly connected below the end of the cross beam far from the bracket. A sound source locator is fixedly connected above the end of the cross beam far from the bracket. An air sampler, a temperature and humidity sensor, and an identification and storage center are fixedly connected in the middle of the bracket. The Internet of Things data transmission module is fixedly connected with the bracket. A wind sensor is fixedly connected in the middle of the cross beam. The environmental noise collector, the variable-direction camera, the sound source locator, the air sampler, the temperature and humidity sensor, and the Internet of Things data transmission module are all signal-connected to the identification and storage center.
3. The combined detection and big data comprehensive analysis device for environmental noise and chemical pollution according to claim 2, characterized in that: The Internet of Things data transmission module uses NB-IoT technology to transmit environmental pollution data to the big data analysis center.
4. The combined detection and big data comprehensive analysis device for environmental noise and chemical pollution according to claim 1, wherein: The convolutional neural network uses the image data and noise soundprint information taken after environmental noise sound source localization to identify the sound source type in the image data, and integrates the sound source type data into the environmental pollution data.
5. The environmental noise and chemical pollution joint detection and big data comprehensive analysis device according to claim 1, characterized in that: The environmental pollution data also includes the sound source type data identified by the convolutional neural network, including the sound source object type, sound source object model, and soundprint information data.
6. The combined detection and big data comprehensive analysis device for environmental noise and chemical pollution according to claim 1, wherein: The big data analysis center includes a database, which is used to classify and store the environmental pollution data uploaded by the Internet of Things data transmission module.
7. The combined detection and big data comprehensive analysis device for environmental noise and chemical pollution according to claim 1, wherein: Data preprocessing includes cleaning, filtering, and converting the environmental pollution data into a standardized feature vector.
8. The combined detection and big data comprehensive analysis device for environmental noise and chemical pollution according to claim 1, wherein: The construction of the long short-term memory network includes, Hidden layer unit construction, which is used to encode the input feature vector to capture long-range dependencies in the encoded sequence; Input gate construction, which is used to select the current input and the hidden layer units at the previous time step to update the hidden layer; Forget gate construction, which is used to select the hidden layer state at the previous time step to discard invalid information; Output gate construction, which is used to select the hidden layer to generate the output encoded sequence.
9. The combined detection and big data comprehensive analysis device for environmental noise and chemical pollution according to claim 1, wherein: The optimization of the long short-term memory network is to use the backpropagation algorithm to minimize the loss function of the network.
10. The combined detection and big data comprehensive analysis device for environmental noise and chemical pollution according to claim 1, characterized in that: Visualization includes creating time series charts, block pollution levels, and diffusion heat maps.