A visual monitoring system and method for pollution sources in industrial parks

Abstract

The present invention belongs to the field of data processing technology, and specifically relates to a system and method for visually monitoring pollution sources in industrial parks. The method includes installing water quality monitoring sensors in each enterprise in the industrial park, monitoring key indicators related to water quality in real time, transmitting the water quality data to an identification module via a wireless network, identifying and processing error values ​​in the water quality data, normalizing the water quality data, classifying the water quality data, calculating a pollution score, setting a pollution level, determining whether the water quality data is abnormal, and if so, comparing the water quality data with the normal range of key indicators, identifying indicators that exceed the standard, and marking them on a GIS map. If there are no abnormalities, a prediction model is trained based on historical data to predict whether abnormalities will occur in the future. The visualization module displays the water quality data and geographic location on a GIS map, specifically marking abnormal data, and providing graphical analysis of historical data. The present invention can accurately monitor pollution sources in industrial parks.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing, and in particular relates to a visual monitoring system and method for pollution sources in an industrial park. Background Art

[0002] Traditional pollution monitoring methods often rely on fixed monitoring sites, which have limited distribution and are difficult to fully cover every corner of the industrial park. In addition, traditional methods do not consider the accuracy of data when collecting data. A similar prior art Chinese patent application with publication number CN116010507A provides a water pollution monitoring method and system based on big data, which includes: analyzing and matching data collection methods according to water pollution factors to determine data collection ports; determining target water area characteristics; performing correlation analysis based on target water area characteristics and water pollution factors, determining target water pollution factors and corresponding target data collection ports based on the correlation, collecting data through the target data collection ports, and constructing a multivariate collection data source; performing standardized preprocessing on the multivariate collection data source to obtain a target data source; inputting the target data source into a water pollution analysis model to obtain water pollution analysis results. Similar prior art includes a Chinese patent application with publication number CN112559657A, which discloses a water pollution source tracing method, device and terminal equipment. The method includes: based on the online water quality monitoring data, obtaining the current monitoring point information where water pollution occurs and the first excessive pollutant concentration corresponding to the current monitoring point; and according to the current monitoring point information, obtaining each upstream monitoring point information corresponding to the current monitoring point and the tracing time corresponding to each upstream monitoring point; according to the online water quality monitoring data, determining the second excessive pollutant concentration of the upstream monitoring point corresponding to each tracing time; when the second excessive pollutant concentration is greater than the first excessive pollutant concentration, determining the upstream monitoring point corresponding to the second excessive pollutant concentration as the water pollution source point. However, when collecting water quality related data, the above two methods do not take into account the problem that water quality data may be erroneous. Therefore, the present invention provides a visual monitoring system and method for pollution sources in industrial parks. Summary of the Invention

[0003] The present invention installs water quality monitoring sensors in each enterprise in the industrial park to monitor key water quality indicators in real time, transmits water quality data to an identification module, identifies erroneous values ​​in the water quality data, and replaces them with key indicators corresponding to the previous set of water quality data. By detecting and replacing erroneous values, the accuracy of water quality data is improved. After normalization, the water quality data is classified and a pollution score is calculated for each water quality category. The pollution level is set to determine whether the water quality data is abnormal. The anomaly model is also used to predict future water quality conditions and formulate preventive measures in advance. Finally, the water quality data is intuitively displayed in combination with a GIS map, and abnormal data is identified using specific identifiers to help managers quickly understand the water quality conditions.

[0004] To achieve the above-mentioned purpose, the present invention provides a method for visually monitoring pollution sources in an industrial park as described below. The method is implemented by installing water quality monitoring sensors at the pollution source outlets of each enterprise in the industrial park and performing the following steps:

[0005] Step S1: real-time monitoring of key indicators related to water quality, combining multiple key indicators collected simultaneously into a set of water quality data, and transmitting the data to the recognition module in real time via a wireless network;

[0006] Step S2: Identify whether the key indicator in the water quality data has an erroneous value. If so, use the key indicator in the previous set of water quality data stored in the water quality database to replace the key indicator identified as an erroneous value. The identification module stores the water quality data in the water quality database, normalizes the water quality data in the water quality database, classifies the normalized water quality data into multiple water quality categories, calculates a pollution score for each water quality category, sets a pollution level for each water quality category based on the pollution score, determines the water quality category to which the newly acquired water quality data belongs and obtains the corresponding pollution level. If the pollution level is greater than a preset pollution level, it is determined that the water quality data is abnormal.

[0007] Step S3: If the water quality data is abnormal, a comparison database is pre-established, the comparison database stores the normal range of each key indicator, the key indicator is compared with the corresponding normal range, the key indicator that exceeds the normal range is identified, the geographical location of the corresponding sensor is obtained, the geographical location and the key indicator that exceeds the range are sent to the visualization module, if the water quality data is not abnormal, a prediction model is generated based on the water quality data in the water quality database, and whether an abnormality will occur in the future is determined based on the prediction model;

[0008] Step S4: The visualization module visualizes the real-time monitored water quality data and the corresponding geographic locations through a GIS map. Key indicators and corresponding geographic locations that are out of range are marked on the GIS map using specific markers. If anomalies are predicted in the future, different specific markers are also used to mark them on the GIS map. The visualization module also performs graphical analysis on historical water quality data.

[0009] As a preferred technical solution of the present invention, identifying whether the key indicator in the water quality data has an erroneous value includes the following steps:

[0010] The water quality data is used as target water quality data, historical water quality data stored in the water quality database is obtained, a key indicator in the target water quality data is obtained as the target key indicator, all historical key indicators corresponding to the target key indicator in the historical water quality data are obtained, all the historical key indicators are arranged in order of numerical value, if the number of all the historical key indicators is an odd number, the middle historical key indicator is used as a first reference value, otherwise, the average of the two middle historical key indicators is used as the first reference value, the difference between the target key indicator and the first reference value is calculated as a first difference, the difference between all the historical key indicators and the first reference value is also calculated as a second difference, the second reference value of all the second difference values ​​is calculated, and the standard value of the target key indicator is calculated using a first formula, the first formula is: S=|AB| / C, where S is the standard value, A is the target key indicator, B is the first reference value, and C is the second reference value. A first threshold is preset, and if the standard value is greater than the first threshold, the target key indicator is determined to be an error value.

[0011] As a preferred technical solution of the present invention, the normalized water quality data is classified into multiple water quality categories, including the following steps:

[0012] Step S21: presetting a predetermined number of water quality categories, randomly selecting a predetermined number of water quality data from all the water quality data as first water quality data, and randomly assigning the first water quality data to each of the water quality categories, that is, using the first water quality data as initial data corresponding to the water quality category;

[0013] Step S22: selecting a water quality data from the remaining water quality data in the water quality database as the second water quality data, calculating a first distance between the second water quality data and all the first water quality data, obtaining a minimum first distance among a plurality of first distances, obtaining the water quality category corresponding to the first water quality data corresponding to the minimum first distance, classifying the second water quality data into the water quality category, and updating the initial data of the water quality category based on the second water quality data;

[0014] Step S23: Repeat step S22 until all the water quality data have been selected.

[0015] As a preferred technical solution of the present invention, calculating the first distance between the second water quality data and all the first water quality data includes the following steps:

[0016] A second formula is used to calculate a first distance between the second water quality data and each of the first water quality data. The second formula is:

[0017]

[0018] Wherein, D is the first distance, Hi is the i-th key indicator of the first water quality data, Ei is the i-th key indicator of the second water quality data, and m refers to the total number of key indicators in the water quality data.

[0019] As a preferred technical solution of the present invention, updating the initial data of the water quality category based on the second water quality data includes the following steps:

[0020] The key indicator in the initial data is updated based on a third formula, where the third formula is:

[0021] Hn=Ho+(D min / D max )(E-Ho),

[0022] Among them, Ho is one of the key indicators in the initial data before updating, E is the corresponding key indicator in the second water quality data, Dmin is the minimum first distance, Dmax is the maximum first distance among all the first distances, and Hn is the corresponding key indicator of the initial data after updating.

[0023] As a preferred technical solution of the present invention, calculating the pollution score for each water quality category includes the following steps:

[0024] Based on historical data, different key indicators are assigned relevant values, wherein the relevant values ​​represent the importance of the key indicators to the pollution score. All water quality data in each water quality category are obtained, and the corresponding average key indicator is calculated for each key indicator. The pollution score of each water quality category is calculated using the fourth formula, which is:

[0025]

[0026] Among them, Gi refers to the average key indicator corresponding to the i-th key indicator, and Wi refers to the relevant value corresponding to the i-th key indicator.

[0027] As a preferred technical solution of the present invention, a prediction model is generated based on the water quality data in the water quality database, including the following steps:

[0028] Step S31: using the plurality of water quality data as first learning data, generating a first model through training, collecting other data related to water pollution, obtaining a plurality of water quality data other than the first learning data and the combination of the other data as second learning data, and generating a second model through training;

[0029] Step S32: obtaining the most recently collected water quality data as third water quality data, inputting the third water quality data into the first model to obtain first prediction data, calculating a first difference between the third water quality data and the first prediction data, obtaining the other data corresponding to the third water quality data, inputting the third water quality data and the other data into the second model to obtain second prediction data, and calculating a second difference between the second prediction data and the third water quality data;

[0030] Step S33: If the second difference is smaller than the first difference, the difference between the first difference and the second difference is calculated as a third difference, and the third difference is compared with a preset second threshold. If the third difference is larger than the second threshold, the second model and the first model are combined to generate a third model as the prediction model.

[0031] Step S34: If the second difference is greater than or equal to the first difference, continue to collect new other data related to water quality, use the new other data and some of the water quality data as new learning data to train the second model to generate a new second model, and return to step S32 until the second difference is less than the first difference.

[0032] As a preferred technical solution of the present invention, judging whether an anomaly will occur in the future based on the prediction model includes the following steps:

[0033] Based on the prediction model, predicted water quality data is obtained, the water quality category of the predicted water quality data is determined, a corresponding pollution score is obtained based on the water quality category, and a corresponding pollution level is obtained based on the pollution score. If the pollution level is greater than a preset pollution level, it is determined that an abnormal situation will occur in the future.

[0034] The present invention also provides a visual monitoring system for pollution sources in industrial parks, comprising the following modules:

[0035] The monitoring module is used to monitor key indicators related to water quality in real time, combine multiple key indicators collected simultaneously into a set of water quality data, and transmit it to the identification module in real time via the wireless network;

[0036] an identification module for identifying whether the key indicator in the water quality data has an erroneous value; if so, using the key indicator in the previous set of water quality data stored in the water quality database to replace the key indicator identified as an erroneous value; the identification module stores the water quality data in the water quality database, normalizes the water quality data in the water quality database, classifies the normalized water quality data into multiple water quality categories, calculates a pollution score for each water quality category, sets a pollution level for each water quality category based on the pollution score, determines the water quality category to which the newly acquired water quality data belongs and obtains the corresponding pollution level; and determines that the water quality data is abnormal if the pollution level is greater than a preset pollution level;

[0037] The prediction module, if the water quality data is abnormal, pre-establishes a comparison database, the comparison database stores the normal range of each key indicator, compares the key indicator with the corresponding normal range, identifies the key indicator that exceeds the normal range, obtains the geographical location of the corresponding sensor, sends the geographical location and the key indicator that exceeds the range to the visualization module, and if the water quality data is not abnormal, generates a prediction model based on the water quality data in the water quality database for training, and determines whether an abnormality will occur in the future based on the prediction model;

[0038] The visualization module is used to visualize the real-time monitored water quality data and the corresponding geographical locations through a GIS map. Key indicators and corresponding geographical locations that are out of range are marked on the GIS map using specific markers. When anomalies are predicted to occur in the future, different specific markers are also used to mark them on the GIS map. The visualization module also performs graphical analysis on historical water quality data.

[0039] Compared with the prior art, the beneficial effects of the present invention are at least as follows:

[0040] In the present invention, firstly, water quality data is monitored in real time in an industrial park according to water quality monitoring sensors, and the collected water quality data is identified, key indicators with errors are identified, and replaced with key indicators collected last time, thereby improving the accuracy of the data and providing data support for subsequent analysis and prediction. When it is determined that the water quality data is abnormal, the water quality data is further compared and analyzed, and the key indicators that exceed the normal range are obtained by comparing with the normal range of each key indicator stored in the comparison database. The installation location of the sensor corresponding to the key indicator, that is, the geographical location, is also obtained, and the key indicator and the geographical location are sent to the visualization module. When there is no abnormality in the water quality data, the prediction model is used to predict future abnormal conditions, and the abnormal conditions and the corresponding geographical locations are displayed through the visualization module. The historical water quality data is also analyzed graphically to help managers intuitively understand the water quality data and water quality trends. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a flowchart of the steps of a method for visually monitoring pollution sources in an industrial park according to the present invention;

[0042] Figure 2 This is a structural diagram of the composition of a visual monitoring system for pollution sources in an industrial park according to the present invention. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0044] It is understood that the terms "first," "second," etc., used herein may be used to describe various elements, but unless otherwise specified, these elements are not limited by these terms. These terms are only used to distinguish a first element from another element. For example, a first xx script may be referred to as a second xx script, and similarly, a second xx script may be referred to as a first xx script without departing from the scope of this application.

[0045] The present invention provides Figure 1 The method for visually monitoring pollution sources in an industrial park is shown, which is implemented by installing water quality monitoring sensors at the pollution source outlets of each enterprise in the industrial park and performing the following steps:

[0046] Step S1: Real-time monitoring of key indicators related to water quality, combining multiple key indicators collected simultaneously into a set of water quality data, and transmitting the data to the identification module in real time via a wireless network.

[0047] Specifically, in order to conduct visual real-time monitoring of pollution sources in the industrial park, water quality monitoring sensors are installed at the pollution source discharge outlets of each enterprise in the industrial park. The number of sensors can be multiple, which are used to monitor key indicators related to water quality in real time. Key indicators include physical parameters such as turbidity, temperature and flow rate, chemical parameters such as chemical oxygen demand, dissolved oxygen, suspended solids and pH value, metal ions such as heavy metal content such as lead, mercury and chromium, nutrient components such as nitrogen content and phosphorus content, microorganisms such as total bacteria and specific pathogens, etc. For example, if different water quality data are collected from multiple sensors in the same time period, these different water quality data are combined into a set of water quality data and transmitted to the recognition module in real time through the wireless network, and the recognition module processes and analyzes these water quality data.

[0048] Step S2: Identify whether there are erroneous values ​​in the key indicators in the water quality data. If so, use the key indicators in the previous set of water quality data stored in the water quality database to replace the key indicators identified as erroneous values. The identification module stores the water quality data in the water quality database, normalizes the water quality data in the water quality database, classifies the normalized water quality data into multiple water quality categories, calculates the pollution score of each water quality category, sets a pollution level for each water quality category based on the pollution score, determines the water quality category to which the newly acquired water quality data belongs and obtains the corresponding pollution level. If the pollution level is greater than the preset pollution level, it is determined that the water quality data is abnormal.

[0049] Specifically, errors may occur in the process of collecting and transmitting data. Analyzing erroneous water quality data will directly affect the analysis results and lead to inaccurate analysis. In order to effectively identify and remove erroneous values, improve data quality, and provide more accurate data for subsequent analysis and prediction, it is necessary to identify whether erroneous values ​​appear in the water quality data. The specific identification method will be explained in detail later. If so, the key indicators in the previous set of water quality data stored previously are used to replace the error values, and then the water quality data without errors is stored in the water quality database for easy subsequent acquisition. Different key indicators correspond to different values. In order to analyze the water quality data more quickly, the water quality data is also normalized. Based on the normalized water quality data, the water quality data is divided into multiple water quality categories, and the pollution score of each water quality category is calculated. The pollution level is set for each water quality category based on the pollution score. For example, 6 levels are set, namely 1-6. The larger the level, the greater the degree of pollution. Later, when new water quality data is obtained, the water quality category of the water quality data can be judged, and the pollution level of the water quality data can be obtained based on the water quality category. When the pollution level is greater than the preset pollution level, it is judged that the water quality data is abnormal, such as excessive emissions.

[0050] Step S3: If the water quality data is abnormal, a comparison database is established in advance. The comparison database stores the normal range of each key indicator, compares the key indicator with the corresponding normal range, identifies the key indicator that exceeds the normal range, obtains the geographical location of the corresponding sensor, and sends the geographical location and the key indicator that exceeds the range to the visualization module. If the water quality data is normal, a prediction model is generated based on the water quality data training in the water quality database, and whether an abnormality will occur in the future is determined based on the prediction model.

[0051] Specifically, when it is determined that the water quality data is abnormal, the water quality data is further compared and analyzed. By comparing it with the normal range of each key indicator stored in the comparison database, the key indicators that are beyond the normal range are obtained, and the installation location of the sensor corresponding to the key indicator, that is, the geographical location, is also obtained. The key indicator and the geographical location are sent to the visualization module. If it is determined that the water quality data is not abnormal, a prediction model is generated based on the water quality data stored in the water quality database. The specific process of training the prediction model will be explained in detail later. Based on the prediction model, it is determined whether an abnormality will be sent. If so, the relevant management personnel can be notified to take measures. The above method can quickly and accurately analyze abnormal water quality data, and use the prediction model to predict future water quality data.

[0052] Step S4: The visualization module visualizes the real-time monitored water quality data and the corresponding geographic locations through a GIS map. Key indicators and corresponding geographic locations that are out of range are marked on the GIS map using specific markers. If anomalies are predicted in the future, different specific markers are also used to mark them on the GIS map. The visualization module also performs graphical analysis on historical water quality data.

[0053] Specifically, to help managers understand the situation more intuitively, the visualization module uses a GIS map to visualize the real-time monitored water quality data and the corresponding geographical locations. Key indicators that are out of range and the corresponding geographical locations are marked with specific markers, such as a prominent red five-pointed star. When anomalies are predicted in the future, different specific markers, such as a green triangle, are used to mark them on the GIS map to help managers quickly locate the source of pollution and key indicators that exceed the standard. The visualization module also performs graphical analysis of historical water quality data to help managers understand pollution trends.

[0054] Furthermore, identifying whether key indicators in water quality data have erroneous values ​​includes the following steps:

[0055] Take water quality data as target water quality data, obtain historical water quality data stored in a water quality database, obtain a key indicator in the target water quality data as the target key indicator, obtain all corresponding historical key indicators in the historical water quality data for the target key indicator, arrange all historical key indicators in order of numerical value, if the number of all historical key indicators is an odd number, use the middle historical key indicator as the first reference value, otherwise, use the average value of the two middle historical key indicators as the first reference value, calculate the difference between the target key indicator and the first reference value as the first difference, also calculate the difference between all historical key indicators and the first reference value as the second difference, calculate the second reference value of all second differences, and use the first formula to calculate the standard value of the target key indicator. The first formula is: S = |AB| / C, where S is the standard value, A is the target key indicator, B is the first reference value, and C is the second reference value. A first threshold is preset. If the standard value is greater than the first threshold, the target key indicator is judged as an error value.

[0056] Specifically, in order to identify the erroneous values ​​of key indicators in water quality data, historical water quality data in the water quality database is obtained. In order to improve the accuracy of identification, each key indicator is analyzed separately. For a target key indicator in the target water quality data, all corresponding historical key indicators in the historical water quality data are obtained, and all historical key indicators are arranged in order of the numerical value of the key indicators. If the number of all historical key indicators is an odd number, the historical key indicator in the middle is obtained as the first reference value. Otherwise, the average value of the two historical key indicators in the middle is used as the first reference value. The difference between the target key indicator and the first reference value is calculated as the first difference. The difference between all historical key indicators and the first reference value is also calculated as the second difference. The difference between all historical key indicators and the first reference value is also calculated as the second difference. The second reference value of the second difference, that is, the median of all second differences, is calculated in the same way as the first reference value. The standard value corresponding to the target key indicator is calculated using the above first formula, and the first threshold is determined based on historical experience or statistical methods. If the standard value is greater than the first threshold, the target key indicator is judged to be an error value, otherwise the target key indicator is a correct value. The above method is used to determine whether other target key indicators in the target water quality data are error values. The above method uses the median of all historical key indicators as the first reference value, because the median has good resistance to error values ​​and is not as easily affected by extreme values ​​as the mean and standard deviation. Therefore, using the median as a reference value to identify error values ​​can improve the stability of error value identification.

[0057] Furthermore, the normalized water quality data is classified into multiple water quality categories, including the following steps:

[0058] Step S21: pre-set a predetermined number of water quality categories, randomly select a predetermined number of water quality data from all water quality data as first water quality data, and randomly assign the first water quality data to each water quality category, that is, use the first water quality data as initial data for the corresponding water quality category;

[0059] Step S22: Select one water quality data from the remaining water quality data in the water quality database as the second water quality data, calculate the first distance between the second water quality data and all the first water quality data, obtain the minimum first distance among the multiple first distances, obtain the water quality category corresponding to the first water quality data corresponding to the minimum first distance, classify the second water quality data into the water quality category, and update the initial data of the water quality category based on the second water quality data;

[0060] Step S23: Repeat step S22 until all water quality data have been selected.

[0061] Specifically, in order to quickly and accurately classify water quality data, a predetermined number of water quality categories are pre-set, for example, 10 categories are pre-set, and 10 water quality data are randomly selected from the water quality data as first water quality data. The first water quality data are used as initial data for each water quality category. Then, one water quality data is selected from the remaining water quality data as second water quality data, and the first distance between the second water quality data and each first water quality data is calculated to obtain multiple first distances. The specific method for calculating the first distance will be introduced later. The first distance represents the similarity between the two water quality data. The smaller the first distance, the greater the similarity between the two water quality data, and the more likely they are to be classified into the same category. Therefore, the second water quality data is classified into the water quality category corresponding to the first water quality data corresponding to the latest first distance. Then, the initial data of the corresponding water quality category is updated based on the second water quality data, so that the two water quality data are closer, which facilitates more accurate classification of the remaining other water quality data. The specific method for updating the initial data will be explained in detail later. Repeat the above step S22 until all water quality data are classified. The above method can quickly and accurately classify water quality data.

[0062] Furthermore, calculating the first distance between the second water quality data and all the first water quality data includes the following steps:

[0063] The first distance between the second water quality data and each first water quality data is calculated using the second formula, which is:

[0064]

[0065] Among them, D is the first distance, Hi is the i-th key indicator of the first water quality data, Ei is the i-th key indicator of the second water quality data, and m refers to the total number of key indicators in the water quality data.

[0066] Specifically, calculating the first distance between the second water quality data and each first water quality data through the above formula, determining the first water quality data that is closer to the second water quality, and classifying the second water quality data into the corresponding water quality category can help us quickly classify the water quality data.

[0067] Furthermore, updating the initial data of the water quality category based on the second water quality data includes the following steps:

[0068] Update the key indicators in the initial data based on the third formula, which is:

[0069] Hn=Ho+(D min / D max )(E-Ho),

[0070] Among them, Ho is a key indicator in the initial data before updating, E is the corresponding key indicator in the second water quality data, Dmin is the minimum first distance, Dmax is the maximum first distance among all first distances, and Hn is the corresponding key indicator of the initial data after updating.

[0071] Specifically, since the initial data is randomly selected, each time a water quality data is added to the data of the same category, the initial data is updated using the above formula so that the initial data better matches the newly added water quality data. Through the above method, as the data in the same category increases, the neighborhood in the same category of data is gradually reduced, making the water quality data in the same category more and more similar, and finally completing the classification of the water quality data.

[0072] Furthermore, the pollution score for each water quality category is calculated, including the following steps:

[0073] Based on historical data, different key indicators are assigned relevant values. The relevant values ​​represent the importance of the key indicators to the pollution score. All water quality data in each water quality category are obtained, and the corresponding average key indicator is calculated for each key indicator. The pollution score of each water quality category is calculated using the fourth formula. The fourth formula is:

[0074]

[0075] Among them, Gi refers to the average key indicator corresponding to the i-th key indicator, and Wi refers to the relevant value corresponding to the i-th key indicator.

[0076] Specifically, since the water quality data of each water quality category are very similar, a pollution score is calculated for each category, and the pollution score of each water quality category is used to represent the pollution score of each water quality data in each water quality category. First, relevant values ​​are assigned to different key indicators based on historical data. The relevant values ​​represent the importance of the key indicators to the calculation of the pollution score. In other words, the relevant values ​​represent the importance of the key indicators to the evaluation of the degree of pollution. Then, all water quality data of each water quality category are obtained, and the corresponding average key indicator is calculated for each key indicator. The pollution score of each water quality category is calculated using the above formula. Through the above method, the pollution score of each water quality category can be calculated quickly and accurately.

[0077] Furthermore, a prediction model is generated based on water quality data training in a water quality database, including the following steps:

[0078] Step S31: using a plurality of water quality data as first learning data, generating a first model through training, collecting other data related to water pollution, obtaining a plurality of water quality data other than the first learning data and other data combinations as second learning data, and generating a second model through training;

[0079] Step S32: obtaining the latest collected water quality data as third water quality data, inputting the third water quality data into the first model to obtain first predicted data, calculating a first difference between the third water quality data and the first predicted data, obtaining other data corresponding to the third water quality data, inputting the third water quality data and the other data into the second model to obtain second predicted data, and calculating a second difference between the second predicted data and the third water quality data;

[0080] Step S33: If the second difference is smaller than the first difference, the difference between the first difference and the second difference is calculated as a third difference, and the third difference is compared with a preset second threshold. If the third difference is greater than the second threshold, the second model and the first model are combined to generate a third model as a prediction model.

[0081] Step S34: If the second difference is greater than or equal to the first difference, continue to collect new other data related to water quality, use the new other data and some water quality data as new learning data to train the second model to generate a new second model, and return to step S32 until the second difference is less than the first difference.

[0082] Specifically, in order to predict the water quality conditions, it is necessary to train a prediction model. In order to quickly train a prediction model with good prediction performance and high prediction accuracy and reduce the cost of training the model, first obtain some water quality data as the first learning data, train the first learning data to generate a first model, and also obtain other data related to water pollution, such as meteorological data, obtain water quality data other than the first learning data and other data in combination to generate second learning data, train the second learning data to generate a second model, and then obtain the latest water quality data, that is, the most recently collected water quality data as the third water quality data, input the third water quality data into the first model to output the first prediction data, input the third water quality data and the corresponding other data into the second model to output the second prediction data, calculate the first difference between the first prediction data and the latest water quality data, calculate the second difference between the second prediction data and the latest water quality data, if the second difference is less than the first difference, it means that the prediction accuracy of the second model is higher, and then calculate the difference between the first difference and the second difference as the third difference, and compare the third difference with the preset The second threshold value of the third difference is greater than the second threshold value, which means that the prediction accuracy of the second model is much higher than that of the first model, and it also means that other data have a good improvement on the prediction result. Therefore, the second model and the first model are combined to generate the third model as the final prediction model, so that the prediction model has a higher prediction accuracy. If the second difference is greater than or equal to the first difference, it means that the second model does not improve the prediction accuracy compared with the first model, and it also means that other data have no improvement on the prediction result. Therefore, other new data related to water pollution are continuously obtained, such as industrial production data, and the new other data are combined with the water quality data as new learning data to train the second model to generate a new second model. Then, return to step S32 until the second difference is less than the first difference, that is, the new second model and the first model have obvious prediction accuracy. The second model and the first model are combined to generate the final prediction model. The prediction accuracy of the prediction model is improved by continuously using new other data through the above method, so as to achieve the purpose of improving the prediction accuracy of the prediction model.

[0083] Furthermore, it is characterized in that judging whether an abnormality will occur in the future based on the prediction model includes the following steps:

[0084] Based on the prediction model, the predicted water quality data is obtained, the water quality category of the predicted water quality data is determined, the corresponding pollution score is obtained based on the water quality category, and the corresponding pollution level is obtained based on the pollution score. If the pollution level is greater than the preset pollution level, it is determined that an abnormal situation will occur in the future.

[0085] Specifically, the prediction model is used to obtain future predicted water quality data, analyze the predicted water quality data to determine the water quality category of the predicted water quality data, obtain the pollution score corresponding to the water quality category, and obtain the corresponding pollution level based on the pollution score. If the pollution level is greater than the preset pollution level, it means that the future water quality data is abnormal, which may cause one or several key indicators to be abnormal. The prediction results are sent to the visualization module in time and displayed by the visualization module to help administrators detect abnormal situations in advance so that preventive measures can be taken.

[0086] According to another aspect of the embodiment of the present invention, Figure 2 As shown, a visual monitoring system for pollution sources in industrial parks is also provided, including a monitoring module, an identification module, a prediction module, and a visualization module, which are used to implement the visual monitoring method for pollution sources in industrial parks described above. The specific functions of each module are as follows:

[0087] The monitoring module is used to monitor key indicators related to water quality in real time, combine multiple key indicators collected simultaneously into a set of water quality data, and transmit it to the identification module in real time via the wireless network;

[0088] an identification module for identifying whether a key indicator in the water quality data has an erroneous value; if so, replacing the key indicator identified as an erroneous value with a key indicator in a previous set of water quality data stored in the water quality database; the identification module stores the water quality data in the water quality database, normalizes the water quality data in the water quality database, classifies the normalized water quality data into multiple water quality categories, calculates a pollution score for each water quality category, sets a pollution level for each water quality category based on the pollution score, determines the water quality category to which the newly acquired water quality data belongs and obtains the corresponding pollution level; and determines that the water quality data is abnormal if the pollution level is greater than a preset pollution level;

[0089] The prediction module, if the water quality data is abnormal, pre-establishes a comparison database that stores the normal range of each key indicator, compares the key indicator with the corresponding normal range, identifies the key indicator that exceeds the normal range, obtains the geographic location of the corresponding sensor, and sends the geographic location and the key indicator that exceeds the range to the visualization module. If the water quality data is normal, a prediction model is generated based on the water quality data in the water quality database to determine whether an abnormality will occur in the future based on the prediction model;

[0090] The visualization module is used to visualize the real-time monitored water quality data and the corresponding geographic locations through GIS maps. Key indicators and corresponding geographic locations that are out of range are marked with specific markers on the GIS map. When anomalies are predicted in the future, different specific markers are also used to mark them on the GIS map. The visualization module also performs graphical analysis of historical water quality data.

[0091] In summary, the present invention provides a system and method for visually monitoring pollution sources in an industrial park. The method includes installing water quality monitoring sensors in each enterprise in the industrial park, monitoring key indicators related to water quality in real time, transmitting the water quality data to an identification module via a wireless network, identifying erroneous values ​​in the water quality data, and replacing them with corresponding key indicators in the previous set of water quality data. The water quality data is normalized and classified, and a pollution score is calculated. The pollution level is set, and it is determined whether the water quality data is abnormal. If the water quality data is abnormal, the normal range of the key indicators is compared, the indicators exceeding the standard are identified, and marked on a GIS map. If there is no abnormality, a prediction model is trained based on historical data to predict whether an abnormality will occur in the future. The visualization module displays the water quality data and geographic location on a GIS map, specifically marks the abnormal data, and provides a graphical analysis of the historical data. The present invention can accurately monitor the pollution sources in the industrial park.

[0092] It should be understood that, although the various steps in the flow chart of each embodiment of the present invention are shown in sequence according to the indication of the arrows, these steps are not necessarily performed in sequence according to the order indicated by the arrows. Unless otherwise specified herein, the execution of these steps is not strictly limited in order, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0093] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program. The above-mentioned 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 in this application 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).

[0094] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0095] The above embodiments merely represent several implementation methods of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art may make various modifications and improvements without departing from the scope of the present invention, and these modifications and improvements fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be based on the appended claims.

[0096] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for visually monitoring pollution sources in industrial parks, wherein water quality monitoring sensors are installed at the pollution source outlets of each enterprise in the industrial park, characterized in that: The steps include: Step S1: Real-time monitoring of key indicators related to water quality, combining multiple key indicators collected simultaneously into a set of water quality data, and transmitting the data to the recognition module in real time via a wireless network; Step S2: Identify whether the key indicator in the water quality data has an erroneous value. If so, use the key indicator in the previous set of water quality data stored in the water quality database to replace the key indicator identified as an erroneous value. The identification module stores the water quality data in the water quality database, normalizes the water quality data in the water quality database, classifies the normalized water quality data into multiple water quality categories, calculates a pollution score for each water quality category, sets a pollution level for each water quality category based on the pollution score, determines the water quality category to which the newly acquired water quality data belongs and obtains the corresponding pollution level. If the pollution level is greater than a preset pollution level, it is determined that the water quality data is abnormal. Step S3: If the water quality data is abnormal, a comparison database is pre-established, the comparison database stores the normal range of each key indicator, the key indicator is compared with the corresponding normal range, the key indicator that exceeds the normal range is identified, the geographical location of the corresponding sensor is obtained, the geographical location and the key indicator that exceeds the range are sent to the visualization module, if the water quality data is not abnormal, a prediction model is generated based on the water quality data in the water quality database, and whether an abnormality will occur in the future is determined based on the prediction model; The process of training and generating a prediction model based on the water quality data in the water quality database includes the following steps: Step S31: using the plurality of water quality data as first learning data, generating a first model through training, collecting other data related to water pollution, obtaining a plurality of water quality data other than the first learning data and the combination of the other data as second learning data, and generating a second model through training; Step S32: obtaining the most recently collected water quality data as third water quality data, inputting the third water quality data into the first model to obtain first prediction data, calculating a first difference between the third water quality data and the first prediction data, obtaining the other data corresponding to the third water quality data, inputting the third water quality data and the other data into the second model to obtain second prediction data, and calculating a second difference between the second prediction data and the third water quality data; Step S33: If the second difference is smaller than the first difference, the difference between the first difference and the second difference is calculated as a third difference, and the third difference is compared with a preset second threshold. If the third difference is larger than the second threshold, the second model and the first model are combined to generate a third model as the prediction model. Step S34: If the second difference is greater than or equal to the first difference, continue to collect new other data related to water quality, use the new other data and some of the water quality data as new learning data to train the second model to generate a new second model, and return to step S32 until the second difference is less than the first difference; Step S4: The visualization module visualizes the real-time monitored water quality data and the corresponding geographic locations through a GIS map. Key indicators and corresponding geographic locations that are out of range are marked on the GIS map using specific markers. If anomalies are predicted in the future, different specific markers are also used to mark them on the GIS map. The visualization module also performs graphical analysis on historical water quality data.

2. The method according to claim 1, characterized in that Identifying whether the key indicator in the water quality data has an erroneous value includes the following steps: The water quality data is used as target water quality data, historical water quality data stored in the water quality database is obtained, a key indicator in the target water quality data is obtained as the target key indicator, all historical key indicators corresponding to the target key indicator in the historical water quality data are obtained, all the historical key indicators are arranged in order of numerical value, if the number of all the historical key indicators is an odd number, the middle historical key indicator is used as a first reference value, otherwise, the average of the two middle historical key indicators is used as the first reference value, the difference between the target key indicator and the first reference value is calculated as a first difference, the difference between all the historical key indicators and the first reference value is also calculated as a second difference, the second reference value of all the second difference values ​​is calculated, and the standard value of the target key indicator is calculated using a first formula, the first formula is: S=|AB| / C, where S is the standard value, A is the target key indicator, B is the first reference value, and C is the second reference value. A first threshold is preset, and if the standard value is greater than the first threshold, the target key indicator is determined to be an error value.

3. The method according to claim 1, characterized in that Classifying the normalized water quality data into multiple water quality categories includes the following steps: Step S21: presetting a predetermined number of water quality categories, randomly selecting a predetermined number of water quality data from all the water quality data as first water quality data, and randomly assigning the first water quality data to each of the water quality categories, that is, using the first water quality data as initial data corresponding to the water quality category; Step S22: selecting a water quality data from the remaining water quality data in the water quality database as the second water quality data, calculating a first distance between the second water quality data and all the first water quality data, obtaining a minimum first distance among a plurality of first distances, obtaining the water quality category corresponding to the first water quality data corresponding to the minimum first distance, classifying the second water quality data into the water quality category, and updating the initial data of the water quality category based on the second water quality data; Step S23: Repeat step S22 until all the water quality data have been selected.

4. The method according to claim 3, characterized in that Calculating a first distance between the second water quality data and all the first water quality data includes the following steps: A second formula is used to calculate a first distance between the second water quality data and each of the first water quality data. The second formula is: ， Wherein, D is the first distance, Hi is the i-th key indicator of the first water quality data, Ei is the i-th key indicator of the second water quality data, and m refers to the total number of key indicators in the water quality data.

5. The method according to claim 3, characterized in that Updating the initial data of the water quality category based on the second water quality data comprises the following steps: The key indicator in the initial data is updated based on a third formula, where the third formula is: ， Among them, Ho is one of the key indicators in the initial data before updating, E is the corresponding key indicator in the second water quality data, Dmin is the minimum first distance, Dmax is the maximum first distance among all the first distances, and Hn is the corresponding key indicator of the initial data after updating.

6. The method according to claim 1, characterized in that Calculating the pollution score for each water quality category involves the following steps: Based on historical data, different key indicators are assigned relevant values, wherein the relevant values ​​represent the importance of the key indicators to the pollution score. All water quality data in each water quality category are obtained, and the corresponding average key indicator is calculated for each key indicator. The pollution score of each water quality category is calculated using the fourth formula, which is: , Among them, Gi refers to the average key indicator corresponding to the i-th key indicator, and Wi refers to the relevant value corresponding to the i-th key indicator.

7. The method according to claim 1, characterized in that Determining whether an anomaly will occur in the future based on the prediction model includes the following steps: Based on the prediction model, predicted water quality data is obtained, the water quality category of the predicted water quality data is determined, a corresponding pollution score is obtained based on the water quality category, and a corresponding pollution level is obtained based on the pollution score. If the pollution level is greater than a preset pollution level, it is determined that an abnormal situation will occur in the future.

8. A visual monitoring system for pollution sources in an industrial park, used to implement a visual monitoring method for pollution sources in an industrial park according to any one of claims 1 to 7, characterized in that: Includes the following modules: The monitoring module is used to monitor key indicators related to water quality in real time, combine multiple key indicators collected simultaneously into a set of water quality data, and transmit it to the identification module in real time via the wireless network; an identification module for identifying whether the key indicator in the water quality data has an erroneous value; if so, using the key indicator in the previous set of water quality data stored in the water quality database to replace the key indicator identified as an erroneous value; the identification module stores the water quality data in the water quality database, normalizes the water quality data in the water quality database, classifies the normalized water quality data into multiple water quality categories, calculates a pollution score for each water quality category, sets a pollution level for each water quality category based on the pollution score, determines the water quality category to which the newly acquired water quality data belongs and obtains the corresponding pollution level; and determines that the water quality data is abnormal if the pollution level is greater than a preset pollution level; The prediction module, if the water quality data is abnormal, pre-establishes a comparison database, the comparison database stores the normal range of each key indicator, compares the key indicator with the corresponding normal range, identifies the key indicator that exceeds the normal range, obtains the geographical location of the corresponding sensor, and sends the geographical location and the key indicator that exceeds the range to the visualization module. If the water quality data is not abnormal, a prediction model is generated based on the water quality data in the water quality database, and whether an abnormality will occur in the future is judged based on the prediction model. The prediction model is generated based on the water quality data in the water quality database, including the following steps: step S31, taking some of the water quality data as the first learning data, generating a first model through training, collecting other data related to water pollution, obtaining some of the water quality data other than the first learning data and the combination of the other data as the second learning data, and generating a second model through training; step S32, obtaining the latest collected water quality data as the third water quality data, and taking the third Input three water quality data into the first model to obtain first prediction data, calculate the first difference between the third water quality data and the first prediction data, obtain the other data corresponding to the third water quality data, input the third water quality data and the other data into the second model to obtain second prediction data, and calculate the second difference between the second prediction data and the third water quality data; step S33, if the second difference is less than the first difference, calculate the difference between the first difference and the second difference as the third difference, compare the third difference with a preset second threshold, if the third difference is greater than the second threshold, combine the second model and the first model to generate a third model as the prediction model; step S34, if the second difference is greater than or equal to the first difference, continue to collect new other data related to water quality, use the new other data and several water quality data as new learning data to train the second model to generate a new second model, and return to step S32 until the second difference is less than the first difference; The visualization module is used to visualize the real-time monitored water quality data and the corresponding geographical locations through a GIS map. Key indicators and corresponding geographical locations that are out of range are marked on the GIS map using specific markers. When anomalies are predicted to occur in the future, different specific markers are also used to mark them on the GIS map. The visualization module also performs graphical analysis on historical water quality data.

Images