Soil pollution detection method, device and medium based on Internet of Things data fusion

Abstract

The soil pollution detection method, device and medium based on the Internet of Things data fusion disclosed in the present invention obtain multi-source data related to soil pollution in the area to be detected; use a preset soil environment fusion algorithm to perform end-to-end fusion and prediction on the multi-source data to obtain estimated data of the soil pollution degree; use a preset spatial interpolation algorithm to infer unobserved data to obtain soil pollution assessment data; use a pulse neural network model to fuse the estimated data with the assessment data to obtain comprehensive soil pollution detection data; use geographic information system technology to combine the detection data with the geographic space information of the area to be detected, perform geographic space analysis, and obtain spatial analysis data; input the soil pollution detection data and the spatial analysis data into a pre-built health risk assessment model to calculate the risk level of soil pollution. Through this method, the potential risk of urban soil pollution to human health can be evaluated efficiently and accurately.

Description

Technical Field

[0001] The present invention relates to the field of soil monitoring, and in particular to a soil pollution detection method, device and medium based on Internet of Things data fusion. Background Art

[0002] Urban soil pollution poses a serious threat to human health and environmental safety. Soil pollution not only causes excessive heavy metals and organic matter in crops and food, but also poses a risk of direct contact and inhalation to the human body through dust from soil particles and sediments, floods, etc. Therefore, it is of great significance to accurately assess urban soil pollution and analyze health risks. Traditional urban soil pollution assessment methods are usually based on field sampling and experimental analysis, which have problems such as limited sampling points, large workload, high cost, and long time. Although modern monitoring technologies such as sensor networks, remote sensing technology, and geographic information systems (GIS) can provide a large amount of soil pollution data, these data come from different sources, the data quality varies, and data fusion and analysis methods have not been well solved.

[0003] In recent years, the rapid development of Internet of Things technology has provided new opportunities for urban soil pollution assessment. Internet of Things technology can obtain data from various sensors in real time and transmit the data to the cloud platform through wireless communication for centralized management and analysis. Different types of sensors such as heavy metal sensors, organic matter sensors, temperature and humidity sensors, etc. can provide information on multiple aspects of soil pollution. However, these data often have problems such as differences, inaccuracies and inconsistencies, and effective data fusion and analysis methods are needed to integrate them into comprehensive assessment results. Geographic Information System (GIS) technology also plays an important role in soil pollution assessment. GIS can combine soil pollution data with geospatial information for geospatial analysis. By combining soil pollution data with geographic terrain data, soil type data, etc., the spatial distribution characteristics and potential risks of soil pollution can be better analyzed, providing a scientific basis for soil pollution control and environmental protection.

[0004] In summary, the urban soil pollution health risk assessment method based on IoT data fusion and GIS technology has important theoretical and practical value. Through effective data fusion and analysis methods, the accuracy and scientificity of soil pollution assessment can be improved, providing strong support for protecting human health and urban environmental safety. At the same time, through the application of GIS technology, the spatial distribution characteristics and risk level of soil pollution can be better analyzed, providing a scientific basis for soil pollution control. Therefore, the development of soil pollution detection methods based on IoT data fusion has important research significance and application prospects. Summary of the invention

[0005] To solve the above technical problems, the present invention relates to a soil pollution detection method, device and medium based on Internet of Things data fusion. The method can efficiently and accurately assess the potential risks of urban soil pollution to human health.

[0006] The embodiment of the present invention provides a soil pollution detection method based on Internet of Things data fusion, the method comprising the following steps:

[0007] Acquire multi-source data related to soil pollution in the area to be detected, wherein the multi-source data includes different types of sensor data and monitoring site data;

[0008] Using a preset soil environment fusion algorithm to perform end-to-end fusion and prediction on the multi-source data to obtain estimated data of the soil pollution degree;

[0009] Using a preset spatial interpolation algorithm to infer unobserved data from the multi-source data to obtain soil pollution assessment data;

[0010] The estimated data and the evaluation data are merged by a preset spiking neural network model to obtain comprehensive soil pollution detection data;

[0011] Combining the detection data with the geographic spatial information of the area to be detected by using geographic information system technology, performing geographic spatial analysis, and obtaining spatial analysis data;

[0012] The soil pollution detection data and the spatial analysis data are input into a pre-built health risk assessment model to calculate the risk level of soil pollution.

[0013] Preferably, the use of a preset soil environment fusion algorithm to perform end-to-end fusion and prediction on the multi-source data to obtain estimated data of the soil pollution degree specifically includes:

[0014] Preprocessing the sensor data, including data cleaning, missing value processing and abnormal data removal;

[0015] Adaptive fusion method is used to fuse the data of different sensors to obtain fused data of soil pollution degree;

[0016] The fused data is predicted based on a pre-trained deep learning model to obtain estimated data on the degree of soil contamination.

[0017] As a preferred solution, the multi-source data is extrapolated with a preset spatial interpolation algorithm to obtain the soil pollution assessment data, which specifically includes:

[0018] Preprocessing the monitoring site data, including data cleaning and outlier removal;

[0019] Using geographic information system technology, spatial interpolation is performed based on the spatial distribution characteristics of monitoring sites to estimate the assessment data of unobserved soil pollution;

[0020] Among them, the data at the unobserved position u in the evaluation data Among them, Z(i) represents the value of the observed data point i, w(i) represents the weight of the data point i, n represents the number of observed data points, and W represents the sum of weighting coefficients.

[0021] Preferably, the estimation data and the evaluation data are fused by a preset spiking neural network model to obtain comprehensive soil pollution detection data, specifically including:

[0022] Performing feature extraction and correlation analysis on the estimated data and the evaluation data, extracting statistical features, frequency domain features and time domain features, obtaining feature data, and determining the correlation between the features to determine a correlation coefficient matrix;

[0023] The characteristic data are fused using a pre-trained pulse neural network model to obtain comprehensive soil pollution risk assessment data;

[0024] Based on the correlation coefficient matrix, the risk assessment data is correlated and calculated to obtain comprehensive soil pollution detection data.

[0025] As a preferred solution, the soil pollution detection data is combined with the geographic space information of the area to be detected by using geographic information system technology to perform geographic space analysis to obtain spatial analysis data, including:

[0026] Associating the soil pollution detection data with the geographic terrain data and soil type data in the geographic spatial information;

[0027] GIS software is used to perform spatial statistical analysis and spatial interaction analysis on the associated data to obtain spatial analysis data containing the spatial distribution characteristics of soil pollution.

[0028] Preferably, the step of inputting the soil pollution detection data and the spatial analysis data into a pre-built health risk assessment model to calculate the risk level of soil pollution includes:

[0029] Calculate the soil pollution degree index, human contact pathway index and comprehensive impact coefficient respectively according to the soil pollution detection data and the spatial analysis data;

[0030] Inputting the soil pollution degree index, the human contact pathway index and the comprehensive impact coefficient into the health risk assessment model to calculate the risk index;

[0031] Determining the risk level according to the risk index;

[0032] Among them, the risk index R = (w1*S+w2*E+w3*T) / (w4*S+w5*E+w6*T+ε), w1~w6 are preset weight coefficients, ε is a preset threshold; the soil pollution degree index coefficient The human exposure pathway indicator coefficient The comprehensive influence coefficient S j is the jth soil pollution index, W j is the weight of the jth soil pollution index, m is the number of soil pollution indexes, E j is the jth human contact pathway indicator, η j is the weight of the jth human contact pathway indicator, p is the number of human contact pathway indicators, T j is the exposure time index of the jth toxicity index, σ j is the weight of the j-th toxicity index exposure time, and q is the number of toxicity index exposure time indicators.

[0033] Preferably, determining the risk level according to the risk index specifically includes:

[0034] When the risk index is not higher than a preset first threshold, determining the risk level as low risk;

[0035] When the risk index is higher than the first threshold value and not higher than a preset second threshold value, determining that the risk level is medium risk;

[0036] When the risk index is higher than the second threshold, determining the risk level as high risk;

[0037] The second threshold is higher than the first threshold.

[0038] The embodiment of the present invention further provides a soil pollution detection device based on Internet of Things data fusion, the device comprising:

[0039] An acquisition module, used to acquire multi-source data related to soil pollution in the area to be detected, wherein the multi-source data includes different types of sensor data and monitoring site data;

[0040] An estimation module, used to perform end-to-end fusion and prediction of the multi-source data using a preset soil environment fusion algorithm to obtain estimated data of the soil pollution degree;

[0041] An inference module, used to infer unobserved data using a preset spatial interpolation algorithm on the multi-source data to obtain soil pollution assessment data;

[0042] A fusion module, used to fuse the estimated data with the evaluation data through a preset spiking neural network model to obtain comprehensive soil pollution detection data;

[0043] A spatial analysis module, used to combine the detection data with the geographic spatial information of the area to be detected by using geographic information system technology, perform geographic spatial analysis, and obtain spatial analysis data;

[0044] The level calculation module is used to input the soil pollution detection data and the spatial analysis data into a pre-built health risk assessment model to calculate the risk level of soil pollution.

[0045] Preferably, the estimation module is specifically used for:

[0046] Preprocessing the sensor data, including data cleaning, missing value processing and abnormal data removal;

[0047] Adaptive fusion method is used to fuse the data of different sensors to obtain fused data of soil pollution degree;

[0048] The fused data is predicted based on a pre-trained deep learning model to obtain estimated data on the degree of soil contamination.

[0049] Preferably, the calculation module is specifically used for:

[0050] Preprocessing the monitoring site data, including data cleaning and outlier removal;

[0051] Using geographic information system technology, spatial interpolation is performed based on the spatial distribution characteristics of monitoring sites to estimate the assessment data of unobserved soil pollution;

[0052] Among them, the data at the unobserved position u in the evaluation data Among them, Z(i) represents the value of the observed data point i, w(i) represents the weight of the data point i, n represents the number of observed data points, and W represents the sum of weighting coefficients.

[0053] As a preferred solution, the fusion module is specifically used for:

[0054] Performing feature extraction and correlation analysis on the estimated data and the evaluation data, extracting statistical features, frequency domain features and time domain features, obtaining feature data, and determining the correlation between the features to determine a correlation coefficient matrix;

[0055] The characteristic data are fused using a pre-trained pulse neural network model to obtain comprehensive soil pollution risk assessment data;

[0056] Based on the correlation coefficient matrix, the risk assessment data is correlated and calculated to obtain comprehensive soil pollution detection data.

[0057] Preferably, the spatial analysis module is specifically used for:

[0058] Associating the soil pollution detection data with the geographic terrain data and soil type data in the geographic spatial information;

[0059] GIS software is used to perform spatial statistical analysis and spatial interaction analysis on the associated data to obtain spatial analysis data containing the spatial distribution characteristics of soil pollution.

[0060] Preferably, the level calculation module is specifically used for:

[0061] Calculate the soil pollution degree index, human contact pathway index and comprehensive impact coefficient respectively according to the soil pollution detection data and the spatial analysis data;

[0062] Inputting the soil pollution degree index, the human contact pathway index and the comprehensive impact coefficient into the health risk assessment model to calculate the risk index;

[0063] Determining the risk level according to the risk index;

[0064] Among them, the risk index R = (w1*S+w2*E+w3*T) / (w4*S+w5*E+w6*T+ε), w1~w6 are preset weight coefficients, ε is a preset threshold; the soil pollution degree index coefficient The human exposure pathway indicator coefficient The comprehensive influence coefficient S j is the jth soil pollution index, W j is the weight of the jth soil pollution index, m is the number of soil pollution indexes, E j is the jth human contact pathway indicator, η j is the weight of the jth human contact pathway indicator, p is the number of human contact pathway indicators, T j is the exposure time index of the jth toxicity index, σ j is the weight of the j-th toxicity index exposure time, and q is the number of toxicity index exposure time indicators.

[0065] Furthermore, the level calculation module is specifically used for:

[0066] When the risk index is not higher than a preset first threshold, determining the risk level as low risk;

[0067] When the risk index is higher than the first threshold value and not higher than a preset second threshold value, determining that the risk level is medium risk;

[0068] When the risk index is higher than the second threshold, determining the risk level as high risk;

[0069] The second threshold is higher than the first threshold.

[0070] An embodiment of the present invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the soil pollution detection method based on Internet of Things data fusion as described in any one of the above embodiments.

[0071] An embodiment of the present invention also provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the soil pollution detection method based on Internet of Things data fusion as described in any one of the above embodiments is implemented.

[0072] The embodiment of the present invention provides a soil pollution detection method, device and medium based on Internet of Things data fusion, which obtains multi-source data related to soil pollution in the area to be detected, and the multi-source data includes different types of sensor data and monitoring site data; uses a preset soil environment fusion algorithm to perform end-to-end fusion and prediction on the multi-source data to obtain estimated data of soil pollution degree; uses a preset spatial interpolation algorithm to infer unobserved data on the multi-source data to obtain soil pollution assessment data; fuses the estimated data with the assessment data through a preset pulse neural network model to obtain comprehensive soil pollution detection data; uses geographic information system technology to combine the detection data with the geographic spatial information of the area to be detected, performs geographic spatial analysis, and obtains spatial analysis data; inputs the soil pollution detection data and the spatial analysis data into a pre-built health risk assessment model to calculate the risk level of soil pollution. Through this method, the potential risk of urban soil pollution to human health can be efficiently and accurately assessed.

[0073] Compared with the prior art, the present invention has the following advantages:

[0074] This application proposes a soil pollution detection method based on IoT data fusion. Traditional soil pollution assessment methods mainly rely on field sampling and experimental analysis, which are costly, labor-intensive and time-consuming. The present invention uses IoT technology to obtain multi-source data, and uses the independently innovative SEI soil environment fusion algorithm and SSI spatial interpolation algorithm to perform data fusion and analysis, effectively integrating data from various sensors and monitoring sites, and improving the accuracy and scientificity of the assessment. This assessment method based on IoT data fusion is innovative and practical, and can obtain a large amount of representative soil pollution data in a short period of time, providing a new perspective and method for soil pollution assessment.

[0075] This application proposes an independent innovative algorithm for urban soil pollution assessment. Through the independent innovative SEI soil environment fusion algorithm and SSI spatial interpolation algorithm, the fusion and extrapolation problems of sensor data and monitoring site data are solved. The SEI soil environment fusion algorithm integrates data from different sensors through an adaptive fusion method, and uses a deep learning algorithm to predict the data to obtain a high-precision estimate of the degree of soil pollution. The SSI spatial interpolation algorithm uses geographic information system technology to perform spatial interpolation based on the spatial distribution characteristics of monitoring sites to extrapolate unobserved soil pollution data. These independent innovative algorithms can provide more comprehensive and accurate soil pollution assessment results, and have high application value and creativity.

[0076] This application realizes a comprehensive soil pollution health risk assessment. The present invention uses the independently innovative SNN model to fuse and analyze the data extracted by the SEI soil environment fusion algorithm and the SSI spatial interpolation algorithm to obtain comprehensive soil pollution health risk assessment data. By comprehensively considering indicators such as the degree of soil pollution, human exposure pathways, exposure time and sensitivity, an SHR health risk assessment model was established to accurately calculate the potential risks of urban soil pollution to human health. This comprehensive assessment method takes multiple factors into account and can more comprehensively and accurately assess the health risks of urban soil pollution, providing a scientific basis for formulating effective soil pollution control and environmental protection measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 It is a flow chart of a soil pollution detection method based on Internet of Things data fusion provided by an embodiment of the present invention;

[0078] Figure 2 It is a structural schematic diagram of a soil pollution detection device based on Internet of Things data fusion provided by an embodiment of the present invention;

[0079] Figure 3 It is a structural schematic diagram of a soil pollution detection device based on Internet of Things data fusion provided by another embodiment of the present invention. DETAILED DESCRIPTION

[0080] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0081] The embodiment of the present invention provides a soil pollution detection method based on Internet of Things data fusion, see Figure 1 , is a flow chart of a soil pollution detection method based on IoT data fusion provided by an embodiment of the present invention, wherein steps S1 to S6 of the method are as follows:

[0082] S1, obtaining multi-source data related to soil pollution in the area to be detected, wherein the multi-source data includes different types of sensor data and monitoring site data;

[0083] S2, using a preset soil environment fusion algorithm to perform end-to-end fusion and prediction on the multi-source data to obtain estimated data of the soil pollution degree;

[0084] S3, using a preset spatial interpolation algorithm to infer unobserved data from the multi-source data to obtain soil pollution assessment data;

[0085] S4, fusing the estimated data with the evaluation data through a preset spiking neural network model to obtain comprehensive soil pollution detection data;

[0086] S5, combining the detection data with the geographic spatial information of the area to be detected by using geographic information system technology, performing geographic spatial analysis, and obtaining spatial analysis data;

[0087] S6, inputting the soil pollution detection data and the spatial analysis data into a pre-built health risk assessment model to calculate the risk level of soil pollution.

[0088] In the specific implementation of this embodiment, the Internet of Things technology is used to obtain multi-source data related to urban soil pollution;

[0089] The SEI soil environment fusion algorithm is used to perform end-to-end fusion and prediction of multi-source data to obtain high-precision soil pollution degree estimation results. First, different types of sensor networks are deployed, such as heavy metal sensors, organic matter sensors, temperature and humidity sensors, etc., to monitor different aspects of soil pollution in real time. The self-developed sensor data acquisition SD function is used to realize interface communication and data reading with different types of sensors. The self-designed SWC wireless communication function is used to realize wireless transmission of sensor data and monitoring site data and cloud platform connection. Data processing algorithm functions are written, including verification, correction and fusion functions. Database operation functions are created to realize data storage and query operations. At the same time, monitoring sites are established to collect soil pollution data, environmental parameter data, etc. at the monitoring sites. The self-designed MSC monitoring site data acquisition function is used to realize the collection and transmission functions of site data. Finally, the sensor data and monitoring site data are transmitted to the cloud platform through the SWC wireless communication function for centralized management and analysis.

[0090] Deploy different types of sensor networks according to different aspects of soil pollution, such as heavy metal sensors, organic matter sensors, temperature and humidity sensors, etc. Deploy these sensors in representative areas of the city to ensure that soil pollution in different areas can be fully monitored. Write a sensor data acquisition function (SD) to implement interface communication and data reading with different types of sensors.

[0091] def sensor_data_collection(sensor_type,sensor_params);

[0092] Establishing a communication connection with the sensor;

[0093] sensor_connection=establish_sensor_connection(sensor_type,sensor_params);

[0094] Read data from the sensor:

[0095] sensor_data=read_sensor_data(sensor_connection);

[0096] Verify and preprocess the data:

[0097] validated_sensor_data=validate_and_preprocess_data(sensor_data);

[0098] return validated_sensor_data;

[0099] Design wireless communication function (SWC) to realize wireless transmission of sensor data and monitoring site data and connection to the cloud platform.

[0100] def wireless_communication(sensor_data,site_data,cloud_platform_params);

[0101] Establish a connection with the wireless communication module:

[0102] wireless_connection=establish_wireless_connection();

[0103] Transmit sensor data and site data via wireless communication module:

[0104] data_transmission_status=transmit_data(wireless_connection,sensor_data,site_data);

[0105] Establish a connection with the cloud platform and upload data to the cloud platform:

[0106] cloud_platform_connection=establish_cloud_platform_connection(cloud_platform_params);

[0107] upload_data_to_cloud_platform(cloud_platform_connection,sensor_data,site_data);

[0108] return data_transmission_status;

[0109] Write soil pollution related data processing algorithm functions, including verification, correction and fusion functions.

[0110] def data_processing(sensor_data,site_data);

[0111] Verify and correct sensor data and site data:

[0112] corrected_sensor_data=correct_sensor_data(sensor_data);

[0113] corrected_site_data=correct_site_data(site_data);

[0114] Fusion of sensor data and site data:

[0115] fused_data=data_fusion(corrected_sensor_data,corrected_site_data);

[0116] return fused_data;

[0117] Create soil pollution-related database operation functions to implement data storage and query operations.

[0118] def database_operations(data,db_connection_params);

[0119] Establish a connection to the database:

[0120] db_connection=establish_db_connection(db_connection_params);

[0121] Create a data table:

[0122] create_table(db_connection);

[0123] Insert data into the database:

[0124] insert_data(db_connection,data);

[0125] Query data from the database:

[0126] query_result=query_data(db_connection);

[0127] return query_result;

[0128] Establish monitoring stations at appropriate geographical locations to collect soil pollution data, environmental parameter data, etc. The monitoring stations are selected based on the city's soil pollution characteristics and population distribution. By setting up appropriate monitoring stations, soil pollution data can be collected in real time, taking into account the spatial distribution of urban soil pollution. Design a monitoring station data collection function (MSC) to realize the collection and transmission functions of station data.

[0129] def monitoring_site_data_collection(site_params,data_collection_device_params);

[0130] Establish a connection with the data logger;

[0131] data_collection_device_connectionestablish_device_connection(data_collection_device_params);

[0132] Read data from the data logger:

[0133] site_data=read_data_collection_device_data(data_collection_device_connection,site_params);

[0134] Verify and preprocess the data;

[0135] validated_site_data=validate_and_preprocess_data(site_data);

[0136] return validated_site_data;

[0137] The sensor data and monitoring site data are transmitted to the cloud platform for centralized management and analysis through the SWC wireless communication function. The sensor data and monitoring site data are used as input parameters, and the wireless communication module is used to establish a connection with the cloud platform, and the data is transmitted and uploaded.

[0138] def wireless_communication(sensor_data,site_data,cloud_platform_params);

[0139] Establish a connection with the wireless communication module:

[0140] wireless_connection=establish_wireless_connection();

[0141] Transmit sensor data and site data via wireless communication module:

[0142] data_transmission_status=transmit_data(wireless_connection,sensor_data,site_data);

[0143] Establish a connection with the cloud platform and upload data to the cloud platform:

[0144] cloud_platform_connection=

[0145] establish_cloud_platform_connection(cloud_platform_params);

[0146] upload_data_to_cloud_platform(cloud_platform_connection,sensor_data,site_data);

[0147] return data_transmission_status;

[0148] Through the above innovative implementation steps, multi-source data related to urban soil pollution can be obtained efficiently and accurately.

[0149] The self-developed SEI Soil Environment Integration algorithm is used to perform end-to-end fusion and prediction of multi-source data. First, the sensor data is preprocessed, including data cleaning, missing value processing, and abnormal data removal. Then, the data from different sensors are fused using an adaptive fusion method to obtain a unified soil pollution degree estimation result. Finally, the fused data is predicted based on a deep learning algorithm to obtain a high-precision soil pollution degree estimation result.

[0150] The SSI spatial interpolation algorithm (Soil Spatial Interpolation) is used to estimate and improve the unobserved soil pollution data. First, the monitoring site data is preprocessed, including data cleaning and outlier removal. Then, using geographic information system technology, spatial interpolation is performed based on the spatial distribution characteristics of the monitoring sites to estimate the unobserved soil pollution data.

[0151] The independently innovated SNN model (Soil Neural Network) is used to fuse and analyze the data extracted by the SEI soil environment fusion algorithm and the SSI spatial interpolation algorithm in the first two steps.

[0152] Geographic Information System (GIS) technology is used to combine comprehensive soil pollution health risk assessment data with geographic spatial information for geospatial analysis. Soil pollution data is associated with geographic topography data, soil type data and other geospatial information, and GIS software is used to perform spatial statistical analysis, spatial interactive analysis, etc., to obtain the spatial distribution characteristics of soil pollution and risk assessment results.

[0153] In another embodiment provided by the present invention, the step S2 specifically includes:

[0154] Preprocessing the sensor data, including data cleaning, missing value processing and abnormal data removal;

[0155] Adaptive fusion method is used to fuse the data of different sensors to obtain fused data of soil pollution degree;

[0156] The fused data is predicted based on a pre-trained deep learning model to obtain estimated data on the degree of soil contamination.

[0157] In the specific implementation of this embodiment, the independently innovated SEI soil environment fusion algorithm (Soil Environment Integration) is used to perform end-to-end fusion and prediction of multi-source data. First, the sensor data is preprocessed, including data cleaning, missing value processing and abnormal data removal. Then, the data from different sensors are fused using an adaptive fusion method to obtain a unified soil pollution degree estimation result. Finally, the fused data is predicted based on a deep learning algorithm to obtain a high-precision soil pollution degree estimation result. The specific implementation steps are as follows:

[0158] The SEI soil environment fusion algorithm is an independently innovative algorithm. Assuming there are n sensors, the sensor data are X1, X2, ..., Xn respectively, and the estimated result of soil pollution degree after fusion is Y. The formula of the SEI soil environment fusion algorithm is as follows:

[0159] Y=f(G(X1),G(X2),...,G(Xn));

[0160] Among them, the G function is the feature extraction and fusion calculation function of the sensor data. The input data of each sensor is subjected to feature extraction and fusion to obtain the fused feature representation;

[0161] G(Xi)=FeatureExtractori(Xi), where i∈{1,2,...,n}.

[0162] The feature extraction and fusion function (G) can use convolutional neural network, autoencoder or other network structures as needed;

[0163] The fusion weight calculation function can be defined based on the statistical characteristics of the sensor data and the machine learning model in the form of:

[0164] W=h(Feature1,Feature2,...,Featuren);

[0165] Among them, Feature1, Feature2, ..., Featuren represent the feature matrix of sensor data, and the h function is a weight calculation function, which can be learned and calculated using a machine learning algorithm.

[0166] The following is the implementation process using the SEI soil environment fusion algorithm formula and function:

[0167] Preprocess the sensor data, including data cleaning, missing value processing and abnormal data removal.

[0168] def data_preprocessing(data);

[0169] Data cleaning:

[0170] cleaned_data=data_cleaning(data);

[0171] Missing value handling;

[0172] filled_data=missing_value_imputation(cleaned_data);

[0173] Abnormal data removal:

[0174] filtered_data=outlier_removal(filled_data);

[0175] return filtered_data;

[0176] Adaptive fusion is performed based on the data features and weights of multiple sensors to obtain a unified soil pollution degree estimation result. Adaptive fusion of different sensor data first extracts features, then uses the adaptive fusion method to fuse the feature matrix to obtain a unified soil pollution degree estimation result.

[0177] def adaptive_fusion(data);

[0178] Define the weight calculation function:

[0179] def calculate_weights(data);

[0180] Feature extraction:

[0181] features=feature_extraction(data);

[0182] Weight calculation:

[0183] weights=weight_calculation(features);

[0184] return weights;

[0185] Calculate weights:

[0186] weights=calculate_weights(data);

[0187] Weighted Fusion:

[0188] def fusion(data,weights):

[0189] fused_data=np.average(data,weights=weights,axis=1):

[0190] return fused_data:

[0191] Fusion processing:

[0192] fused_data=fusion(data,weights);

[0193] return fused_data;

[0194] The fused data is predicted based on the deep learning algorithm. First, a deep learning model is built, then the model is trained using the training data, and finally the trained model is used to predict the data to obtain a high-precision estimate of the degree of soil contamination.

[0195] def deep_learning_prediction(data);

[0196] Building a deep learning model:

[0197] model = build_model();

[0198] Define the training function:

[0199] def train_model(model,X_train,y_train,epochs,batch_size);

[0200] model.fit(X_train,y_train,epochs=epochs,batch_size=batch_size);

[0201] return model;

[0202] Train the model:

[0203] trained_model=train_model(model,X_train,y_train,epochs,batch_size);

[0204] Predicting the extent of soil contamination:

[0205] predicted_data=trained_model.predict(data);

[0206] return predicted_data;

[0207] Through the above implementation steps and functions, the independently innovative SEI soil environment fusion algorithm can effectively fuse different sensor data, preprocess the data of multiple sensors, adaptively fuse and predict based on deep learning algorithms, and obtain high-precision soil pollution degree estimation results.

[0208] In another embodiment provided by the present invention, the step S3 specifically includes:

[0209] Preprocessing the monitoring site data, including data cleaning and outlier removal;

[0210] Using geographic information system technology, spatial interpolation is performed based on the spatial distribution characteristics of monitoring sites to estimate the assessment data of unobserved soil pollution;

[0211] Among them, the data at the unobserved position u in the evaluation data Among them, Z(i) represents the value of the observed data point i, w(i) represents the weight of the data point i, n represents the number of observed data points, and W represents the sum of weighting coefficients.

[0212] In the specific implementation of this embodiment, the SSI spatial interpolation algorithm (Soil Spatial Interpolation) is used to infer and improve the unobserved soil pollution data. First, the monitoring site data is preprocessed, including data cleaning and outlier removal. Then, using geographic information system technology, spatial interpolation is performed based on the spatial distribution characteristics of the monitoring sites to infer the unobserved soil pollution data. Combining sensor data and monitoring site data, a more comprehensive soil pollution assessment result is obtained. The specific implementation steps of this step are as follows:

[0213] The SSI spatial interpolation algorithm is an independent and innovative spatial interpolation method used to infer unobserved data.

[0214] The data at the unobserved position u in the evaluation data Among them, Z(i) represents the value of the observed data point i, w(i) represents the weight of the data point i, n represents the number of observed data points, and W represents the sum of weighting coefficients.

[0215] The following is the implementation process using the SSI spatial interpolation algorithm formula and function:

[0216] Collect the observed soil data points around and their corresponding geographic location information. These data include soil properties and environmental factors, such as soil heavy metal content, soil pH, soil moisture, etc.

[0217] Preprocess the monitoring site data, including data cleaning and outlier removal.

[0218] def station_data_preprocessing(data);

[0219] Data cleaning:

[0220] cleaned_data=data_cleaning(data);

[0221] Outlier removal:

[0222] filtered_data=outlier_removal(cleaned_data);

[0223] return filtered_data;

[0224] The SSI spatial interpolation algorithm is used to estimate and improve the unobserved soil pollution data. First, the spatial interpolation model is constructed based on the spatial distribution characteristics of the monitoring sites using geographic information system technology. Then, the model is used to interpolate the unobserved data to obtain the estimated soil pollution data. Finally, the sensor data and monitoring site data are combined to obtain a more comprehensive soil pollution assessment result.

[0225] def spatial_interpolation(station_data,sensor_data);

[0226] Soilspatial_interpolation;

[0227] Build a spatial interpolation model:

[0228] model=build_interpolation_model(station_data);

[0229] Define the interpolation function:

[0230] def interpolate(model,X);

[0231] Use the spatial interpolation model to estimate unobserved data:

[0232] interpolated_data=model.predict(X);

[0233] return interpolated_data;

[0234] Construct input features for unobserved data:

[0235] X_unobserved=create_unobserved_features();

[0236] Perform spatial interpolation to estimate unobserved soil pollution data;

[0237] interpolated_data=interpolate(model,X_unobserved);

[0238] Combining sensor data with monitoring site data;

[0239] combined_data=combine_data(sensor_data,station_data,interpolated_data);

[0240] return combined data;

[0241] The preprocessing function and spatial interpolation function are called to finally return the result of combining sensor data, monitoring site data and estimated soil pollution data.

[0242] def main(sensor_data,station_data);

[0243] Monitoring site data preprocessing;

[0244] preprocessed_station_data=station_data_preprocessing(station_data);

[0245] Spatial interpolation:

[0246] combined_data=spatial_interpolation(preprocessed_station_data,sensor_data);

[0247] return combined_data;

[0248] The SSI spatial interpolation algorithm uses geographic information system technology to perform spatial interpolation based on the spatial distribution characteristics of monitoring sites to infer unobserved soil pollution data. These independently innovative algorithms can provide more comprehensive and accurate soil pollution assessment results and have high application value and creativity.

[0249] In another embodiment provided by the present invention, the step S4 specifically includes:

[0250] Performing feature extraction and correlation analysis on the estimated data and the evaluation data, extracting statistical features, frequency domain features and time domain features, obtaining feature data, and determining the correlation between the features to determine a correlation coefficient matrix;

[0251] The characteristic data are fused using a pre-trained pulse neural network model to obtain comprehensive soil pollution risk assessment data;

[0252] Based on the correlation coefficient matrix, the risk assessment data is correlated and calculated to obtain comprehensive soil pollution detection data.

[0253] In the specific implementation of this embodiment, the independently innovated pulse neural network model SNN model (SoilNeural Network) is used to fuse and analyze the data extracted by the SEI soil environment fusion algorithm and the SSI spatial interpolation algorithm in the first two steps. First, the extracted data is subjected to feature extraction and correlation analysis, and the important features and relationships of the feature data are analyzed. Then, the extracted data is fused using the SNN model algorithm to obtain comprehensive soil pollution health risk assessment data. The specific implementation steps are as follows:

[0254] The SNN model is an independently innovative model, and the algorithm formula is:

[0255] Y = f(g(X1,X2));

[0256] Among them, X1 represents the data extracted by the SEI soil environment fusion algorithm, X2 represents the data obtained by the SSI spatial interpolation algorithm, g represents the process of feature extraction and association analysis, and f represents the activation function of the SNN model.

[0257] The following is the function implementation process of data fusion and correlation analysis using the SNN model:

[0258] The extracted data are subjected to feature extraction and correlation analysis. First, the data extracted by the SEI soil environment fusion algorithm and the SSI spatial interpolation algorithm are used to extract statistical features, frequency domain features, and time domain features, and obtain the numerical values ​​of each feature. Then, the correlation between the features is calculated to obtain the correlation coefficient matrix between the features.

[0259] def feature_extraction(data);

[0260] Extract statistical features (such as mean, standard deviation):

[0261] statistical_features=extract_statistical_features(data);

[0262] Extract frequency domain features (such as power spectrum density, frequency peak);

[0263] frequency_domain_features=extract_frequency_domain_features(data);

[0264] Extract time domain features (such as autocorrelation function, cross-correlation function):

[0265] time_domain_features=extract_time_domain_features(data);

[0266] Merge features:

[0267] features=combine_features(statistical_features,frequency_domain_features,time_domain_features);

[0268] return features;

[0269] def correlation_analysis(data);

[0270] Calculate the correlation coefficient between features;

[0271] correlation_matrix=calculate_correlation(data);

[0272] return correlation_matrix;

[0273] Build and train SNN models and training. First, build a neural network model based on the number of extracted features, and select the appropriate network structure and parameter settings. Then, compile the model, train the model using the training set, and iteratively optimize the model parameters.

[0274] def build_SNN_model(input_dim);

[0275] model = Sequential();

[0276] Add hidden layers;

[0277] model.add(Dense(64,activation='relu',input_dim=input_dim));

[0278] Add an output layer:

[0279] model.add(Dense(1,activation='sigmoid'));

[0280] return model;

[0281] def train_SNN_model(model,X_train,y_train);

[0282] Compile the model;

[0283] model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy']);

[0284] Train the model;

[0285] model.fit(X_train,y_train,epochs=10,batch_size=32);

[0286] return model;

[0287] Predict and analyze the results of the SNN model. First, use the trained SNN model to predict the data extracted by the SEI soil environment fusion algorithm and the SSI spatial interpolation algorithm to obtain comprehensive soil pollution health risk assessment data. Then, according to actual needs, further analyze and interpret the prediction results, such as formulating risk assessment standards and performing anomaly detection.

[0288] def predict_SNN_model(model,X);

[0289] Predicted soil pollution health risk assessment data:

[0290] predictions=model.predict(X);

[0291] return predictions;

[0292] def analyze_predictions(predictions);

[0293] Analyze the forecast results;

[0294] The corresponding analysis function can be defined according to actual needs;

[0295] Such as establishing risk assessment standards and conducting anomaly detection.

[0296] analyzed_results=analyze(predictions);

[0297] return analyzed_results;

[0298] Call functions such as feature extraction, association analysis, SNN model training, prediction and analysis, and finally return the analysis results.

[0299] def main(X1,X2,y);

[0300] Feature extraction and association analysis;

[0301] features=feature_extraction(X1);

[0302] correlation_matrix=correlation_analysis(X2);

[0303] Data Fusion:

[0304] X=merge_data(features, correlation_matrix);

[0305] SNN model training:

[0306] model=build_SNN_model(X.shape[1]);

[0307] trained_model=train_SNN_model(model,X,y);

[0308] Prediction and analysis:

[0309] predictions=predict_SNN_model(trained_model,X);

[0310] analyzed_results=analyze_predictions(predictions);

[0311] return analyzed_results;

[0312] In this implementation step, the data extracted by the SEI soil environment fusion algorithm and the SSI spatial interpolation algorithm are integrated, and the independently innovative SNN model is used for data fusion and correlation analysis. Through the SNN model, we can comprehensively consider multiple factors of soil properties and environmental factors, and extract the characteristics and correlation information of soil pollution to obtain comprehensive soil pollution health risk assessment data. The creativity of this method lies in the use of the SNN model, which can fuse data from different sources and find the correlation between them, thereby improving the accuracy and reliability of the assessment results.

[0313] In another embodiment provided by the present invention, the step S5 specifically includes:

[0314] Associating the soil pollution detection data with the geographic terrain data and soil type data in the geographic spatial information;

[0315] GIS software is used to perform spatial statistical analysis and spatial interaction analysis on the associated data to obtain spatial analysis data containing the spatial distribution characteristics of soil pollution.

[0316] In the specific implementation of this embodiment, the comprehensive soil pollution health risk assessment data is combined with geographic information space information using geographic information system (GIS) technology to perform geospatial analysis. Soil pollution data is associated with geographic terrain data, soil type data and other geospatial information, and spatial statistical analysis and spatial interaction analysis are performed using GIS software to obtain the spatial distribution characteristics of soil pollution and risk assessment results. The specific implementation steps are as follows:

[0317] Obtain geospatial information from the Geographic Information System (GIS) database, such as terrain data, soil type data, etc. Then load the soil pollution health risk assessment data and geospatial information data and associate the two.

[0318] def prepare_data(risk_assessment_data,geospatial_data);

[0319] Load soil pollution health risk assessment data:

[0320] risk_assessment_data=load_risk_assessment_data(risk_assessment_data);

[0321] Load geospatial information data:

[0322] geospatial_data=load_geospatial_data(geospatial_data);

[0323] To associate two datasets:

[0324] combined_data=merge_data(risk_assessment_data,geospatial_data);

[0325] return combined_data;

[0326] Define the spatial analysis algorithm formula:

[0327] Y=f(X1,X2,...,Xn);

[0328] Among them, Y represents the spatial distribution characteristics and risk assessment results of soil pollution, X1, X2, ..., Xn represent geographic spatial information, and f represents the function or algorithm of spatial analysis.

[0329] Use GIS software to perform spatial statistical analysis and select appropriate spatial statistical methods and algorithms according to needs, such as spatial clustering analysis, spatial interpolation analysis, spatial association analysis, etc.

[0330] def spatial_statistical_analysis(data);

[0331] Use GIS software to conduct spatial statistical analysis, such as spatial clustering analysis, spatial interpolation analysis, spatial association analysis, etc.

[0332] result=GIS.spatial_statistical_analysis(data);

[0333] return result;

[0334] Use GIS software to conduct spatial interaction analysis and select appropriate spatial interaction methods and algorithms according to needs, such as spatial buffer analysis, heterogeneous spatial analysis, etc.

[0335] def spatial_interaction_analysis(data);

[0336] Use GIS software to conduct spatial interaction analysis, such as spatial buffer analysis, heterogeneous spatial analysis, etc.

[0337] result=GIS.spatial_interaction_analysis(data);

[0338] return result;

[0339] Display and report the results of geospatial analysis. First, use GIS software to visualize the results of geospatial analysis, such as generating soil pollution heat maps, spatial distribution maps, etc. Then, generate reports and statements to use the results of geospatial analysis for soil pollution monitoring, assessment and decision support.

[0340] def visualize_results(result);

[0341] Use GIS software to visualize the results of geographic spatial analysis, such as generating soil pollution heat maps, spatial distribution maps, etc.

[0342] visualization=GIS.visualize_results(result);

[0343] return visualization;

[0344] def generate_report(visualization);

[0345] Generate reports and statements to use the results of geospatial analysis for soil contamination monitoring, assessment and decision support.

[0346] report=generate_report(visualization);

[0347] return report;

[0348] Call functions such as data preparation, spatial statistical analysis, spatial interactive analysis, result display and reporting, and finally return the generated reports and statements.

[0349] def main(risk_assessment_data,geospatial_data);

[0350] Data preparation;

[0351] combined_data=prepare_data(risk_assessment_data,geospatial_data).

[0352] Spatial statistical analysis:

[0353] statistical_result=spatial_statistical_analysis(combined_data).

[0354] Spatial interaction analysis:

[0355] interaction_result=spatial_interaction_analysis(combined_data);

[0356] Results presentation and reporting;

[0357] visualization=visualize_results(statistical_result);

[0358] report=generate_report(visualization);

[0359] return report;

[0360] Through the above implementation steps and algorithm formulas, geographic information system (GIS) technology can be used to combine comprehensive soil pollution health risk assessment data with geospatial information for geospatial analysis.

[0361] In another embodiment provided by the present invention, the step S6 specifically includes:

[0362] Calculate the soil pollution degree index, human contact pathway index and comprehensive impact coefficient respectively according to the soil pollution detection data and the spatial analysis data;

[0363] Inputting the soil pollution degree index, the human contact pathway index and the comprehensive impact coefficient into the health risk assessment model to calculate the risk index;

[0364] Determining the risk level according to the risk index;

[0365] Risk index R = (w1*S+w2*E+w3*T) / (w4*S+w5*E+w6*T+ε), w1~w6 are preset weight coefficients, ε is a preset threshold; the soil pollution degree index coefficient The human exposure pathway indicator coefficient The comprehensive influence coefficient S j is the jth soil pollution index, W j is the weight of the jth soil pollution index, m is the number of soil pollution indexes, E j is the jth human contact pathway indicator, η j is the weight of the jth human contact pathway indicator, p is the number of human contact pathway indicators, T j is the exposure time index of the jth toxicity index, σ j is the weight of the j-th toxicity index exposure time, and q is the number of toxicity index exposure time indicators.

[0366] In the specific implementation of this embodiment, an SHR health risk assessment model is established based on comprehensive soil pollution health risk assessment data. The model comprehensively considers multiple factors such as soil properties, environmental factors, and geographic spatial information, and accurately calculates the potential risks of urban soil pollution to human health through a mathematical model. The model can comprehensively assess the impact of soil pollution on human health based on the characteristics and associated information of soil pollution, and calculate the corresponding health risks.

[0367] The SHR health risk assessment model (Soil Health Risk) is an independently innovative model.

[0368] Formula: R = f(S, E, T);

[0369] In the formula,

[0370] R represents the potential risk of urban soil pollution to human health;

[0371] S represents the soil pollution degree index coefficient;

[0372] E represents the human exposure pathway indicator coefficient;

[0373] T represents the comprehensive impact coefficient for different toxicity indicators and exposure times.

[0374] Specifically in the SHR health risk assessment model, the calculation process is based on the following algorithm formula:

[0375] Calculation of soil pollution index:

[0376] The soil pollution degree index coefficient The human exposure pathway indicator coefficient The comprehensive influence coefficient S j is the jth soil pollution index, W j is the weight of the jth soil pollution index, m is the number of soil pollution indexes, E j is the jth human contact pathway indicator, η j is the weight of the jth human contact pathway indicator, p is the number of human contact pathway indicators, T j is the exposure time index of the jth toxicity index, σ j is the weight of the j-th toxicity index exposure time, and q is the number of toxicity index exposure time indicators.

[0377] Risk index R = (w1*S+w2*E+w3*T) / (w4*S+w5*E+w6*T+ε), w1~w6 are preset weight coefficients, and ε is a preset threshold.

[0378] Define the algorithm formula of the health risk assessment model. Assume that S represents the comprehensive index of soil properties, E represents the comprehensive index of environmental factors, and T represents the comprehensive index of geographic spatial information. By analyzing the comprehensive soil pollution health risk assessment data, determine how to combine these indicators to calculate the health risk assessment result R. Establish a mathematical model to determine how to calculate the health risk assessment result R by analyzing the comprehensive soil pollution health risk assessment data. R = (w1*S+w2*E+w3*T) / (w4*S+w5*E+w6*T+ε), w1~w6 are preset weight coefficients used to adjust the contribution of different indicators to health risks. ε is a threshold used to adjust the stability of risk assessment results. Define functions to calculate soil pollution degree indicators, human contact pathway indicators, comprehensive impact coefficients, risk indexes, and judge risk levels.

[0379] Define functions to calculate soil pollution degree index, human exposure pathway index, comprehensive impact coefficient, risk index and judge risk level.

[0380] def calculate_soil_pollution_index(data);

[0381] Calculate soil pollution level indicators based on comprehensive soil pollution health risk assessment data;

[0382] Including pollutant content in soil, toxicity index, ecological environment standards, etc.;

[0383] Returns the soil pollution degree index S;

[0384] def calculate_exposure_pathways_index(data);

[0385] Calculate human exposure pathway indicators based on comprehensive soil pollution health risk assessment data;

[0386] Including air breathing, food ingestion, skin contact and other contact routes;

[0387] Returns human exposure pathway indicator E;

[0388] def calculate_comprehensive_impact_factor(data);

[0389] Calculate the comprehensive impact coefficient based on the comprehensive soil pollution health risk assessment data;

[0390] Including influencing factors such as soil properties, environmental factors, and geospatial information;

[0391] Returns the comprehensive influence coefficient T;

[0392] def calculate_risk_index(S,E,T);

[0393] The risk index is calculated based on the soil pollution degree index S, the human contact pathway index E, and the comprehensive impact coefficient T:

[0394] Returns the risk index R;

[0395] def determine_risk_level(R);

[0396] According to the risk index R, refer to the relevant threshold standards to judge the risk level.

[0397] Depending on the specific situation, the risk level can be divided into several levels, such as low risk, medium risk, high risk, etc.

[0398] Returns the risk level.

[0399] def SHR_model(data);

[0400] Calculate soil pollution degree index, human contact route index, comprehensive impact coefficient, risk index and judge risk level.

[0401] Call the function defined above to perform calculations.

[0402] Returns the health risk assessment result R and risk level.

[0403] In this implementation step, we established the algorithm formula of the SHR health risk assessment model based on the comprehensive soil pollution health risk assessment data. By defining the algorithm formula and function, we can calculate the soil pollution degree index S, the human contact pathway index E, the comprehensive impact coefficient T, and use a more complex algorithm formula to calculate the risk index R. Finally, according to the value of the risk index, the risk level is judged with reference to the threshold standard. The creativity of this method lies in the accurate calculation of the potential risks of urban soil pollution to human health through mathematical models, and the comprehensive consideration of multiple factors such as soil properties, environmental factors, and geographic spatial information.

[0404] In another embodiment provided by the present invention, determining the risk level according to the risk index specifically includes:

[0405] When the risk index is not higher than a preset first threshold, determining the risk level as low risk;

[0406] When the risk index is higher than the first threshold value and not higher than a preset second threshold value, determining that the risk level is medium risk;

[0407] When the risk index is higher than the second threshold, determining the risk level as high risk;

[0408] The second threshold is higher than the first threshold.

[0409] In the specific implementation of this embodiment, when determining the risk level,

[0410] If the risk index is less than or equal to the threshold of 1, the risk level is low risk;

[0411] If the risk index is greater than threshold 1 and less than or equal to threshold 2, the risk level is medium risk;

[0412] If the risk index is greater than the threshold of 2, the risk level is high risk.

[0413] It should be noted that, in this embodiment, the first threshold and the second threshold are 2 and 1 respectively. In other embodiments, the first threshold and the second threshold can be adjusted according to actual conditions.

[0414] The embodiments of the present invention are uniquely creative in terms of IoT data fusion and innovative algorithm application. By using machine learning algorithms, SSI spatial interpolation algorithms and SNN models, combined with geographic information system technology, accurate and comprehensive soil pollution health risk assessment data can be obtained. Based on these data, an SHR health assessment model is established to accurately calculate the potential risks of urban soil pollution to human health.

[0415] In another embodiment provided by the present invention, see Figure 2 , is a schematic diagram of the structure of a soil pollution detection device based on Internet of Things data fusion provided by an embodiment of the present invention, the device comprising:

[0416] An acquisition module, used to acquire multi-source data related to soil pollution in the area to be detected, wherein the multi-source data includes different types of sensor data and monitoring site data;

[0417] An estimation module, used to perform end-to-end fusion and prediction of the multi-source data using a preset soil environment fusion algorithm to obtain estimated data of the soil pollution degree;

[0418] An inference module, used to infer unobserved data using a preset spatial interpolation algorithm on the multi-source data to obtain soil pollution assessment data;

[0419] A fusion module, used to fuse the estimated data with the evaluation data through a preset spiking neural network model to obtain comprehensive soil pollution detection data;

[0420] A spatial analysis module, used to combine the detection data with the geographic spatial information of the area to be detected by using geographic information system technology, perform geographic spatial analysis, and obtain spatial analysis data;

[0421] The level calculation module is used to input the soil pollution detection data and the spatial analysis data into a pre-built health risk assessment model to calculate the risk level of soil pollution.

[0422] It should be noted that the soil pollution detection device based on Internet of Things data fusion provided in the embodiment of the present invention can execute the soil pollution detection method based on Internet of Things data fusion described in any of the above embodiments, and the specific functions of the soil pollution detection device based on Internet of Things data fusion are not described here.

[0423] See also Figure 3 , is a schematic diagram of the structure of a terminal device provided in an embodiment of the present invention. The terminal device of this embodiment includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a soil pollution detection program based on IoT data fusion. When the processor executes the computer program, the steps in the above-mentioned soil pollution detection method based on IoT data fusion are implemented, such as Figure 1 Alternatively, the processor implements the functions of each module in the above-mentioned device embodiments when executing the computer program.

[0424] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program in the terminal device. For example, the computer program may be divided into various modules, and the specific functions of each module are not described in detail again.

[0425] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The terminal device may include, but is not limited to, a processor and a memory. Those skilled in the art will appreciate that the schematic diagram is merely an example of a terminal device and does not constitute a limitation on the terminal device. The terminal device may include more or fewer components than shown in the diagram, or may combine certain components, or different components. For example, the terminal device may also include an input / output device, a network access device, a bus, etc.

[0426] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the terminal device, and uses various interfaces and lines to connect various parts of the entire terminal device.

[0427] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the terminal device by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0428] Wherein, if the module / unit integrated in the terminal device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Wherein, the computer program includes computer program code, and the computer program code can be in code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0429] It should be noted that the device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art may understand and implement it without paying any creative effort.

[0430] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A soil pollution detection method based on Internet of Things data fusion, characterized in that: The method comprises the following steps: Acquire multi-source data related to soil pollution in the area to be detected, wherein the multi-source data includes different types of sensor data and monitoring site data; Using a preset soil environment fusion algorithm to perform end-to-end fusion and prediction on the multi-source data to obtain estimated data of the soil pollution degree; Using a preset spatial interpolation algorithm to infer unobserved data from the multi-source data to obtain soil pollution assessment data; The estimated data and the evaluation data are merged by a preset spiking neural network model to obtain comprehensive soil pollution detection data; Combining the detection data with the geographic spatial information of the area to be detected by using geographic information system technology, performing geographic spatial analysis, and obtaining spatial analysis data; Inputting the soil pollution detection data and the spatial analysis data into a pre-built health risk assessment model to calculate the risk level of soil pollution; The use of a preset soil environment fusion algorithm to perform end-to-end fusion and prediction on the multi-source data to obtain estimated data on the degree of soil pollution specifically includes: Preprocessing the sensor data, including data cleaning, missing value processing and abnormal data removal; Adaptive fusion method is used to fuse the data of different sensors to obtain fused data of soil pollution degree; Predicting the fused data based on a pre-trained deep learning model to obtain estimated data on the degree of soil contamination; The method of using a preset spatial interpolation algorithm to infer unobserved data from the multi-source data to obtain soil pollution assessment data specifically includes: Preprocessing the monitoring site data, including data cleaning and outlier removal; Using geographic information system technology, spatial interpolation is performed based on the spatial distribution characteristics of monitoring sites to estimate the assessment data of unobserved soil pollution; Among them, the data at the unobserved position u in the evaluation data , where Z(i) represents the value of the observed data point i, w(i) represents the weight of the data point i, n represents the number of observed data points, and W represents the sum of weighted coefficients; The estimated data and the evaluation data are fused by a preset pulse neural network model to obtain comprehensive soil pollution detection data, specifically including: Performing feature extraction and correlation analysis on the estimated data and the evaluation data, extracting statistical features, frequency domain features and time domain features, obtaining feature data, and determining the correlation between the features to determine a correlation coefficient matrix; The characteristic data are fused using a pre-trained pulse neural network model to obtain comprehensive soil pollution risk assessment data; Based on the correlation coefficient matrix, the risk assessment data is correlated and calculated to obtain comprehensive soil pollution detection data.

2. The soil pollution detection method based on Internet of Things data fusion according to claim 1 is characterized in that: The method of combining the soil pollution detection data with the geographic space information of the area to be detected by using geographic information system technology to perform geographic space analysis to obtain spatial analysis data includes: Associating the soil pollution detection data with the geographic terrain data and soil type data in the geographic spatial information; GIS software is used to perform spatial statistical analysis and spatial interaction analysis on the associated data to obtain spatial analysis data containing the spatial distribution characteristics of soil pollution.

3. The soil pollution detection method based on Internet of Things data fusion according to claim 1 is characterized in that: The step of inputting the soil pollution detection data and the spatial analysis data into a pre-built health risk assessment model to calculate the risk level of soil pollution includes: Calculate the soil pollution degree index, human contact pathway index and comprehensive impact coefficient respectively according to the soil pollution detection data and the spatial analysis data; Inputting the soil pollution degree index, the human contact pathway index and the comprehensive impact coefficient into the health risk assessment model to calculate the risk index; Determining the risk level according to the risk index; Among them, risk index R = (w1 * S + w2 * E + w3 * T) / (w4 * S + w5 * E + w6 * T +ε), w1~w6 are preset weight coefficients, ε is the preset threshold; the soil pollution degree index coefficient , the human exposure pathway indicator coefficient , the comprehensive influence coefficient , For the j Soil pollution indicators, For the j The weight of each soil pollution index, m is the number of soil pollution indicators, For the j Individual human contact pathway indicators, For the j The weight of the individual human exposure pathway indicator, p is the number of human exposure pathway indicators, For the j Toxicity index exposure time index, For the j The weight of exposure time of each toxicity indicator, q The number of exposure time indicators for toxicity indicators.

4. The soil pollution detection method based on Internet of Things data fusion according to claim 3 is characterized in that: Determining the risk level according to the risk index specifically includes: When the risk index is not higher than a preset first threshold, determining the risk level as low risk; When the risk index is higher than the first threshold value and not higher than a preset second threshold value, determining that the risk level is medium risk; When the risk index is higher than the second threshold, determining the risk level as high risk; The second threshold is higher than the first threshold.

5. A soil pollution detection device based on Internet of Things data fusion, characterized in that: The device comprises: An acquisition module, used to acquire multi-source data related to soil pollution in the area to be detected, wherein the multi-source data includes different types of sensor data and monitoring site data; An estimation module, used to perform end-to-end fusion and prediction of the multi-source data using a preset soil environment fusion algorithm to obtain estimated data of the soil pollution degree; An inference module, used to infer unobserved data using a preset spatial interpolation algorithm on the multi-source data to obtain soil pollution assessment data; A fusion module, used to fuse the estimated data with the evaluation data through a preset spiking neural network model to obtain comprehensive soil pollution detection data; A spatial analysis module, used to combine the detection data with the geographic spatial information of the area to be detected by using geographic information system technology, perform geographic spatial analysis, and obtain spatial analysis data; A level calculation module, used to input the soil pollution detection data and the spatial analysis data into a pre-built health risk assessment model to calculate the risk level of soil pollution; The estimation module is specifically used for: Preprocessing the sensor data, including data cleaning, missing value processing and abnormal data removal; Adaptive fusion method is used to fuse the data of different sensors to obtain fused data of soil pollution degree; Predicting the fused data based on a pre-trained deep learning model to obtain estimated data on the degree of soil contamination; The calculation module is specifically used for: Preprocessing the monitoring site data, including data cleaning and outlier removal; Using geographic information system technology, spatial interpolation is performed based on the spatial distribution characteristics of monitoring sites to estimate the assessment data of unobserved soil pollution; Among them, the data at the unobserved position u in the evaluation data , where Z(i) represents the value of the observed data point i, w(i) represents the weight of the data point i, n represents the number of observed data points, and W represents the sum of weighted coefficients; The fusion module is specifically used for: Performing feature extraction and correlation analysis on the estimated data and the evaluation data, extracting statistical features, frequency domain features and time domain features, obtaining feature data, and determining the correlation between the features to determine a correlation coefficient matrix; The characteristic data are fused using a pre-trained pulse neural network model to obtain comprehensive soil pollution risk assessment data; Based on the correlation coefficient matrix, the risk assessment data is correlated and calculated to obtain comprehensive soil pollution detection data.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the soil pollution detection method based on Internet of Things data fusion as described in any one of claims 1 to 4.

7. A terminal device, characterized in that: It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it implements the soil pollution detection method based on Internet of Things data fusion as described in any one of claims 1 to 4.

Images