Fixed pollution source monitoring data analysis method, system and medium based on the Internet of Things

Abstract

The present application relates to the field of environmental monitoring technology, and discloses a method, system, and medium for analyzing fixed pollution source monitoring data based on the Internet of Things. The method includes: deploying Internet of Things devices at pollution source monitoring points to collect monitoring data in real time, and sending the monitoring data to a data processing module for processing; extracting first monitoring data from historical monitoring data for training a first model, and simultaneously training multiple second models based on different feature quantities of the first monitoring data, and training multiple third models based on different types of second monitoring data; after obtaining real-time monitoring data, first determine whether it is abnormal through the first model, and if abnormal, use the second model to determine the type of abnormality, and use the third model to find the cause of the abnormality, and finally generate early warning information. The present invention can improve the efficiency and accuracy of fixed pollution source monitoring and provide strong support for environmental monitoring.

Description

Technical Field

[0001] The present application relates to the field of environmental monitoring technology, and in particular to a method, system, and medium for analyzing fixed pollution source monitoring data based on the Internet of Things. Background Art

[0002] With the acceleration of industrialization, emissions from stationary pollution sources, such as industrial parks, are increasingly impacting the atmospheric environment. Traditional pollution source monitoring methods rely primarily on manual sampling and laboratory analysis, which suffer from low efficiency, poor real-time performance, and incomplete data. In recent years, the development of the Internet of Things (IoT) has provided new means for pollution source monitoring. By deploying a large number of sensors at pollution sources and in the environment, real-time and continuous monitoring of pollutants is possible.

[0003] A similar prior art Chinese patent application with publication number CN119619433A provides an online pollution source monitoring and analysis system and method, including: a map construction unit for three-dimensional map modeling of the target monitoring area; a human-computer interaction unit for issuing monitoring control commands; a collection and monitoring unit for calculating data sampling density and environmental data collection; a threshold warning unit for comparing data to determine whether it is a suspected pollution source; a pollution confirmation unit for setting an identification framework for the received environmental data of the suspected pollution source to determine whether the suspected pollution source is a real pollution source; and a diffusion analysis unit for matching the environmental data of the real pollution source with the pollutant diffusion model to obtain the diffusion range of the real pollution source.

[0004] Similar prior art includes a Chinese patent application with publication number CN118839175A, which provides a method and system for online pollution source monitoring. The method comprises: obtaining target pollution area data and performing regional segmentation through simulation model restoration processing to obtain simulated segmented pollution areas; then, using a uniform array of measurement points to measure oxygen flux in the segmented areas and analyze flux fluctuations to generate a normal fluctuation range; based on this fluctuation range, the pollution source is tracked back to locate the initial pollution source data. Subsequently, the initial pollution source is regionally calibrated based on the oxygen flux data of each area, and candidate directions of pollution diffusion are calculated; finally, based on the diffusion direction and the initial positioning data, the pollution area is dynamically monitored and adjusted to generate the final pollution source monitoring results.

[0005] However, the above two technical solutions do not take into account the problem of low monitoring data transmission efficiency, which leads to reduced monitoring efficiency. Therefore, the present invention provides a method, system and medium for analyzing fixed pollution source monitoring data based on the Internet of Things. Summary of the Invention

[0006] The present application provides a method, system and medium for analyzing fixed pollution source monitoring data based on the Internet of Things, which are used to improve the efficiency and reliability of monitoring data transmission and further improve the efficiency of pollution source monitoring.

[0007] In a first aspect, the present application provides a method for analyzing fixed pollution source monitoring data based on the Internet of Things, the method comprising:

[0008] Step S1: deploying IoT devices at pollution source monitoring points to collect monitoring data in real time, and sending the monitoring data to a data processing module. The data processing module cleans, denoises, and normalizes the collected monitoring data to obtain processed monitoring data.

[0009] Step S2: extracting first monitoring data from the historical monitoring data, training a first model based on the first monitoring data, extracting multiple different feature quantities from the first monitoring data, training multiple different second models based on the different feature quantities, extracting different types of second monitoring data from the historical monitoring data, and training multiple third models based on the different types of second monitoring data;

[0010] Step S3: acquiring real-time monitoring data, inputting the real-time monitoring data into a first model to obtain a first result, comparing the first result with a preset first threshold, and if the first result is less than the first threshold, inputting the real-time monitoring data into multiple second models to obtain second results, obtaining the second model corresponding to the largest second result, and obtaining the corresponding anomaly type based on the second model;

[0011] Step S4: input the real-time monitoring data into the third model to obtain multiple third results, obtain the abnormal cause corresponding to the third model corresponding to the largest third result, and generate early warning information based on the abnormal type and the abnormal cause.

[0012] In combination with the first aspect, in a first implementation of the first aspect of the present application, sending the monitoring data to the data processing module includes:

[0013] A corresponding data collection terminal is set up for each monitoring point, and the IoT devices at each monitoring point send the collected monitoring data to the corresponding data collection terminal;

[0014] Each data collection terminal sends broadcast information at a preset time interval. The broadcast information includes terminal location information, terminal logo and terminal memory size.

[0015] After collecting broadcast information from other data collection terminals, each data collection terminal updates its own adjacent terminal information table and stores relevant information of adjacent terminals.

[0016] In combination with the first aspect, in a second implementation of the first aspect of the present application, sending the monitoring data to the data processing module includes:

[0017] After receiving the monitoring data sent by the IoT device, the data collection terminal groups all the data collection terminals based on the location information, and selects a target terminal as a transfer terminal for each group of data collection terminals based on the first value;

[0018] When the monitoring data is first transmitted, an initial first value is set for each data collection terminal in each group based on the first rule, a target terminal corresponding to the largest first value is selected for forwarding the monitoring data, transmission feedback is received after forwarding, the transmission feedback including positive feedback and negative feedback, and the first value of the target terminal is updated based on the transmission feedback;

[0019] Each time the monitoring data is subsequently sent, based on the updated first value, the target terminal corresponding to the largest first value is selected to forward the monitoring data, and after forwarding, transmission feedback is received, and the first value is updated based on the transmission feedback.

[0020] In combination with the first aspect, in a third implementation of the first aspect of the present application, updating the first value of the target terminal based on the transmission feedback includes:

[0021] Use the first formula to update the first value. The first formula is: , where V t+1 It refers to the new first value obtained after selecting a target terminal at time t+1, Vt refers to the first value after selecting the target terminal at time t, (1-ɑ) is the proportion of retaining the old first value, ɑ is the proportion of the feedback value, F t+1 is the feedback function for selecting the target terminal at time t+1, γ is the discount factor for future feedback, and R is the future feedback function after the target terminal is selected at time t+1.

[0022] In combination with the first aspect, in a fourth implementation of the first aspect of the present application, the instant feedback function is expressed by a second formula, which is: , where F is the immediate feedback function, L is the size of the remaining cache space of the target terminal, S is the size of the cache space of the target terminal, D1 is the average distance between the target terminal and other data collection terminals in the same group, D2 refers to the result value of the second distance minus the first distance, the first distance refers to the distance between the target terminal and the data processing module, the second distance refers to the average distance between other data collection terminals and the data processing module in the same group, a1 is the weight of the remaining space, a2 is the weight of the distance, T s Refers to monitoring the success of data transmission, M is a negative number, indicating the feedback value after transmission failure, T F Refers to monitoring data transmission failure or transmission errors.

[0023] In combination with the first aspect, in a fifth implementation of the first aspect of the present application, before sending the monitoring data to the data processing module, the method includes:

[0024] The data collection terminal divides the monitoring data into multiple data groups according to a predetermined size, adds corresponding flag bits and detection bits after each data group, generates corresponding detection information, and sends the data groups and detection information to the data processing module.

[0025] In combination with the first aspect, in a sixth implementation of the first aspect of the present application, before the data processing module processes the monitoring data, the method further includes:

[0026] After receiving the data group and detection information, the data processing module obtains the corresponding detection bit from the data group, and generates corresponding detection information based on the detection bit, and compares the generated detection information with the received detection information. If the data of the detection bit and the corresponding bit of the detection information are the same, it means that there is no error in the corresponding data group during the sending process. Otherwise, it means that there is an error in the corresponding data group during the sending process, and the data processing module re-requests the erroneous data group from the data collection terminal.

[0027] In combination with the first aspect, in a seventh implementation of the first aspect of the present application, sending the monitoring data to the data processing module includes:

[0028] In the process of the data collection terminal sending data to the data processing module, the erroneous data bits of the erroneous data group are also counted. When it is detected that the number of errors in a certain data bit exceeds the preset number, fixed data is preset. When the data group is generated, the data of the corresponding erroneous data bit is set as fixed data.

[0029] In a second aspect, the present application provides a fixed pollution source monitoring data analysis system based on the Internet of Things, the system comprising:

[0030] The data collection module is used to deploy IoT devices at pollution source monitoring points to collect monitoring data in real time and send the monitoring data to the data processing module;

[0031] The data processing module is used to clean, denoise and normalize the collected monitoring data to obtain the processed monitoring data;

[0032] a model generation module, configured to extract first monitoring data from historical monitoring data, train a first model based on the first monitoring data, extract a plurality of different feature quantities from the first monitoring data, train a plurality of different second models based on the different feature quantities, extract different types of second monitoring data from the historical monitoring data, and train a plurality of third models based on the different types of second monitoring data;

[0033] A first monitoring module is configured to obtain real-time monitoring data, input the real-time monitoring data into a first model to obtain a first result, compare the first result with a preset first threshold, and if the first result is less than the first threshold, input the real-time monitoring data into multiple second models to obtain a second result, obtain the second model corresponding to the largest second result, and obtain a corresponding anomaly type based on the second model;

[0034] The second monitoring module is used to input real-time monitoring data into the third model to obtain multiple third results, obtain the abnormal cause corresponding to the third model corresponding to the largest third result, and generate early warning information based on the abnormal type and abnormal cause.

[0035] A third aspect of the present application provides a computer-readable storage medium, which stores instructions. When the computer-readable storage medium is run on a computer, it enables the computer to execute the above-mentioned method for analyzing fixed pollution source monitoring data based on the Internet of Things.

[0036] Compared with the prior art, the beneficial effects of the present invention are at least as follows:

[0037] In the technical solution provided by the present application, by deploying Internet of Things devices at pollution source monitoring points, real-time and continuous monitoring of pollutants is achieved; by training multiple models, including a first model for quickly determining whether there is an anomaly, a second model for determining the type of anomaly, and a third model for identifying the specific cause of the anomaly. This multi-model collaborative approach can more accurately identify anomalies and provide strong support for taking targeted governance measures; during the data transmission process, by setting up data collection terminals, adopting broadcast information exchange and dynamic selection of target terminals based on feedback, the data transmission path can be dynamically adjusted according to the network environment and transmission quality, improving the efficiency and reliability of data transmission and reducing the number of data retransmissions; after collecting monitoring data, the data collection terminal selects other data collection terminals as target terminals to continue sending data to the data processing module, dispersing the pressure of data processing, avoiding overload of the data processing module, and improving the overall performance of the system; during the data transmission process, by dividing the monitoring data into multiple data groups, adding detection bits and generating detection information, it can effectively detect whether the data is erroneous during transmission, and promptly request retransmission of the erroneous data group to ensure that the data processing module receives accurate monitoring data. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0039] Figure 1 This is a schematic diagram of an embodiment of a method for analyzing fixed pollution source monitoring data based on the Internet of Things in an embodiment of the present application;

[0040] Figure 2 This is a flowchart of establishing an adjacent terminal information table in an embodiment of the present application;

[0041] Figure 3 This is a flow chart of the data collection terminal sending monitoring data in an embodiment of the present application;

[0042] Figure 4 This is a schematic diagram of an embodiment of a fixed pollution source monitoring data analysis system based on the Internet of Things in an embodiment of the present application. DETAILED DESCRIPTION

[0043] The embodiments of the present application provide a method, system and medium for analyzing data of stationary pollution source monitoring based on the Internet of Things. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0044] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiments of the present application, an embodiment of a method for analyzing fixed pollution source monitoring data based on the Internet of Things includes:

[0045] Step S1: deploy IoT devices at pollution source monitoring points to collect monitoring data in real time, and send the monitoring data to a data processing module. The data processing module cleans, denoises, and normalizes the collected monitoring data to obtain processed monitoring data.

[0046] Specifically, in order to achieve real-time and continuous monitoring of pollution source emissions and improve monitoring efficiency and data accuracy, it is necessary to obtain more comprehensive pollution source emission data to provide comprehensive data support for subsequent anomaly detection and identification. Internet of Things devices are deployed at pollution source monitoring points, that is, pollution source emission outlets. Internet of Things devices include gas sensors, particulate matter sensors, flow sensors, etc., which are used to collect multi-dimensional monitoring data to achieve real-time and continuous collection of pollution source emission data. The collected monitoring data is sent to the data processing module. In the process of sending monitoring data, in order to improve the transmission efficiency of monitoring data, multiple data collection terminals are set for each monitoring point. The data collection terminals send the monitoring data collected by the Internet of Things devices to the data processing module. In order to further improve the transmission efficiency of monitoring data, when transmitting monitoring data, high-performance data collection terminals are selected to forward monitoring data collected by other data collection terminals. Furthermore, in order to improve the reliability of monitoring data transmission, corresponding verification information is generated for the monitoring data to ensure that the data processing module receives accurate monitoring data. The specific monitoring data transmission process will be explained in detail later. After receiving the monitoring data, the data processing module cleans, denoises and normalizes the monitoring data to obtain the processed monitoring data.

[0047] Step S2: extract first monitoring data from historical monitoring data, train a first model based on the first monitoring data, extract multiple different feature quantities from the first monitoring data, train multiple different second models based on different feature quantities, extract different types of second monitoring data from historical monitoring data, and train multiple third models based on different types of second monitoring data.

[0048] Specifically, the first monitoring data refers to the monitoring data under normal conditions (such as monitoring data when the equipment is operating normally and emissions meet standards). The first model can accurately determine whether the input monitoring data deviates from the normal range by learning the data characteristics under normal emission conditions, and quickly identify whether there are abnormal emissions based on the first model.

[0049] Since the first model is used to identify all types of anomalies, its accuracy is limited. The first monitoring data reflects the data distribution pattern of the pollution source under normal conditions. The occurrence of abnormal conditions usually causes the data distribution to deviate from the normal state. By deeply analyzing the distribution pattern of the first monitoring data, it is possible to identify which feature quantities are most sensitive to distinguishing between normal and abnormal conditions, thereby extracting feature quantities suitable for detecting abnormal conditions. Therefore, feature quantities of various different abnormal conditions are extracted from the first monitoring data. Feature quantities include the mean value, fluctuation range, and periodic changes of pollution concentration. Different types of abnormal conditions will produce different disturbance patterns on the first monitoring data, which is the monitoring data under normal conditions. For example, abnormally high emission concentration, abnormal fluctuation of emission flow, abnormal proportion of pollutant components, and periodic anomalies. These different disturbance patterns will introduce specific feature changes based on the first monitoring data. Different feature quantities are most sensitive to specific types of abnormal conditions. Therefore, multiple second models are trained based on different feature quantities. Each second model focuses on detecting a specific type of abnormal situation. Since the second models are specially trained for different types of abnormal conditions, each second model focuses on detecting a specific type of anomaly. Therefore, the second models can more deeply learn the features of the corresponding type of anomaly, thereby improving detection accuracy.

[0050] One anomaly type may correspond to multiple causes. For example, the second model identifies an abnormal increase in emission concentration, which may be caused by equipment failure (sensor failure, valve damage, etc.), operational errors (operator error, incorrect process parameter settings, etc.), external environmental changes (weather changes, ambient interference, etc.), and changes in pollutant sources (new pollution sources, changes in pollutant composition, etc.). The second model focuses primarily on the anomaly type itself and cannot deeply analyze the specific causes of the anomaly. The second monitoring data refers to different monitoring data corresponding to different anomaly causes. Multiple third models are trained based on the second monitoring data. By modeling different anomaly causes separately, the specific causes of the anomaly can be more accurately identified. Once the specific causes of the anomaly are understood, more targeted remediation measures can be taken. Each third model corresponds to a specific anomaly type and is used to identify the specific cause of the anomaly.

[0051] Step S3: obtain real-time monitoring data, input the real-time monitoring data into the first model to obtain a first result, compare the first result with a preset first threshold, if the first result is less than the first threshold, input the real-time monitoring data into multiple second models to obtain a second result, obtain the second model corresponding to the largest second result, and obtain the corresponding abnormality type based on the second model.

[0052] Specifically, the real-time monitoring data is analyzed using the first model and multiple second models. The first model can make an overall evaluation of the real-time monitoring data, quickly identify real-time monitoring data that is suspected to be abnormal, and improve the efficiency of anomaly detection. The first result output by the first model is usually a probability value, which is used to indicate the probability that the current input real-time monitoring data is normal data. The larger the first result, the closer the real-time monitoring data is to the first monitoring data, and the smaller the probability of an abnormal situation, which may be a normal state. On the contrary, the real-time monitoring data and the first monitoring data are significantly different, and the probability of an abnormal situation is greater, which may indicate an anomaly. If the first result is less than the preset first threshold, it means that there is a high possibility of an abnormality at present. Therefore, the real-time monitoring data is input into the second model, and the second model can detect different types of abnormalities. Special detection is performed on the normal state to provide a preliminary judgment on the abnormal type. The real-time monitoring data is input into multiple second models to obtain corresponding multiple second results. Each second model corresponds to an abnormal type. The second model outputs a second result which is a probability value, which is used to indicate the probability that the current input real-time monitoring data corresponds to the abnormal type. The second model corresponding to the largest second result is obtained. Assuming that there are four second models in total, the corresponding abnormal types are: abnormal increase in emission concentration, abnormal fluctuation in emission flow, abnormal proportion of pollutant composition and periodic abnormality. The output second results are distributed as 0.7, 0.2, 0.1 and 0.05. According to the second result, it can be seen that the possibility of the abnormal situation of abnormal increase in emission concentration is the highest at present. Therefore, it is judged that the abnormality of increased emission concentration has occurred.

[0053] Step S4: input the real-time monitoring data into the third model to obtain multiple third results, obtain the abnormal cause corresponding to the third model corresponding to the largest third result, and generate early warning information based on the abnormal type and the abnormal cause.

[0054] Specifically, after determining that a certain type of abnormality has occurred, the real-time monitoring data is input into multiple third models to obtain multiple third results. The third models can identify different types of abnormal causes. The third result output by the third model is a probability value, which is used to indicate the probability that the currently input monitoring data is the abnormal cause corresponding to the third model. The abnormal cause corresponding to the largest third result is used as the abnormal cause of the corresponding abnormal type. Based on the abnormal type and abnormal cause, early warning information is generated and sent to relevant staff to help relevant staff formulate treatment measures.

[0055] Through the coordination of the above methods, this application first realizes real-time and continuous monitoring of pollution sources by deploying Internet of Things devices, thereby improving monitoring efficiency and data real-timeness; cleans, denoises and normalizes the collected data through the data processing module, thereby enhancing the accuracy and comprehensiveness of the data; accurately identifies abnormal situations through multi-model collaborative analysis (the first model judges the abnormality, the second model determines the type of abnormality, and the third model identifies the cause of the abnormality), providing support for targeted governance; during data transmission, the data collection terminal dynamically selects the target terminal, optimizes the transmission path, improves efficiency and reliability, and reduces the number of retransmissions; the data grouping and detection information mechanism ensures transmission accuracy; overall, this method has strong adaptability and high reliability, provides strong guarantees for the long-term and stable monitoring of fixed pollution sources, and can effectively improve the level of environmental monitoring and governance.

[0056] The method of the present application can also adapt to different pollution source monitoring scenarios and network environments. By dynamically adjusting model parameters and data transmission strategies, the adaptability and reliability of the system are improved, providing a guarantee for long-term and stable monitoring of fixed pollution sources.

[0057] In a specific embodiment, sending the monitoring data to the data processing module specifically includes the following steps:

[0058] A corresponding data collection terminal is set up for each monitoring point, and the IoT devices at each monitoring point send the collected monitoring data to the corresponding data collection terminal.

[0059] Each data collection terminal sends broadcast information at a preset time interval. The broadcast information includes terminal location information, terminal logo and terminal memory size.

[0060] After collecting broadcast information from other data collection terminals, each data collection terminal updates its own adjacent terminal information table and stores relevant information of adjacent terminals.

[0061] Specifically, in order to improve the transmission efficiency of monitoring data, a data collection terminal is first set up for each detection point to collect the monitoring data collected by the IoT devices at the monitoring point, such as Figure 2As shown in the figure, it is a flowchart for establishing an adjacent terminal information table. In order to establish and maintain the adjacent terminal information table and obtain information about the surrounding adjacent collection terminals, each data collection terminal sends broadcast information. The broadcast information includes terminal location information, terminal flag and terminal memory size. The distance between the two data collection terminals is calculated based on the location information of the two terminals. The terminal flag information refers to the ID of the data collection terminal. The transit terminal is used to forward the monitoring data of other data collection terminals. Each data collection terminal has a buffer memory for temporarily storing monitoring data waiting to be sent. When the data collection terminal receives the monitoring data, it stores the monitoring data in the buffer memory and queues to wait for the corresponding monitoring data to be sent. After collecting the broadcast information from other data collection terminals, each data collection terminal updates its own adjacent terminal information table. The adjacent terminal information table is used to store relevant information about adjacent terminals. The relevant information includes adjacent terminal location information, adjacent terminal flag, the distance between the adjacent terminal and itself and the remaining space size of the adjacent terminal buffer memory. Through the exchange of broadcast information, each data collection terminal can obtain the information of adjacent data collection terminals, which provides data support for the subsequent search for the target terminal for forwarding monitoring data.

[0062] In a specific embodiment, sending the monitoring data to the data processing module further includes the following steps:

[0063] After receiving the monitoring data sent by the Internet of Things device, the data collection terminal groups all the data collection terminals based on the location information, and selects a target terminal as a transfer terminal for each group of data collection terminals based on the first value.

[0064] When the monitoring data is sent for the first time, an initial first value is set for the data collection terminal in each group based on the first rule, and the target terminal corresponding to the largest first value is selected to forward the monitoring data. After forwarding, transmission feedback is received, and the transmission feedback includes positive feedback and negative feedback. The first value of the target terminal is updated based on the transmission feedback.

[0065] Each time the monitoring data is subsequently sent, based on the updated first value, the target terminal corresponding to the largest first value is selected to forward the monitoring data, and after forwarding, transmission feedback is received, and the first value is updated based on the transmission feedback.

[0066] Specifically, in order to reduce the pressure on the data processing module, after collecting the monitoring data, the data collection terminal selects a target terminal as a transit terminal for each group of data collection terminals based on the first numerical value, and the target terminal continues to send the monitoring data to the data processing module, which can reduce the communication pressure of the data processing module and thus improve the transmission efficiency and reliability of the monitoring data.

[0067] At the beginning, since the data collection terminal does not understand the network environment and cannot determine which data collection terminal is the best target terminal, it is initially set to a first value for the adjacent terminal according to the first rule. The first rule is to set a first value for the adjacent terminal based on the length of the data packet to be sent. The smaller the length of the data packet to be sent, the larger the first value. The target terminal corresponding to the largest first value is selected to monitor data forwarding, and receive instant transmission feedback after forwarding. The transmission feedback includes positive feedback and negative feedback. Positive feedback includes successful transmission of monitoring data to the target terminal, low transmission time and successful transmission of monitoring data to the data processing module. Negative feedback includes failure of monitoring data transmission, error of monitoring data transmission and high transmission time. The specific positive feedback and negative feedback will be detailed later according to the feedback function. Description: Based on the transmission feedback, the first value corresponding to the selected target terminal is updated. If the first value corresponding to the positive feedback is increased, it means that the target terminal is a good choice, and it may be selected again when the monitoring data is transmitted next time. If negative feedback is obtained, it means that the monitoring data transmission fails or the transmission time is too long, and the first value will decrease, indicating that the target terminal is not a good choice. Other options will be tried when the monitoring data is transmitted next time. In addition to immediate feedback, possible future feedback will also be considered when updating the first value. For example, after selecting a target terminal, if the subsequent transmission process goes smoothly, then the first value of this target terminal will be higher. The update of the first value will comprehensively consider immediate feedback and future feedback. The specific update formula will be explained in detail later, such as Figure 3 As shown in FIG, it is a flow chart of the data collection terminal sending monitoring data.

[0068] Each data collection terminal will select a target terminal based on the current first value, send monitoring data, receive feedback, and then update the first value. The data collection terminal will continuously explore new target terminals and also use previous transmission experience. As the monitoring data is continuously forwarded, the data collection terminal accumulates enough experience and selects the best target terminal based on the first value each time it forwards the monitoring data, ultimately achieving efficient and reliable data transmission.

[0069] In a specific embodiment, updating the first value of the target terminal based on the transmission feedback includes the following steps:

[0070] Use the first formula to update the first value. The first formula is: , where V t+1 It refers to the new first value obtained after selecting a target terminal at time t+1, Vt refers to the first value after selecting the target terminal at time t, (1-ɑ) is the proportion of retaining the old first value, ɑ is the proportion of the feedback value, F t+1is the immediate feedback function of selecting the target terminal at time t+1, γ is the discount factor of future feedback, and R is the future feedback function after the target terminal is selected at time t+1.

[0071] Specifically, the first value is calculated using the above first formula, where V t+1 It refers to the new first value obtained after selecting a target terminal at time t+1. Vt refers to the first value after selecting the same target terminal at time t. (1-ɑ) is the proportion of the old first value retained, ranging from 0 to 1. ɑ is the proportion of the feedback value. The feedback value is calculated based on the feedback function and represents the feedback value of the new experience. The larger ɑ is, the more emphasis is placed on the new experience. Conversely, the more emphasis is placed on the past experience. F t+1 is the immediate feedback function for selecting the target terminal at time t+1, γ is the discount factor for future feedback, and its value range is between 0 and 1. A higher γ indicates more emphasis on future feedback, and a lower γ indicates more attention on immediate feedback. R is the future feedback function after selecting the target terminal at time t+1.

[0072] In a specific embodiment, the instant feedback function is expressed by a second formula, which is: , where F is the immediate feedback function, L is the size of the remaining cache space of the target terminal, S is the size of the cache space of the target terminal, D1 is the average distance between the target terminal and other data collection terminals in the same group, D2 refers to the result value of the second distance minus the first distance, the first distance refers to the distance between the target terminal and the data processing module, the second distance refers to the average distance between other data collection terminals and the data processing module in the same group, a1 is the weight of the remaining space, a2 is the weight of the distance, T s Refers to monitoring the success of data transmission, M is a negative number, indicating the feedback value after transmission failure, T F Refers to monitoring data transmission failure or transmission errors.

[0073] Specifically, the instant feedback function includes positive feedback and negative feedback. If the monitoring data transmission is successful, the corresponding reward value is calculated as positive feedback. If the monitoring data transmission fails or an error occurs in the transmission, the value of the feedback function is set to a negative number M as negative feedback. Through the above feedback function, the data collection terminal is encouraged to select a data collection terminal with large remaining cache space, close distance to the data processing module, and good transmission quality as the target terminal to forward the monitoring data, making the transmission of the monitoring data more efficient and reliable.

[0074] Furthermore, the future feedback function is expressed using the third formula, which is: R=N*(1-1 / G), where N is the future transmission success ratio of the target terminal predicted based on historical communication data, which can reflect the reliability of the target terminal's future transmission; G is the future transmission duration of the target terminal predicted based on historical communication data, which can reflect the efficiency of the target terminal's future transmission.

[0075] In a specific embodiment, before the monitoring data is sent to the data processing module, the following steps are specifically included:

[0076] The data collection terminal divides the monitoring data into multiple data groups according to a predetermined size, adds corresponding flag bits and detection bits after each data group, generates corresponding detection information, and sends the data groups and detection information to the data processing module.

[0077] Specifically, in order to ensure the accuracy of the monitoring data during the transmission process, before sending the monitoring data to the data processing module, the data collection terminal divides the monitoring data into multiple data groups according to a predetermined size, and adds a corresponding detection bit after each data group. For example, in the process of sending detection data a certain time, 8 data groups need to be sent, and the 8 data groups are B1-B8 respectively. An 8-bit binary number is randomly generated, assuming it is 10010011 as the corresponding detection information, and each binary number is added to the end of the 8 data groups as the corresponding detection bit. The generated 8-bit binary data is used as the detection information to determine whether an error occurs during the transmission process, and the data group and the detection information are sent to the data processing module to facilitate the data processing module to verify the data group based on the detection information.

[0078] In a specific embodiment, before the data processing module processes the monitoring data, the following steps are specifically included:

[0079] After receiving the data group and detection information, the data processing module obtains the corresponding detection bit from the data group, and generates corresponding detection information based on the detection bit, and compares the generated detection information with the received detection information. If the data of the detection bit and the corresponding bit of the detection information are the same, it means that there is no error in the corresponding data group during the sending process. Otherwise, it means that there is an error in the corresponding data group during the sending process, and the data processing module re-requests the erroneous data group from the data collection terminal.

[0080] Specifically, assuming that the detection information received by the data processing module is 10010011, and the detection information generated based on the detection bit of the data group is 10011011, by comparing the received detection information and the generated detection information, it is found that the 5th bit data is different, indicating that the corresponding data group has an error during the transmission process. The data processing module re-requests the erroneous data group from the data collection terminal to ensure that the data collection terminal receives the correct data group.

[0081] In a specific embodiment, sending the monitoring data to the data processing module further includes the following steps:

[0082] In the process of the data collection terminal sending data to the data processing module, the erroneous data bits of the erroneous data group are also counted. When it is detected that the number of errors in a certain data bit exceeds the preset number, fixed data is preset. When the data group is generated, the data of the corresponding erroneous data bit is set as fixed data.

[0083] Specifically, in order to ensure that the data processing module receives the correct data group, the data collection terminal counts the data groups that often have errors. For example, when sending the data group, it is found that the data in the 5th position often has errors, and the number of errors exceeds the preset number. For example, a total of 10 data are sent, and 5 or more errors occur. Then a fixed data is preset, assuming it is 1. When generating the data group, the corresponding 5th bit of each data group is set to fixed data, for example, it can be set to a fixed 0 or 1, and then the data group is sent to the data processing module. After receiving the data group, the data processing module deletes the data at the corresponding position. Assuming that the 5th bit is fixed data, the 5th bit data is deleted, and then subsequent operations are performed based on the deleted data. The above method can reduce the number of data retransmissions and improve the efficiency of data transmission.

[0084] The above describes the method for analyzing fixed pollution source monitoring data based on the Internet of Things in the embodiment of the present application. The following describes the system for analyzing fixed pollution source monitoring data based on the Internet of Things in the embodiment of the present application. Figure 4 In the embodiment of the present application, an embodiment of the fixed pollution source monitoring data analysis system based on the Internet of Things includes:

[0085] The data collection module is used to deploy IoT devices at pollution source monitoring points to collect monitoring data in real time and send the monitoring data to the data processing module.

[0086] The data processing module is used to clean, denoise and normalize the collected monitoring data to obtain the processed monitoring data.

[0087] A model generation module is used to extract first monitoring data from historical monitoring data, train a first model based on the first monitoring data, extract multiple different feature quantities from the first monitoring data, train multiple different second models based on different feature quantities, extract different types of second monitoring data from historical monitoring data, and train multiple third models based on different types of second monitoring data.

[0088] The first monitoring module is used to obtain real-time monitoring data, input the real-time monitoring data into the first model to obtain a first result, compare the first result with a preset first threshold, and if the first result is less than the first threshold, input the real-time monitoring data into multiple second models to obtain a second result, obtain the second model corresponding to the largest second result, and obtain the corresponding abnormality type based on the second model.

[0089] The second monitoring module is used to input real-time monitoring data into the third model to obtain multiple third results, obtain the abnormal cause corresponding to the third model corresponding to the largest third result, and generate early warning information based on the abnormal type and abnormal cause.

[0090] The present application also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the steps of a method for analyzing fixed pollution source monitoring data based on the Internet of Things.

[0091] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0092] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program code.

[0093] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for analyzing fixed pollution source monitoring data based on the Internet of Things, characterized in that: The method comprises: Step S1: deploying Internet of Things devices at pollution source monitoring points to collect monitoring data in real time and sending the monitoring data to a data processing module, including: after the data collection terminal receives the monitoring data sent by the Internet of Things device, grouping all data collection terminals based on location information, and selecting a target terminal as a transfer terminal for each group of data collection terminals based on a first value; when sending monitoring data for the first time, setting an initial first value for the data collection terminals in each group based on a first rule, selecting a target terminal corresponding to the largest first value for forwarding the monitoring data, receiving transmission feedback after forwarding, the transmission feedback including positive feedback and negative feedback, and updating the first value of the target terminal based on the transmission feedback, including: updating the first value using a first formula, the first formula being: , where V t+1 It refers to the new first value obtained after selecting a target terminal at time t+1, V t It refers to the first value after the target terminal is selected at time t, (1-ɑ) is the proportion of the old first value retained, ɑ is the proportion of the feedback value, F t+1 is the immediate feedback function for selecting the target terminal at time t+1, γ is the discount factor for future feedback, and R is the future feedback function after the target terminal is selected at time t+1; each time the monitoring data is subsequently sent, based on the updated first value, the target terminal corresponding to the largest first value is selected for forwarding the monitoring data, and after forwarding, transmission feedback is received. The first value is updated based on the transmission feedback, and the data processing module cleans, denoises, and normalizes the collected monitoring data to obtain processed monitoring data; Step S2: extracting first monitoring data from the historical monitoring data, training a first model based on the first monitoring data, extracting multiple different feature quantities from the first monitoring data, training multiple different second models based on the different feature quantities, extracting different types of second monitoring data from the historical monitoring data, and training multiple third models based on the different types of second monitoring data; Step S3: acquiring real-time monitoring data, inputting the real-time monitoring data into a first model to obtain a first result, comparing the first result with a preset first threshold, and if the first result is less than the first threshold, inputting the real-time monitoring data into multiple second models to obtain second results, obtaining the second model corresponding to the largest second result, and obtaining the corresponding anomaly type based on the second model; Step S4: input the real-time monitoring data into the third model to obtain multiple third results, obtain the abnormal cause corresponding to the third model corresponding to the largest third result, and generate early warning information based on the abnormal type and the abnormal cause.

2. The method according to claim 1, characterized in that Send monitoring data to the data processing module, including: A corresponding data collection terminal is set up for each monitoring point, and the IoT devices at each monitoring point send the collected monitoring data to the corresponding data collection terminal; Each data collection terminal sends broadcast information at a preset time interval. The broadcast information includes terminal location information, terminal logo and terminal memory size. After collecting broadcast information from other data collection terminals, each data collection terminal updates its own adjacent terminal information table and stores relevant information of adjacent terminals.

3. The method according to claim 1, characterized in that The instant feedback function is expressed by the second formula, which is: , where F is the immediate feedback function, L is the size of the remaining cache space of the target terminal, S is the size of the cache space of the target terminal, D1 is the average distance between the target terminal and other data collection terminals in the same group, D2 refers to the result value of the second distance minus the first distance, the first distance refers to the distance between the target terminal and the data processing module, the second distance refers to the average distance between other data collection terminals and the data processing module in the same group, a1 is the weight of the remaining space, a2 is the weight of the distance, T s Refers to monitoring the success of data transmission, M is a negative number, indicating the feedback value after transmission failure, T F Refers to monitoring data transmission failure or transmission errors.

4. The method according to claim 1, wherein Before the monitoring data is sent to the data processing module, it also includes: The data collection terminal divides the monitoring data into multiple data groups according to a predetermined size, adds corresponding flag bits and detection bits after each data group, generates corresponding detection information, and sends the data groups and detection information to the data processing module.

5. The method according to claim 1, wherein Before the data processing module processes the monitoring data, it also includes: After receiving the data group and detection information, the data processing module obtains the corresponding detection bit from the data group, and generates corresponding detection information based on the detection bit, and compares the generated detection information with the received detection information. If the data of the detection bit and the corresponding bit of the detection information are the same, it means that there is no error in the corresponding data group during the sending process. Otherwise, it means that there is an error in the corresponding data group during the sending process, and the data processing module re-requests the erroneous data group from the data collection terminal.

6. The method according to claim 1, characterized in that Send monitoring data to the data processing module, including: In the process of the data collection terminal sending data to the data processing module, the erroneous data bits of the erroneous data group are also counted. When it is detected that the number of errors in a certain data bit exceeds the preset number, fixed data is preset. When the data group is generated, the data of the corresponding erroneous data bit is set as fixed data.

7. A fixed pollution source monitoring data analysis system based on the Internet of Things, used to implement the fixed pollution source monitoring data analysis method based on the Internet of Things as described in any one of claims 1 to 6, characterized in that: The system comprises: The data collection module is used to deploy IoT devices at pollution source monitoring points to collect monitoring data in real time and send the monitoring data to the data processing module; The data processing module is used to clean, denoise and normalize the collected monitoring data to obtain the processed monitoring data; a model generation module, configured to extract first monitoring data from historical monitoring data, train a first model based on the first monitoring data, extract a plurality of different feature quantities from the first monitoring data, train a plurality of different second models based on the different feature quantities, extract different types of second monitoring data from the historical monitoring data, and train a plurality of third models based on the different types of second monitoring data; A first monitoring module is configured to obtain real-time monitoring data, input the real-time monitoring data into a first model to obtain a first result, compare the first result with a preset first threshold, and if the first result is less than the first threshold, input the real-time monitoring data into multiple second models to obtain a second result, obtain the second model corresponding to the largest second result, and obtain a corresponding anomaly type based on the second model; The second monitoring module is used to input real-time monitoring data into the third model to obtain multiple third results, obtain the abnormal cause corresponding to the third model corresponding to the largest third result, and generate early warning information based on the abnormal type and abnormal cause.

8. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the method for analyzing fixed pollution source monitoring data based on the Internet of Things according to any one of claims 1 to 6 is implemented.

Images