Intelligent environmental protection management and control method and system

Abstract

The invention provides an intelligent environmental protection management and control method and system, and relates to the technical field of environmental protection management and control, and the intelligent environmental protection management and control method comprises the steps: obtaining multi-dimensional environment detection information of a management and control region; performing feature fusion on the multi-dimensional environment detection information to generate fused feature information; pollution identification is carried out based on the fusion feature information, a pollution identification result is generated, and the pollution identification result is used for representing whether pollution discharge of the management and control area at the current moment reaches the standard or not and / or whether pollution discharge of the management and control area at the future moment reaches the standard or not; and according to the pollution identification result, a corresponding control scheme is generated, and the control scheme comprises pollution warning based on the control area whose pollution discharge does not reach the standard. According to the invention, the environment condition can be effectively monitored, accurate pollution management and timely response are realized, and errors and hysteresis caused by manual detection are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental protection management and control, and in particular to an intelligent environmental protection management and control method and system. Background Art

[0002] With the advancement of urbanization and the rapid development of social economy, the number of engineering projects undertaken by heavy industrial enterprises such as steel companies has increased significantly. During the implementation of engineering projects, environmental pollution caused by emissions has received more and more attention.

[0003] In related technologies, relevant technical personnel usually conduct regular inspections of pollution emissions to achieve environmental protection control. However, this method requires relevant technical personnel to have considerable knowledge and experience, and manual inspections are prone to empirical errors and have a certain lag, which may lead to failure to timely discover and correct excessive pollution emissions. Summary of the invention

[0004] The problem to be solved by the present invention is that in the process of environmental protection control, human detection is prone to empirical errors and hysteresis, which may lead to failure to timely discover and correct the situation where pollution emissions exceed the standard.

[0005] In order to solve the above problems, in a first aspect, the present invention provides an intelligent environmental protection management and control method, comprising:

[0006] Obtain multi-dimensional environmental detection information of the controlled area;

[0007] Performing feature fusion on the multi-dimensional environment detection information to generate fused feature information;

[0008] Perform pollution identification based on the fused feature information to generate a pollution identification result, wherein the pollution identification result is used to indicate whether the pollution discharge of the control area at the current moment meets the standard, and / or whether the pollution discharge of the control area at a future moment meets the standard;

[0009] A corresponding control plan is generated based on the pollution identification result, wherein the control plan includes issuing a pollution warning based on the control area where pollution discharge does not meet the standards.

[0010] Optionally, the performing feature fusion on the multi-dimensional environment detection information to generate fused feature information includes:

[0011] The multi-dimensional environment detection information is input into an information fusion model to generate the fusion feature information, wherein the information fusion model is pre-trained and includes a first deep learning network model.

[0012] Optionally, the performing feature fusion on the multi-dimensional environment detection information to generate fused feature information includes:

[0013] Performing feature vectorization processing on the multi-dimensional environment detection information to obtain multiple feature vectors;

[0014] Extracting statistical features from the plurality of feature vectors, wherein the statistical features include a mean and a variance;

[0015] normalizing the statistical features;

[0016] The standardized statistical features are concatenated to generate a unified feature vector, and the fused feature information includes the unified feature vector.

[0017] Optionally, generating a corresponding control plan according to the pollution identification result includes:

[0018] Based on the control area where the pollution discharge does not meet the standard at a future moment, an interception alarm notification is generated, and the interception alarm notification is sent to the supervision terminal;

[0019] Based on the controlled area that does not meet the pollution discharge standards at the current moment, a verification alarm notification is generated and sent to the supervision terminal.

[0020] Optionally, generating a corresponding control plan according to the pollution identification result includes:

[0021] Obtain the emission monitoring concentration, preset concentration amplitude, the net pollutant content of the stage and the estimated value of the change of the net pollutant content of the stage in the control area where the pollutant discharge does not meet the standard within a preset time period;

[0022] According to the emission monitoring concentration, the preset concentration amplitude, the net pollutant content of the stage and the estimated value of the change in the net pollutant content of the stage, it is judged whether there is diluted pollution discharge in the control area where the pollution discharge does not meet the standard;

[0023] If so, a key inspection and alarm plan is generated based on the controlled area where diluted pollution exists.

[0024] Optionally, judging whether there is diluted pollution discharge in the control area where pollution discharge does not meet the pollution discharge standard according to the emission monitoring concentration, the preset concentration amplitude, the net pollution discharge content of the stage and the estimated value of the change of the net pollution discharge content of the stage includes:

[0025] Determine whether the emission monitoring concentration in the control area where the pollution discharge does not meet the standards has decreased;

[0026] If the emission monitoring concentration of the control area where the pollution discharge does not meet the standards decreases, then monitor whether the decrease in the emission monitoring concentration of the control area where the pollution discharge does not meet the standards is greater than or equal to the preset concentration amplitude;

[0027] If the decrease in the emission monitoring concentration of the control area that does not meet the pollution discharge standard is greater than or equal to the preset concentration amplitude, it is determined whether the increase in the net pollution discharge content of the corresponding control area at that stage is greater than the estimated value of the change in the net pollution discharge content at that stage;

[0028] If so, it is determined that there is dilution discharge in the corresponding controlled area.

[0029] Optionally, performing pollution identification based on the fused feature information to generate a pollution identification result includes:

[0030] The fused feature information is input into an environmental management and control model, and the pollution identification result is output, wherein the environmental management and control model is pre-trained and includes a second deep learning network model.

[0031] Optionally, the step of pre-training to obtain the environmental protection control model includes:

[0032] Obtaining environmental detection feature samples and pollution identification result labels of the environmental detection feature samples, and iteratively training the initial environmental protection control model until a first difference value between the output pollution identification result and the pollution identification result label meets a corresponding error standard, so as to obtain the trained environmental protection control model;

[0033] Alternatively, obtain an environmental detection feature sample corresponding to a prediction determining factor and a pollution identification result label of the environmental detection feature sample, wherein the prediction determining factor includes first time information corresponding to the environmental detection feature sample, a production plan corresponding to the environmental detection feature sample, and second time information corresponding to the production plan; based on the environmental detection feature sample and the pollution identification result label corresponding to the prediction determining factor, iteratively train the initial environmental protection control model until the second difference value between the output pollution identification result and the pollution identification result label meets the corresponding error standard to obtain the trained environmental protection control model.

[0034] The information fusion model and the environmental protection control model are obtained by joint training in advance, and the steps of obtaining the information fusion model and the environmental protection control model by joint training in advance include:

[0035] Obtain multidimensional environmental detection samples and pollution identification result labels corresponding to the multidimensional environmental detection samples, and based on the multidimensional environmental detection samples and the pollution identification result labels corresponding to the multidimensional environmental detection samples, jointly train the initial information fusion model and the environmental protection control model to obtain the trained information fusion model and the environmental protection control model.

[0036] Alternatively, obtain the multidimensional environmental detection samples corresponding to the prediction determination factors and the pollution identification result labels corresponding to the multidimensional environmental detection samples, wherein the prediction determination factors include the first time information corresponding to the multidimensional environmental detection samples, the production plan corresponding to the multidimensional environmental detection samples, and the second time information corresponding to the production plan; based on the multidimensional environmental detection samples corresponding to the prediction determination factors and the pollution identification result labels corresponding to the multidimensional environmental detection samples, jointly train the initial information fusion model and the environmental protection control model to obtain the trained information fusion model and the environmental protection control model.

[0037] Optionally, it also includes:

[0038] If the pollution discharge in the controlled area meets the standards, the emission monitoring data and production scale data of the controlled area that meets the pollution discharge standards are counted and stored.

[0039] The present invention obtains multi-dimensional environmental detection information of the control area, and after receiving the sensor data of each detection point, performs feature fusion on the multi-dimensional environmental detection information to generate fused feature information, and performs pollution identification based on the fused feature information. It can reduce the noise and error of a single sensor data, realize automatic analysis and alarm judgment, reduce the empirical error caused by manual detection, and output the pollution discharge status (meeting the standard or not) of the control area, realize the identification of the current pollution discharge risk of the control area or predict the risk of future pollution discharge, so as to formulate a corresponding alarm plan. The alarm plan may include alarm level, alarm content, etc., and then notify the relevant departments to carry out corresponding processing. For the control area where there is currently a pollution discharge risk, accurate pollution management and timely response can be achieved. For the control area where there is a pollution discharge risk in the future, the lag caused by manual detection can be reduced, so that reasonable treatment measures such as emission interception can be taken in advance.

[0040] In a second aspect, the present invention further provides an intelligent environmental protection management and control system, which applies the intelligent environmental protection management and control method as described above, and comprises:

[0041] The detection information acquisition module is used to obtain multi-dimensional environmental detection information of the control area;

[0042] A feature fusion module, used to perform feature fusion on the multi-dimensional environment detection information to generate fused feature information;

[0043] A pollution identification module, used to perform pollution identification based on the fused feature information and generate a pollution identification result, wherein the pollution identification result is used to indicate whether the pollution discharge of the control area at the current moment meets the standard and / or whether the pollution discharge of the control area at a future moment meets the standard;

[0044] A control module is used to generate a corresponding control plan based on the pollution identification result, wherein the control plan includes a pollution alarm based on the control area where the pollution discharge does not meet the standard

[0045] In a third aspect, the present invention provides an electronic device, including a memory and a processor;

[0046] The memory is used to store computer programs;

[0047] The processor is used to implement the intelligent environmental protection management and control method as described in the first aspect when executing the computer program.

[0048] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the intelligent environmental protection management and control method as described in the first aspect is implemented.

[0049] The intelligent environmental protection management and control system, electronic device and computer-readable storage medium provided by the present invention have the same beneficial effects as the intelligent environmental protection management and control method relative to the prior art, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A schematic diagram of a process of an intelligent environmental protection management method in an embodiment of the present invention is shown;

[0051] Figure 2 The structure diagram of the intelligent environmental protection management and control system in an embodiment of the present invention is shown;

[0052] Figure 3 A schematic structural diagram of an electronic device in an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0053] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0054] It should be noted that, in the present invention, 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 terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0055] In the description of this specification, the description with reference to the terms "embodiment", "one embodiment" and "one implementation" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or implementation are included in at least one embodiment or implementation of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or implementation. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or implementations in a suitable manner.

[0056] Reference Figure 1 As shown, an embodiment of the present invention proposes an intelligent environmental protection management method;

[0057] The intelligent environmental protection control method comprises:

[0058] S100: Obtain multi-dimensional environmental detection information of the controlled area.

[0059] Specifically, the method for obtaining multi-dimensional environmental detection information is to receive the detection information of the current detection point collected by the sensors of each detection point. The detection sensors include but are not limited to TSP (total suspended particulate) monitors, moisture monitors, air quality monitors, video surveillance equipment, and environmental protection vehicle positioning equipment, etc. The corresponding multi-dimensional detection information includes but is not limited to video image data and vehicle positioning data.

[0060] S200: Perform feature fusion on the multi-dimensional environment detection information to generate fused feature information.

[0061] Specifically, by fusing multiple data sources, richer environmental features can be obtained to help better understand and analyze environmental problems. Feature fusion can reduce the noise and errors of single sensor data. For example, air quality monitors may be affected by the local environment, while video surveillance equipment can provide additional contextual information, thereby improving the overall detection accuracy. By integrating multiple data, pollution sources and environmental changes can be identified more accurately.

[0062] S300: Perform pollution identification based on the fused feature information and generate a pollution identification result, wherein the pollution identification result is used to characterize whether the pollution discharge of the controlled area meets the standard at the current moment, and / or whether the pollution discharge of the controlled area meets the standard at a future moment.

[0063] Specifically, the input fusion feature information can be processed by constructing a pollution identification model and using machine learning or deep learning models. The model can be a classifier (such as a support vector machine, decision tree, neural network, etc.), and its goal is to determine whether the environment meets the standard based on the input features. The model needs to be trained and use historical data (including known pollution conditions) to learn the relationship between features and pollution status. The fusion feature information is input into the trained pollution identification model. The model will infer based on the input features and output the pollution identification result. The identification result is usually a classification result, indicating the pollution discharge status of the control area. The pollution discharge status can be determined by the current pollution discharge situation or the pollution discharge prediction at a future moment.

[0064] S400: Generate a corresponding control plan based on the pollution identification result, wherein the control plan includes issuing a pollution alarm based on the control area where pollution discharge does not meet the standards.

[0065] Specifically, based on the pollution identification results, it is determined that the pollution discharge in at least one control area does not meet the standards. This means that the emission monitoring points in the area may exceed the emission standards in the future or are already very likely to have pollution discharge that does not meet the standards. For the identified pollution discharge control areas that do not meet the standards, an alarm plan for the non-compliant emission monitoring points is generated. The alarm plan may include one or more of the following: Alarm level: set the alarm level according to the degree of exceeding the standard (such as mild, moderate, and severe), alarm content: describe in detail the pollutants that exceed the standard, the location of the monitoring point, the exceeding value, and other information, response measures: formulate corresponding response measures for different levels of alarms, for example: mild alarm: recommend strengthening monitoring, moderate alarm: require relevant enterprises to conduct self-inspection, evaluate the effectiveness of improving production processes, etc., severe alarm: immediately initiate emergency response, conduct on-site inspections and rectifications, or intercept the emissions that may exceed the standards to ensure that they will not cause more serious environmental problems, notify relevant departments, etc.

[0066] When this embodiment is applied in practice, by acquiring multi-dimensional environmental detection information of the control area, after receiving the sensor data of each detection point, the multi-dimensional environmental detection information is feature fused to generate fused feature information, and pollution identification is performed based on the fused feature information. This can reduce the noise and error of a single sensor data, realize automatic analysis and alarm assessment, reduce the empirical error caused by manual detection, and output the pollution discharge status (meeting the standard or not) of the control area, realize the identification of the current pollution discharge risk of the control area or predict the risk of future pollution discharge, so as to formulate a corresponding alarm plan. The alarm plan may include alarm level, alarm content, etc., and then notify the relevant departments to take corresponding measures. For the control area where there is currently a pollution discharge risk, accurate pollution management and timely response can be achieved. For the control area where there is a pollution discharge risk in the future, the lag caused by manual detection can be reduced, so as to facilitate the early adoption of reasonable treatment measures such as emission interception.

[0067] As an optional embodiment of the present invention, the step of performing feature fusion on the multi-dimensional environment detection information to generate fused feature information includes:

[0068] The multi-dimensional environment detection information is input into an information fusion model to generate fusion feature information, wherein the information fusion model is pre-trained and includes a first deep learning network model.

[0069] Specifically, the information fusion model can be trained by a deep neural network learning method using reinforcement learning, or can be trained by supervised and unsupervised deep neural network learning methods. The information fusion model is trained based on multi-dimensional reservation information samples, using reinforcement learning, or supervised or unsupervised methods for separate training. It can also be obtained by joint training with an environmental protection control model based on multi-dimensional detection samples. The training material of the information fusion model (multi-dimensional detection samples) comes from the detection data uploaded historically by multiple sensors.

[0070] As an optional embodiment of the present invention, the step of performing feature fusion on the multi-dimensional environment detection information to generate fused feature information includes:

[0071] The multi-dimensional environment detection information is subjected to feature vectorization processing to obtain a plurality of feature vectors.

[0072] Specifically, multi-dimensional environmental detection information (such as SP monitoring data, moisture monitoring data, air quality monitoring data, etc.) is converted into numerical feature vectors. This process can use some common feature extraction methods, such as Fourier transform, wavelet transform, etc., to convert the original signal into a series of numerical features.

[0073] Statistical features are extracted from the plurality of feature vectors, wherein the statistical features include a mean and a variance.

[0074] From the multiple feature vectors obtained in the previous step, statistical features are extracted, including: mean: reflecting the overall level of the feature vector, variance: reflecting the degree of dispersion of the feature vector, and the statistical features are standardized.

[0075] Specifically, the extracted statistical features (mean and variance) are standardized. The purpose of standardization is to eliminate the influence of dimension and make different features comparable. Commonly used standardization methods are:

[0076] For a given dataset, calculate the mean and variance of each feature.

[0077] The mean is the average of the eigenvalues, and the variance is the average of the squared deviations of the eigenvalues ​​from the mean.

[0078] Mean:

[0079] variance:

[0080] Each feature is normalized using the calculated mean and variance.

[0081] The normalization formula is:

[0082] Among them, u represents the mean, N represents the number of eigenvalues, i represents the number of items, and x i represents the eigenvalue, σ 2 represents variance, z represents the standardized value, and σ represents the standard deviation.

[0083] Apply the normalization formula to each feature in the dataset to get the normalized dataset, check the normalization results to make sure that the mean of the normalized features is close to 0 and the standard deviation is close to 1. This can be verified by calculating the mean and standard deviation after normalization.

[0084] The standardized statistical features are concatenated to generate a unified feature vector, and the fused feature information includes the unified feature vector.

[0085] Specifically, the standardized statistical features (mean and variance) are concatenated into a unified feature vector. This unified feature vector contains the fusion feature information of the original multi-dimensional environment detection information. When the existing features are fused, first determine whether the dimensions are the same. If they are different, linearly transform them into features of the same dimension, and then directly add the corresponding elements.

[0086] When this embodiment is applied in practice, the process is to quantize the original multi-dimensional environment detection information, extract statistical features and standardize it, and finally generate a unified feature vector for subsequent analysis and application. This feature fusion method can effectively extract and retain the key features of the original information, and provide valuable input features for subsequent pattern recognition, anomaly detection and other tasks.

[0087] As an optional embodiment of the present invention, generating a corresponding control plan according to the pollution identification result includes:

[0088] Obtain the emission monitoring concentration, preset concentration amplitude, the net pollutant content of the stage and the estimated value of the change of the net pollutant content of the stage in the control area where the pollutant discharge does not meet the standard within a preset time period;

[0089] The preset time period is a preset future time period, whereby corresponding concentration data is automatically acquired after reaching this time period.

[0090] According to the emission monitoring concentration, the preset concentration amplitude, the net pollutant content of the stage and the estimated value of the change in the net pollutant content of the stage, it is judged whether there is diluted pollution discharge in the control area where the pollution discharge does not meet the standard;

[0091] If so, a key inspection and alarm plan is generated based on the controlled area where diluted pollution exists.

[0092] The net content of sewage discharge is the content of pollutants in the polluted mixture, such as the concentration of a certain pollutant in sewage. The statistics can be based on the monitoring of pollution sensors installed at the sewage outlet.

[0093] After the controlled areas that do not meet the pollution discharge standards are monitored to have excessive emissions, they may take improper measures to evade verification in order to improve subsequent pollution discharges. One of such measures is to dilute the pollution discharge.

[0094] It should be noted that for the control areas of diluted pollution discharge, separate pollution alarms are issued, and the alarm priority is higher than that of other control areas. Feedback can be given to the user terminal or an alarm signal can be issued. The volume or color of the alarm prompt on the user terminal (compared with other control areas) should be higher, indicating that this control area should be given priority attention and treatment.

[0095] Wherein, judging whether there is diluted pollution discharge in the control area where pollution discharge does not meet the pollution discharge standard based on the emission monitoring concentration, the preset concentration amplitude, the net pollution discharge content of the stage and the estimated value of the change of the net pollution discharge content of the stage includes:

[0096] Determine whether the emission monitoring concentration in the control area where the pollution discharge does not meet the standards has decreased;

[0097] If the emission monitoring concentration of the control area where the pollution discharge does not meet the standards decreases, then monitor whether the decrease in the emission monitoring concentration of the control area where the pollution discharge does not meet the standards is greater than or equal to the preset concentration amplitude;

[0098] The pollutant monitoring concentration decreases. Only when it decreases to a certain level, the next step is executed. If the decrease is small, it may be caused by the error of the sensor and other monitoring, so further determination is required. If the decrease is greater than or equal to the preset concentration amplitude, it is considered that there is a significant decrease, and there is a possibility that dilution has caused the decrease.

[0099] If the decrease in the emission monitoring concentration of the control area that does not meet the pollution discharge standard is greater than or equal to the preset concentration amplitude, it is determined whether the increase in the net pollution discharge content of the corresponding control area at that stage is greater than the estimated value of the change in the net pollution discharge content at that stage;

[0100] When the decrease in the emission monitoring concentration is greater than or equal to the preset concentration amplitude, it is possible that the pollutant has been diluted. The net content of the pollutant discharged during the stage is a reasonable allowable pollutant discharge content, which is determined based on the standard data reported for production and the allowable emission data corresponding to the pollutant discharge standard under the production. Therefore, a large difference between the two indicates a serious exceedance of the standard, and it is very likely that the pollutant has been diluted. Therefore, it can be determined whether there is diluted pollutant discharge without obtaining the dilution liquid (such as water use data) (this type of data is often difficult to verify).

[0101] If so, it is determined that there is dilution discharge in the corresponding controlled area.

[0102] Specifically, when the increase in the net pollutant content during the second set time period is large, it means that it exceeds the normal amount under the production scale, such as increasing water consumption, resulting in an increase in the total amount of pollutants, which in turn causes a decrease in the emission monitoring concentration.

[0103] For example, the emission monitoring concentration decreased from A% to B%, and the decrease exceeded the preset concentration amplitude c%. Correspondingly, it was found that the net pollution content of the stage increased from D to E, which was greater than the estimated pollution discharge F of the stage. Therefore, the artificially extra water DE was the reason for the decrease in the emission monitoring concentration. According to relevant regulations on pollution prevention and control, it is not allowed to discharge the wastewater after dilution.

[0104] It should be noted that the preset values ​​in the present invention can be set according to actual needs and are not limited here.

[0105] When this embodiment is applied in practice, by selectively monitoring the increase in the net content of pollutants in the monitoring section when monitoring the emission concentration, it can be determined that there may be dilution of pollutants, and this control area can be distinguished, so that this hidden behavior can be effectively discovered.

[0106] As an optional embodiment of the present invention, generating a corresponding control plan according to the pollution identification result includes:

[0107] Based on the control area where the pollution discharge does not meet the standard at a future moment, an interception alarm notification is generated, and the interception alarm notification is sent to the supervision terminal;

[0108] Based on the controlled area that does not meet the pollution discharge standards at the current moment, a verification alarm notification is generated and sent to the supervision terminal.

[0109] When this embodiment is applied in practice, the control area that does not meet the pollution discharge standards in the future needs to remind the control personnel to intercept and deal with it in advance. For the current pollution discharge that does not meet the standards, it is necessary to verify and deal with it, such as notifying the environmental protection department and issuing warnings to production.

[0110] As an optional embodiment of the present invention, the corresponding control area includes the control area where pollution discharge does not meet the standards.

[0111] When this embodiment is applied in practice, the corresponding control areas mentioned above should include the control areas where pollution discharge does not meet the standards, that is, alarm schemes are generated first for the control areas where pollution discharge does not meet the standards. Of course, continuous alarm monitoring can also be carried out for some control areas where pollution discharge is considered normal or areas of key concern.

[0112] As an optional embodiment of the present invention, the performing pollution identification based on the fused feature information and generating a pollution identification result includes:

[0113] The fused feature information is input into an environmental management and control model, and the pollution identification result is output, wherein the environmental management and control model is pre-trained and includes a second deep learning network model.

[0114] Specifically, this model can include a neural network submodule named second_deep_learning_network, which is a pre-trained deep learning network. This subnetwork can be composed of three linear layers and an activation function. After training an environmental protection control model including a deep learning network, initialize the model, load the pre-trained network weights, input the fused feature information into the model, and generate pollution identification results.

[0115] The steps of pre-training the environmental protection control model include:

[0116] Obtaining environmental detection feature samples and pollution identification result labels of the environmental detection feature samples, and iteratively training the initial environmental protection control model until a first difference value between the output pollution identification result and the pollution identification result label meets a corresponding error standard, so as to obtain the trained environmental protection control model;

[0117] Alternatively, obtain an environmental detection feature sample corresponding to a prediction determining factor and a pollution identification result label of the environmental detection feature sample, wherein the prediction determining factor includes first time information corresponding to the environmental detection feature sample, a production plan corresponding to the environmental detection feature sample, and second time information corresponding to the production plan; based on the environmental detection feature sample and the pollution identification result label corresponding to the prediction determining factor, iteratively train the initial environmental protection control model until the second difference value between the output pollution identification result and the pollution identification result label meets the corresponding error standard to obtain the trained environmental protection control model.

[0118] In order to determine whether pollution discharge will meet the standards at a future moment, these predictive determinants (time and production plan) are used to train the model. This training enables the model to consider not only environmental testing samples and pollution labels, but also time and production plan information, making the model more comprehensive and accurate, and able to predict future pollution discharge situations.

[0119] Specifically, the model can be trained separately, and the environmental detection feature sample is input into the preset first model to be trained, and recognition is performed based on the first model to be trained to obtain a predicted recognition result. The predicted recognition result is calculated differently from the pollution recognition result label of the environmental detection feature sample to obtain an error result. Based on the error result, it is determined whether the error result meets the error standard indicated by the preset error threshold range. If the error result does not meet the error standard indicated by the preset error threshold range, return to the step of inputting the environmental detection feature sample into the preset first model to be trained, and recognition is performed based on the first model to be trained to obtain a predicted recognition result, and the training is stopped until the error result meets the error standard indicated by the preset error threshold range to obtain an environmental protection control model.

[0120] As an optional embodiment of the present invention, the information fusion model and the environmental protection control model are obtained by joint training in advance, and the steps of obtaining the information fusion model and the environmental protection control model by joint training in advance include:

[0121] Obtain multidimensional environmental detection samples and pollution identification result labels corresponding to the multidimensional environmental detection samples, and based on the multidimensional environmental detection samples and the pollution identification result labels corresponding to the multidimensional environmental detection samples, jointly train the initial information fusion model and the environmental protection control model to obtain the trained information fusion model and the environmental protection control model.

[0122] To identify whether the pollution discharge at the current moment meets the standards, samples and labels under normal circumstances can be used. Generally, there is no need to limit the time and production plan.

[0123] Alternatively, obtain the multidimensional environmental detection samples corresponding to the prediction determination factors and the pollution identification result labels corresponding to the multidimensional environmental detection samples, wherein the prediction determination factors include the first time information corresponding to the multidimensional environmental detection samples, the production plan corresponding to the multidimensional environmental detection samples, and the second time information corresponding to the production plan; based on the multidimensional environmental detection samples corresponding to the prediction determination factors and the pollution identification result labels corresponding to the multidimensional environmental detection samples, jointly train the initial information fusion model and the environmental protection control model to obtain the trained information fusion model and the environmental protection control model.

[0124] In order to determine whether pollution discharge will meet the standards at a future moment, these predictive determinants (time and production plan) are used to train the model. This training enables the model to consider not only environmental testing samples and pollution labels, but also time and production plan information, making the model more comprehensive and accurate, and able to predict future pollution discharge situations.

[0125] Specifically, the multidimensional environmental detection sample (or the multidimensional environmental detection sample corresponding to the prediction determinant) is input into the preset second model to be trained, and feature fusion processing is performed based on the second model to be trained to obtain the feature information after prediction fusion, wherein the second model to be trained is the initial model to be trained of the information fusion model. The feature information after prediction fusion is input into the preset first model to be trained, and recognition processing is performed based on the first model to be trained to obtain the prediction recognition result, wherein the first model to be trained is the initial model to be trained of the environmental protection control model, and the prediction recognition result is calculated with the pollution recognition result label of the multidimensional detection sample to obtain the error result. Based on the error result, it is judged whether the error result meets the error standard indicated by the preset error threshold range. If the error result does not meet the error standard indicated by the preset error threshold range, return to the step of inputting the multidimensional detection sample into the preset second model to be trained, performing feature fusion processing based on the second model to be trained, and obtaining the feature information after prediction fusion, until the error result meets the error standard indicated by the preset error threshold range, stop training, and obtain the information fusion model and the environmental protection control model.

[0126] The above joint training can improve the matching degree between models, save data training resources, and help reduce costs.

[0127] It should be noted that the multi-dimensional detection sample refers to a sample containing multi-dimensional detection information used for model training, specifically information consistent with the above detection information attributes. The pollution identification result label of the multi-dimensional detection sample refers to the label of the environmental pollution identification result corresponding to the multi-dimensional detection sample manually annotated.

[0128] The information fusion model can be trained by a deep neural network learning method using reinforcement learning, or can be trained by supervised and unsupervised deep neural network learning methods. The information fusion model is trained based on multi-dimensional reservation information samples, using reinforcement learning, or supervised or unsupervised methods for separate training. It can also be obtained by joint training with an environmental protection control model based on multi-dimensional detection samples. Specifically, the training material of the information fusion model (multi-dimensional detection samples) comes from the detection data uploaded historically by multiple sensors.

[0129] As an optional embodiment of the present invention, it also includes:

[0130] If the pollution discharge of the controlled area meets the standard at the current moment, the multi-dimensional environmental detection information and production scale data of the controlled area that meets the pollution discharge standard are counted and stored.

[0131] Specifically, when the pollution discharge in the controlled area meets the standards, it indicates that the emissions in the controlled area are qualified within a statistical control and testing cycle. At this time, the emission monitoring data and production scale data are recorded, and the two correspond to each other, indicating that the emission monitoring data under a certain production scale can be used as a comparison basis for subsequent pollution monitoring.

[0132] like Figure 2 As shown, the present invention further provides an intelligent environmental protection management and control system 200, which applies the intelligent environmental protection management and control method described in the above embodiment, including:

[0133] The detection information acquisition module 210 is used to obtain multi-dimensional environmental detection information of the control area.

[0134] The feature fusion module 220 is used to perform feature fusion on the multi-dimensional environment detection information to generate fused feature information.

[0135] The pollution identification module 230 is used to perform pollution identification based on the fused feature information and generate a pollution identification result, wherein the pollution identification result is used to characterize whether the pollution discharge of the controlled area at the current moment meets the standard, and / or whether the pollution discharge of the controlled area at a future moment meets the standard.

[0136] The control module 240 is used to generate a corresponding control plan based on the pollution identification result, wherein the control plan includes a pollution alarm based on the control area where the pollution discharge does not meet the standard.

[0137] The specific implementation of this embodiment can refer to the corresponding implementation method mentioned above and will not be described again here.

[0138] like Figure 3 As shown, an electronic device 300 provided in an embodiment of the present invention includes a memory 310 and a processor 320; the memory 310 is used to store computer programs; the processor 320 is used to implement the intelligent environmental protection management method as described above when executing the computer program.

[0139] In other words, an electronic device 300 includes a memory 310 and a processor 320 coupled to the memory 310; the memory 310 is configured to store a computer program; and the processor 320 is configured to perform the following operations when executing the computer program:

[0140] Obtain multi-dimensional environmental detection information of the controlled area;

[0141] Performing feature fusion on the multi-dimensional environment detection information to generate fused feature information;

[0142] Perform pollution identification based on the fused feature information to generate a pollution identification result, wherein the pollution identification result is used to indicate whether the pollution discharge of the control area at the current moment meets the standard, and / or whether the pollution discharge of the control area at a future moment meets the standard;

[0143] A corresponding control plan is generated based on the pollution identification result, wherein the control plan includes issuing a pollution warning based on the control area where pollution discharge does not meet the standards.

[0144] An embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the intelligent environmental protection management and control method as described above is implemented.

[0145] In other words, a non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor performs the following operations:

[0146] Obtain multi-dimensional environmental detection information of the controlled area;

[0147] Performing feature fusion on the multi-dimensional environment detection information to generate fused feature information;

[0148] Perform pollution identification based on the fused feature information to generate a pollution identification result, wherein the pollution identification result is used to indicate whether the pollution discharge of the control area at the current moment meets the standard, and / or whether the pollution discharge of the control area at a future moment meets the standard;

[0149] A corresponding control plan is generated based on the pollution identification result, wherein the control plan includes issuing a pollution warning based on the control area where pollution discharge does not meet the standards.

[0150] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0151] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

[0152] Although the present invention is disclosed as above, the protection scope of the present invention is not limited thereto. Those skilled in the art may make various changes and modifications without departing from the spirit and scope of the present invention, and these changes and modifications will fall within the protection scope of the present invention.

Claims

1. An intelligent environmental protection management and control method, characterized in that: include: Obtain multi-dimensional environmental detection information of the controlled area; Performing feature fusion on the multi-dimensional environment detection information to generate fused feature information; Perform pollution identification based on the fused feature information to generate a pollution identification result, wherein the pollution identification result is used to indicate whether the pollution discharge of the control area at the current moment meets the standard, and / or whether the pollution discharge of the control area at a future moment meets the standard; A corresponding control plan is generated based on the pollution identification result, wherein the control plan includes a pollution warning based on the control area where pollution discharge does not meet the standards.

2. The intelligent environmental protection management and control method according to claim 1 is characterized in that: The performing feature fusion on the multi-dimensional environment detection information to generate fused feature information comprises: The multi-dimensional environment detection information is input into an information fusion model to generate the fusion feature information, wherein the information fusion model is pre-trained and includes a first deep learning network model.

3. The intelligent environmental protection management and control method according to claim 2 is characterized in that: The performing pollution identification based on the fused feature information and generating a pollution identification result comprises: The fused feature information is input into an environmental management and control model, and the pollution identification result is output, wherein the environmental management and control model is pre-trained and includes a second deep learning network model.

4. The intelligent environmental protection management and control method according to claim 3 is characterized in that: The steps of pre-training the environmental protection control model include: Obtaining environmental detection feature samples and pollution identification result labels of the environmental detection feature samples, and iteratively training the initial environmental protection control model until a first difference value between the output pollution identification result and the pollution identification result label meets a corresponding error standard, so as to obtain the trained environmental protection control model; Alternatively, obtain the environmental detection feature sample corresponding to the prediction determination factor and the pollution identification result label of the environmental detection feature sample, wherein the prediction determination factor includes the first time information corresponding to the environmental detection feature sample, the production plan corresponding to the environmental detection feature sample, and the second time information corresponding to the production plan; based on the environmental detection feature sample and the pollution identification result label corresponding to the prediction determination factor, iteratively train the initial environmental protection control model until the second difference value between the output pollution identification result and the pollution identification result label meets the corresponding error standard to obtain the trained environmental protection control model.

5. The intelligent environmental protection management and control method according to claim 3 is characterized in that: The information fusion model and the environmental protection control model are obtained by joint training in advance, and the steps of obtaining the information fusion model and the environmental protection control model by joint training in advance include: Acquire multidimensional environmental detection samples and pollution identification result labels corresponding to the multidimensional environmental detection samples, and jointly train the initial information fusion model and the environmental protection control model based on the multidimensional environmental detection samples and the pollution identification result labels corresponding to the multidimensional environmental detection samples to obtain the trained information fusion model and the environmental protection control model; Alternatively, obtain the multidimensional environmental detection samples corresponding to the prediction determination factors and the pollution identification result labels corresponding to the multidimensional environmental detection samples, wherein the prediction determination factors include the first time information corresponding to the multidimensional environmental detection samples, the production plan corresponding to the multidimensional environmental detection samples, and the second time information corresponding to the production plan; based on the multidimensional environmental detection samples corresponding to the prediction determination factors and the pollution identification result labels corresponding to the multidimensional environmental detection samples, jointly train the initial information fusion model and the environmental protection control model to obtain the trained information fusion model and the environmental protection control model.

6. The intelligent environmental protection management and control method according to any one of claims 1 to 5, characterized in that: Generating a corresponding control plan according to the pollution identification result includes: Based on the control area where the pollution discharge does not meet the standard at a future moment, an interception alarm notification is generated, and the interception alarm notification is sent to the supervision terminal; Based on the controlled area that does not meet the pollution discharge standards at the current moment, a verification alarm notification is generated and sent to the supervision terminal.

7. The intelligent environmental protection management and control method according to claim 1 is characterized in that: The performing feature fusion on the multi-dimensional environment detection information to generate fused feature information comprises: Performing feature vectorization processing on the multi-dimensional environment detection information to obtain multiple feature vectors; Extracting statistical features from the plurality of feature vectors, wherein the statistical features include a mean and a variance; normalizing the statistical features; The standardized statistical features are concatenated to generate a unified feature vector, and the fused feature information includes the unified feature vector.

8. The intelligent environmental protection management and control method according to any one of claims 1 to 5, characterized in that: Generating a corresponding control plan according to the pollution identification result includes: Obtaining the emission monitoring concentration, preset concentration amplitude, stage net pollutant discharge content and stage net pollutant discharge content change estimate for the control area that does not meet the pollutant discharge standards within a preset time period; According to the emission monitoring concentration, the preset concentration amplitude, the net pollutant content of the stage and the estimated value of the change in the net pollutant content of the stage, it is judged whether there is diluted pollution discharge in the control area where the pollution discharge does not meet the standard; If so, a key inspection and alarm plan is generated based on the controlled area where diluted pollution exists.

9. The intelligent environmental protection management and control method according to claim 8 is characterized in that: The determining whether there is diluted pollution discharge in the control area where pollution discharge does not meet the pollution discharge standard according to the emission monitoring concentration, the preset concentration amplitude, the net pollution discharge content of the stage and the estimated value of the change of the net pollution discharge content of the stage includes: Determine whether the emission monitoring concentration in the control area where the pollution discharge does not meet the standards has decreased; If the emission monitoring concentration of the control area where the pollution discharge does not meet the standards decreases, then monitor whether the decrease in the emission monitoring concentration of the control area where the pollution discharge does not meet the standards is greater than or equal to the preset concentration amplitude; If the decrease in the emission monitoring concentration of the control area that does not meet the pollution discharge standard is greater than or equal to the preset concentration amplitude, it is determined whether the increase in the net pollution discharge content of the corresponding control area at that stage is greater than the estimated value of the change in the net pollution discharge content at that stage; If so, it is determined that there is dilution discharge in the corresponding controlled area.

10. An intelligent environmental protection management and control system, characterized in that: The intelligent environmental protection control method as claimed in claims 1 to 9 is applied, comprising: The detection information acquisition module is used to obtain multi-dimensional environmental detection information of the control area; A feature fusion module, used to perform feature fusion on the multi-dimensional environment detection information to generate fused feature information; A pollution identification module, used to perform pollution identification based on the fused feature information and generate a pollution identification result, wherein the pollution identification result is used to indicate whether the pollution discharge of the control area at the current moment meets the standard and / or whether the pollution discharge of the control area at a future moment meets the standard; A control module is used to generate a corresponding control plan according to the pollution identification result, wherein: The control plan includes issuing pollution warnings based on the control area where pollution discharge does not meet the standards.

Images