Remote monitoring method and device for smart water meters integrated with NB-IoT

Abstract

The present invention discloses a remote monitoring method and device for smart water meters integrated with NB-IoT, which relates to the technical field related to remote monitoring. The method comprises: integrating a first NB-IoT communication module and a second NB-IoT communication module, wherein the first module corresponds to a first transmission channel and the second module corresponds to a second transmission channel; acquiring a distributed water meter network, extracting and training historical monitoring data, and acquiring a remote transmission channel discriminator; performing real-time monitoring of the distributed water meter network and outputting a real-time data set; discriminating the real-time data set according to the discriminator to obtain first and second types of monitoring data; transmitting the first type of data to a first edge device based on the first transmission channel, and transmitting the second type of data to a second edge device based on the second transmission channel, so as to solve the technical problem that data of different urgency levels cannot be quickly distinguished and transmitted in a complex network environment, thereby affecting the real-time and accuracy of the data.

Description

Technical Field

[0001] The present invention relates to the field of remote monitoring technology, and in particular to a remote monitoring method and device for smart water meters integrated with NB-IoT. Background Art

[0002] With the rapid development of smart cities and the Internet of Things (IoT), traditional water meters are increasingly facing challenges such as low data collection efficiency, difficulty in remote monitoring, and poor data accuracy. Traditional water meters typically rely on manual meter reading, which incurs frequent intervention and management costs, and cannot monitor abnormal conditions such as water flow and leaks in real time. These issues lead to wasted water resources, inconvenience for users, and high costs for operators. With the continuous advancement of intelligent technologies, particularly the application of the Internet of Things (IoT), remote monitoring issues with traditional water meters are being addressed. However, existing smart water meter systems still face technical challenges such as unstable communication networks, data transmission delays, and delayed emergency response. In particular, in distributed water meter networks, the diverse transmission requirements of water meters in different locations and environmental conditions make a single communication module and transmission channel incapable of meeting all monitoring needs. Therefore, improving the data transmission efficiency and real-time responsiveness of smart water meter systems in complex environments remains a major challenge. Summary of the Invention

[0003] This application provides a remote monitoring method and device for smart water meters integrated with NB-IoT, which is used to solve the technical problem that data of different urgency levels cannot be quickly distinguished and transmitted in a complex network environment, thereby affecting the real-time and accuracy of the data.

[0004] In view of the above problems, the present application provides a remote monitoring method and device for smart water meters integrated with NB-IoT.

[0005] The first aspect of the present application provides a remote monitoring method for smart water meters integrated with NB-IoT, the method comprising: integrating a first NB-IoT communication module and a second NB-IoT communication module, wherein the first NB-IoT communication module corresponds to a first remote transmission channel, and the second NB-IoT communication module corresponds to a second remote transmission channel; obtaining a distributed water meter network, extracting a historical monitoring data set of the distributed water meter network, training the historical monitoring data set, and obtaining a remote transmission channel discriminator; performing real-time monitoring of the distributed water meter network and outputting a real-time monitoring data set; discriminating the real-time monitoring data set according to the remote transmission channel discriminator to obtain first-category monitoring data and second-category monitoring data, wherein the first-category monitoring data is data greater than or equal to a preset urgency level, and the second-category monitoring data is data less than the preset urgency level; transmitting the first-category monitoring data to a first edge device based on the first remote transmission channel, and transmitting the second-category monitoring data to a second edge device based on the second remote transmission channel.

[0006] The second aspect of the present application provides an intelligent water meter remote monitoring device integrated with NB-IoT, the device comprising: a communication module integration component: integrating a first NB-IoT communication module and a second NB-IoT communication module, wherein the first NB-IoT communication module corresponds to a first remote transmission channel, and the second NB-IoT communication module corresponds to a second remote transmission channel; a discriminator training component: acquiring a distributed water meter network, extracting a historical monitoring data set of the distributed water meter network, training the historical monitoring data set, and obtaining a remote transmission channel discriminator; a real-time monitoring component: performing real-time monitoring of the distributed water meter network and outputting a real-time monitoring data set; a data judgment component: judging the real-time monitoring data set according to the remote transmission channel discriminator, and obtaining first-category monitoring data and second-category monitoring data, wherein the first-category monitoring data is data greater than or equal to a preset urgency level, and the second-category monitoring data is data less than the preset urgency level; a data transmission component: transmitting the first-category monitoring data to a first edge device based on the first remote transmission channel, and transmitting the second-category monitoring data to a second edge device based on the second remote transmission channel.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] The system integrates a first NB-IoT communication module and a second NB-IoT communication module, wherein the first NB-IoT communication module corresponds to a first remote transmission channel and the second NB-IoT communication module corresponds to a second remote transmission channel; obtains a distributed water meter network, extracts a historical monitoring data set of the distributed water meter network, trains the historical monitoring data set, and obtains a remote transmission channel discriminator; performs real-time monitoring of the distributed water meter network and outputs a real-time monitoring data set; discriminates the real-time monitoring data set according to the remote transmission channel discriminator to obtain first-category monitoring data and second-category monitoring data, wherein the first-category monitoring data is data greater than or equal to a preset urgency level and the second-category monitoring data is data less than the preset urgency level; transmits the first-category monitoring data to a first edge device based on the first remote transmission channel, and transmits the second-category monitoring data to a second edge device based on the second remote transmission channel. This system achieves the technical effect of improving the real-time transmission efficiency and reliability of monitoring data and ensuring the accuracy and timeliness of remote monitoring by integrating multi-channel NB-IoT communication modules. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in 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 only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0010] Figure 1 Schematic diagram of the flow chart of the remote monitoring method of smart water meters integrated with NB-IoT provided in the embodiment of the present application.

[0011] Figure 2 Schematic diagram of the structure of the smart water meter remote monitoring device integrated with NB-IoT provided in the embodiment of the present application.

[0012] Explanation of the accompanying drawings: communication module integration component 1, discriminator training component 2, real-time monitoring component 3, data judgment component 4, data transmission component 5. DETAILED DESCRIPTION

[0013] This application provides a remote monitoring method and device for smart water meters integrated with NB-IoT, which is used to solve the technical problem that data of different urgency levels cannot be quickly distinguished and transmitted in a complex network environment, thereby affecting the real-time and accuracy of the data.

[0014] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making any creative work are within the scope of protection of this application.

[0015] Example 1, as Figure 1 As shown, the present application provides a remote monitoring method for a smart water meter integrated with NB-IoT, the method comprising:

[0016] Integrate a first NB-IoT communication module and a second NB-IoT communication module, wherein the first NB-IoT communication module corresponds to a first remote transmission channel, and the second NB-IoT communication module corresponds to a second remote transmission channel.

[0017] Specifically, the first NB-IoT communication module and the second NB-IoT communication module are integrated. The first NB-IoT communication module is connected to the first remote transmission channel, and the second NB-IoT communication module is connected to the second remote transmission channel. The two communication modules are used to transmit data through different remote transmission channels, thereby ensuring that the appropriate transmission channel can be selected for data transmission based on different transmission requirements and environmental conditions, improving the stability and efficiency of the data transmission process. This multi-channel design enables adaptation to more complex network environments and various application scenarios, ensuring the reliability and real-time performance of data transmission.

[0018] In one possible implementation, the central processing unit is connected to the first NB-IoT communication module and the second NB-IoT communication module via an SPI communication bus.

[0019] Specifically, the central processing unit (CPU) connects to two NB-IoT communication modules via the SPI (Serial Peripheral Interface) communication bus. The SPI bus is a high-speed, synchronous communication protocol that enables the CPU to exchange data and control the NB-IoT communication modules. Through this connection, the CPU can simultaneously manage the working status of both communication modules and coordinate their data transmission tasks. This approach not only increases data transmission rates but also ensures efficient collaboration between the communication modules, thereby ensuring stable operation and fast response under different long-distance transmission channels.

[0020] A distributed water meter network is obtained, a historical monitoring data set of the distributed water meter network is extracted, and the historical monitoring data set is trained to obtain a remote transmission channel discriminator.

[0021] Specifically, the distributed water meter network is first acquired and the monitoring database of each sensor in this distributed water meter network is accessed to extract the historical monitoring data set of the entire distributed water meter network. This historical data set is then analyzed and trained using a DNN neural network to learn the patterns and characteristics of the data. Through training, a remote transmission channel discriminator is established. This discriminator can determine the optimal remote transmission channel in different situations based on the characteristics of the historical data, thereby ensuring stable and efficient data transmission.

[0022] In one possible implementation, training the historical monitoring dataset to obtain a long-range transmission channel discriminator includes:

[0023] By performing feature extraction on the historical monitoring data set, multiple data features are obtained, and the multiple data features include time features, data types, sensor status, and urgency levels; a DNN neural network is initialized, and each of the multiple data features is used as an input node to perform multi-level training on the DNN neural network to obtain data training results; the data training results are tuned by tuning hyperparameters until the model classification accuracy is greater than the preset classification accuracy, thereby obtaining a remote transmission channel discriminator.

[0024] Specifically, multiple data features are extracted from the historical monitoring data set, including time features, data types, sensor status, and urgency. Among them, the time feature is the timestamp of data collection or the time information of periodic changes, which helps to identify the time pattern of data changes. The data type represents the type of data, such as water flow, pressure, temperature, etc. The sensor status reflects the health status of the sensor or whether it has a fault, which affects the reliability of the data. The urgency is graded according to the severity of the monitoring data to help determine whether an immediate response is required. Subsequently, the historical monitoring data set is divided into a training set and a validation set, and the constructed initial remote transmission channel discriminator is trained using the training set. After the training is completed, the validation set is used for verification to evaluate the performance of the model. The initial long-range transmission channel discriminator is constructed based on the deep neural network (DNN) model, including input layer, output layer, hidden layer, etc. Among them, the input layer is used to receive external data and pass the data to the next layer of the neural network. It includes multiple input nodes, each node represents an input feature, such as timestamp, data type, sensor status, etc. The hidden layer is located between the input layer and the output layer, responsible for processing and converting the input data. Through multiple hidden layers, the neural network can extract higher-level abstract features from the input data. It includes multiple neurons inside to learn the relationship between input features and target outputs. The relationship between them gradually improves the network's ability to fit the data. The output layer is the final layer of the neural network. It generates the final prediction result based on the output of the hidden layer. It contains an output node for calculating the urgency of the current data. During the training process, the weights of the initial remote transmission channel discriminator are initialized using a random initialization method, and the training set is input into the initial remote transmission channel discriminator for forward propagation. It is passed through the input layer, hidden layer, and output layer layer by layer to calculate the predicted urgency of the current data. The mean square error loss function is then used to calculate the loss between the prediction result and the true value, and the gradient of the loss to the weight of each layer is calculated layer by layer through the back propagation algorithm. Afterwards, an optimizer (such as Adam) is used to optimize the parameters of the discriminator to minimize the loss function; through multiple training sessions, the discriminator gradually adjusts its weights until the preset classification accuracy is reached; after the training is completed, a validation set is used for testing to evaluate the accuracy of the discriminator in the discrimination tasks of different transmission channels. If the accuracy exceeds the preset classification accuracy, the initial remote transmission channel discriminator is regarded as the final remote transmission channel discriminator; if the accuracy does not meet the requirements, hyperparameters such as the learning rate and training batch are adjusted to continue optimizing the discriminator performance; finally, the trained discriminator will be used for transmission channel selection of real-time data to optimize the remote transmission efficiency of water meter data.

[0025] The distributed water meter network is monitored in real time, and a real-time monitoring data set is output.

[0026] Specifically, the distributed water meter network is continuously monitored in real time. Through the deployment of various sensors and data acquisition devices, such as pressure sensors and flow sensors, real-time data of each water meter (such as water flow, pressure, temperature, etc.) is continuously obtained. These real-time collected data are organized into a data set as a real-time monitoring data set. These data reflect the current operating status and working environment of the water meter for subsequent analysis, judgment and processing.

[0027] The real-time monitoring data set is discriminated according to the remote transmission channel discriminator to obtain first-category monitoring data and second-category monitoring data, wherein the first-category monitoring data is data greater than or equal to a preset urgency level, and the second-category monitoring data is data less than the preset urgency level.

[0028] Specifically, after obtaining the real-time monitoring data set, the real-time monitoring data set is analyzed using the trained remote transmission channel discriminator. The discriminator judges the urgency of each piece of data based on the characteristics of each piece of real-time monitoring data in the real-time monitoring data set (such as time characteristics, data type, sensor status, etc.), thereby obtaining multiple urgency levels; then, these urgency levels are compared with the preset urgency levels. If the urgency of a piece of data is greater than or equal to the preset threshold, the data is added to the first category of monitoring data. This type of data usually represents a more serious anomaly or requires immediate processing. If the urgency of a piece of data is less than the preset threshold, the data is added to the second category of monitoring data. This type of data usually indicates that the monitoring results are within the normal range or are slightly abnormal and do not require immediate response. Through this discrimination process, the monitoring data can be divided into two categories according to the degree of urgency, which facilitates subsequent priority processing and transmission based on the importance of the data.

[0029] The first type of monitoring data is transmitted to a first edge device based on the first remote transmission channel, and the second type of monitoring data is transmitted to a second edge device based on the second remote transmission channel.

[0030] Specifically, according to the urgency of the monitoring data, a suitable remote transmission channel is selected to transmit the data to the corresponding edge device; for the first type of monitoring data (i.e., data with a higher degree of urgency), these data are transmitted to the first edge device through the first remote transmission channel. This type of data usually requires priority processing, so it is transmitted to an edge device that can quickly respond to and handle these emergencies; for the second type of monitoring data (i.e., data with a lower degree of urgency), these data are transmitted to the second edge device through the second remote transmission channel. This type of data usually does not need to be processed immediately and is transmitted to another edge device for subsequent processing or storage. In this way, the monitoring data can be reasonably allocated and transmitted according to the urgency of the data, ensuring that important data can be responded to and processed quickly, while less urgent data is reasonably delayed or processed, thereby optimizing resource utilization and response efficiency.

[0031] In one possible implementation, transmitting the first type of monitoring data to a first edge device based on the first remote transmission channel includes:

[0032] Download a pre-trained anomaly detection model and embed the pre-trained anomaly detection model in the first edge device; perform anomaly detection on the first type of monitoring data based on the pre-trained anomaly detection model to obtain anomaly detection data; obtain a first warning signal based on the anomaly detection data, and the first edge device processes the distributed water meter network based on the first warning signal.

[0033] Specifically, the system terminal downloads a trained anomaly detection model from the cloud or other storage media. This model is trained on historical monitoring data sets and can effectively identify abnormal patterns or abnormal monitoring data in the water meter network. The model can be based on common machine learning algorithms, such as neural networks, decision trees, random forests, support vector machines, etc., and is obtained through a similar training process as mentioned above; after the download is completed, the pre-trained anomaly detection model will be embedded in the first edge device, which is responsible for receiving data transmitted from the water meter network and processing the data locally in real time; after the model is embedded, the first edge device can perform anomaly detection on real-time data based on the model; after receiving the first type of monitoring data, the first edge device will use the embedded anomaly detection model to analyze the data. The model will analyze the data based on the learned typical anomalies (such as equipment failure, leakage, etc.) and data features (such as water flow) , pressure, temperature, etc.), and identify whether there are typical anomalies in the first category of monitoring data; if the first category of monitoring data is detected to be a typical anomaly, the model will use this typical anomaly to mark the corresponding monitoring data, and add the marked data to the anomaly detection data. These data indicate that there may be problems in the water meter network, such as equipment failure, leakage, abnormal fluctuations, etc.; after identifying the anomaly detection data, the first edge device will generate a first warning signal based on these abnormal data. This warning signal indicates that there is a potential emergency or failure in the water meter network and requires timely response and processing; finally, the first edge device will process the distributed water meter network according to the obtained first warning signal. The processing measures include sending warning information to management personnel, starting emergency response procedures, adjusting the working status of water meters, etc., to ensure that potential emergencies can be responded to quickly and avoid greater losses or failure expansion. Through the above steps, the first edge device can not only detect anomalies in the data in real time, but also take timely processing measures according to the warning signal to ensure the stable operation of the water meter network.

[0034] In one possible implementation, after obtaining the anomaly detection data, the method further includes:

[0035] The abnormality detection data is sent to the second edge device, wherein the first edge device and the second edge device are communicatively connected; a correlation analysis is performed on the received abnormality detection data according to the second edge device to obtain associated detection data of the abnormality detection data; and the abnormality detection data is updated according to the associated detection data to obtain updated abnormality detection data.

[0036] Specifically, after the first edge device detects abnormal data, it sends this abnormal detection data to the second edge device via a communication connection (such as a local area network, wireless network, or other transmission method). This data transmission is usually real-time, ensuring that the second edge device can receive the abnormal data detected by the first edge device in a timely manner for further processing and analysis. After the second edge device receives the abnormal detection data from the first edge device, it performs correlation analysis on the data. The purpose of this analysis is to find potential relationships between the data. For example, check whether multiple water meter devices have abnormalities at the same time, whether water meter devices in a specific area or location frequently have problems, whether they are associated with certain specific events (such as equipment failures, environmental changes, etc.), and whether there are regular or sporadic abnormalities. Through these analyses, the second edge device can identify the correlation between different abnormal data, providing more information and context for subsequent processing. After completing the correlation analysis, the second edge device will use the obtained correlation detection data to update the abnormal detection data, that is, add any abnormal data that may have been missed to the abnormal detection data. In this way, the abnormal detection data will become more accurate and comprehensive, ensuring that various problems in the water meter network can be identified and responded to in a timely and efficient manner.

[0037] In one possible implementation, performing correlation analysis on the received anomaly detection data according to the second edge device includes:

[0038] Obtain the abnormal electric energy meter position, abnormal event label and abnormal timing information of the abnormal detection data; perform correlation analysis according to the abnormal electric energy meter position, abnormal event label and abnormal timing information, and obtain the neighborhood electric energy meter detection data corresponding to the abnormal electric energy meter position, the accompanying event detection data of the abnormal event label, and the same timing detection data corresponding to the abnormal timing information; output the associated detection data based on the neighborhood electric energy meter detection data, the accompanying event detection data and the same timing detection data.

[0039] Specifically, after receiving the abnormal detection data, the abnormal electricity meter location, abnormal event label and abnormal timing information are extracted, wherein the abnormal electricity meter location is the water meter location or area where the abnormality occurs, the abnormal event label is the specific event label related to the abnormal data, that is, the typical abnormality marked above, such as equipment failure, leakage detection, temperature abnormality, etc., and the abnormal timing information is the time information of the abnormality, which can help analyze whether the abnormality occurs at a specific time point or period, so as to identify the potential timing rules or the influence of external factors; then, the extracted information is subjected to correlation analysis, and for the abnormal electricity meter location, according to the location Find the detection data of neighboring electric energy meters. Neighboring electric energy meters refer to other water meters or devices whose geographical location is less than or equal to the preset neighborhood distance from the electric energy meter where the abnormality occurs. If a water meter fails, the detection data of neighboring electric energy meters (such as water flow, temperature, pressure, etc.) of other electric energy meters close to it will be identified and collected. The data of these neighboring electric energy meters are used to determine whether there is the same abnormal pattern or chain reaction as the abnormal electric energy meter. For abnormal event labels, the collected neighboring electric energy meter detection data are analyzed through the anomaly detection model to determine whether each neighboring electric energy meter has an abnormality, and assign a label to each neighboring electric energy meter with an abnormality. An event label, these event labels help distinguish the types of anomalies; then the abnormal event label is input into the abnormal association table to identify the accompanying events that may be caused by such events, that is, similar anomalies or anomalies of the same type that may occur. This abnormal association table is constructed based on historical experience and records the association patterns between multiple anomalies. For example, when water pressure anomalies occur, they may be accompanied by pipeline leakage, pump failure and other anomalies; by comparing the event labels of the neighboring electric energy meters with abnormalities with the identified accompanying events that may cause them, if there are accompanying events that may cause them in the event label, the monitoring data of the electric energy meter corresponding to the event label is added to the table. Added to the accompanying event detection data, these data help understand whether the abnormal event is a local problem or a common problem in a larger range; for abnormal time series information, based on the time series information, check whether the neighboring electricity meter detection data within this time period also has abnormalities in the same way, and add the abnormal detection data to the same time series detection data. These data help judge the universality of the abnormality; finally, by integrating the obtained neighboring electricity meter detection data, accompanying event detection data and the same time series detection data, associated detection data is generated to update the abnormal detection data and make the abnormal detection data more complete.

[0040] In one possible implementation, the first edge device and the second edge device are connected to a cloud processor, and the method further includes:

[0041] Obtain a first preset expiration time limit and a second preset expiration time limit, wherein the first preset expiration time limit is greater than the second preset expiration time limit; and perform transmission time limit management on the cloud processor and the first edge device and the second edge device respectively according to the first preset expiration time limit and the second preset expiration time limit.

[0042] Specifically, the first edge device and the second edge device are respectively connected to the cloud processor through a stable communication network, which enables the edge device to share data with the cloud processor for real-time monitoring and data analysis; in order to perform transmission time limit management, the first preset expiration time limit and the second preset expiration time limit are first obtained, wherein the first preset expiration time limit applies to the first edge device, and this time limit is relatively long, because the first edge device is responsible for processing data with a high degree of urgency. Generally, the frequency of receiving such data is low, so a longer retention time can be allowed to reduce frequent data uploads, thereby improving the computing efficiency of the edge device and reducing the transmission frequency; the second preset expiration time limit applies to the second edge device, and this time limit is shorter, because the second edge device is responsible for processing data with a high degree of urgency. Lower data, the reception frequency of this type of data is high. In order to ensure timely processing and response, the data retention time cannot be too long; for the first edge device, after receiving the data, it will decide whether the data should be uploaded to the cloud based on the preset first expiration time limit. If the data stays in the edge device for longer than this expiration time limit (for example, the data has not been processed for a long time or has timed out), the data will no longer be stored in the first edge device, but will be uploaded to the cloud processor immediately; for the second edge device, if the data stays in the second edge device for longer than the corresponding expiration time limit, the data will also be uploaded directly to the cloud processor. In this way, the edge device can be put in a light-load computing state, improve edge processing efficiency, and ensure the real-time and stability of the entire data transmission process.

[0043] In one possible implementation, based on the first preset expiration time limit and the second preset expiration time limit, respectively managing the transmission time limit between the cloud processor and the first edge device and the second edge device, the method includes:

[0044] Detect whether the first edge device includes a first failure identification dataset. If it includes the first failure identification dataset, send the first failure identification dataset to the cloud processor for storage and call, wherein the first failure identification dataset is a dataset whose current data entry time is longer than the first preset failure time limit; detect whether the second edge device includes a second failure identification dataset. If it includes the second failure identification dataset, send the second failure identification dataset to the cloud processor for storage and call, wherein the second failure identification dataset is a dataset whose current data entry time is longer than the second preset failure time limit.

[0045] Specifically, it is first detected whether the first edge device contains a first failure identification data set. This data set contains data stored in the device and whose retention time exceeds the first preset failure time limit, which is timeout data. If the first edge device contains such a data set, in order to ensure edge processing efficiency, these failure identification data will be sent to the cloud processor for storage and call. Similarly, it is also detected whether the second edge device contains a second failure identification data set. This data set also contains data whose retention time exceeds the second preset failure time limit. If such data exists in the second edge device, these failure identification data will be sent to the cloud processor for storage and call. Through this mechanism, it can be ensured that the edge device does not store timeout data for a long time, avoiding the reduction of edge processing efficiency due to data backlog.

[0046] Example 2, based on the same inventive concept as the remote monitoring method of smart water meter integrated with NB-IoT in the above embodiment, Figure 2 As shown, the present application provides a smart water meter remote monitoring device integrated with NB-IoT, wherein the device includes:

[0047] Communication module integration component 1: integrates the first NB-IoT communication module and the second NB-IoT communication module, wherein the first NB-IoT communication module corresponds to the first remote transmission channel, and the second NB-IoT communication module corresponds to the second remote transmission channel; discriminator training component 2: obtains a distributed water meter network, extracts a historical monitoring data set of the distributed water meter network, trains the historical monitoring data set, and obtains a remote transmission channel discriminator; real-time monitoring component 3: performs real-time monitoring of the distributed water meter network and outputs a real-time monitoring data set; data judgment component 4: discriminates the real-time monitoring data set according to the remote transmission channel discriminator to obtain first-category monitoring data and second-category monitoring data, wherein the first-category monitoring data is data greater than or equal to a preset urgency level, and the second-category monitoring data is data less than the preset urgency level; data transmission component 5: transmits the first-category monitoring data to the first edge device based on the first remote transmission channel, and transmits the second-category monitoring data to the second edge device based on the second remote transmission channel.

[0048] Furthermore, the communication module integrated component 1 further includes:

[0049] The central processing unit is connected to the first NB-IoT communication module and the second NB-IoT communication module through an SPI communication bus.

[0050] Furthermore, the discriminator training component 2 further includes:

[0051] By performing feature extraction on the historical monitoring data set, multiple data features are obtained, and the multiple data features include time features, data types, sensor status, and urgency levels; a DNN neural network is initialized, and each of the multiple data features is used as an input node to perform multi-level training on the DNN neural network to obtain data training results; the data training results are tuned by tuning hyperparameters until the model classification accuracy is greater than the preset classification accuracy, thereby obtaining a remote transmission channel discriminator.

[0052] Furthermore, the data transmission component 5 further includes:

[0053] Download a pre-trained anomaly detection model and embed the pre-trained anomaly detection model in the first edge device; perform anomaly detection on the first type of monitoring data based on the pre-trained anomaly detection model to obtain anomaly detection data; obtain a first warning signal based on the anomaly detection data, and the first edge device processes the distributed water meter network based on the first warning signal.

[0054] Furthermore, the data transmission component 5 further includes:

[0055] The abnormality detection data is sent to the second edge device, wherein the first edge device and the second edge device are communicatively connected; a correlation analysis is performed on the received abnormality detection data according to the second edge device to obtain associated detection data of the abnormality detection data; and the abnormality detection data is updated according to the associated detection data to obtain updated abnormality detection data.

[0056] Furthermore, the data transmission component 5 further includes:

[0057] Obtain the abnormal electric energy meter position, abnormal event label and abnormal timing information of the abnormal detection data; perform correlation analysis according to the abnormal electric energy meter position, abnormal event label and abnormal timing information, and obtain the neighborhood electric energy meter detection data corresponding to the abnormal electric energy meter position, the accompanying event detection data of the abnormal event label, and the same timing detection data corresponding to the abnormal timing information; output the associated detection data based on the neighborhood electric energy meter detection data, the accompanying event detection data and the same timing detection data.

[0058] Furthermore, the data transmission component 5 further includes:

[0059] Obtain a first preset expiration time limit and a second preset expiration time limit, wherein the first preset expiration time limit is greater than the second preset expiration time limit; and perform transmission time limit management on the cloud processor and the first edge device and the second edge device respectively according to the first preset expiration time limit and the second preset expiration time limit.

[0060] Furthermore, the data transmission component 5 further includes:

[0061] Detect whether the first edge device includes a first failure identification dataset. If it includes the first failure identification dataset, send the first failure identification dataset to the cloud processor for storage and call, wherein the first failure identification dataset is a dataset whose current data entry time is longer than the first preset failure time limit; detect whether the second edge device includes a second failure identification dataset. If it includes the second failure identification dataset, send the second failure identification dataset to the cloud processor for storage and call, wherein the second failure identification dataset is a dataset whose current data entry time is longer than the second preset failure time limit.

[0062] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0063] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

[0064] This specification and drawings are merely illustrative of the present application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, this application is intended to include such modifications and variations as fall within the scope of this application and its equivalents.

Claims

1. A remote monitoring method for smart water meters integrated with NB-IoT, characterized in that: The method comprises: Integrate a first NB-IoT communication module and a second NB-IoT communication module, wherein the first NB-IoT communication module corresponds to a first remote transmission channel and the second NB-IoT communication module corresponds to a second remote transmission channel; Acquire a distributed water meter network, extract a historical monitoring data set of the distributed water meter network, train the historical monitoring data set, and acquire a remote transmission channel discriminator; Performing real-time monitoring on the distributed water meter network and outputting a real-time monitoring data set; The real-time monitoring data set is discriminated according to the remote transmission channel discriminator to obtain first-category monitoring data and second-category monitoring data, wherein the first-category monitoring data is data with a degree of urgency greater than or equal to a preset degree of urgency, and the second-category monitoring data is data with a degree of urgency less than the preset degree of urgency; Transmitting the first type of monitoring data to a first edge device based on the first remote transmission channel, and transmitting the second type of monitoring data to a second edge device based on the second remote transmission channel; The historical monitoring data set is trained to obtain a long-range transmission channel discriminator, the method comprising: Extracting features from the historical monitoring data set to obtain multiple data features, including time features, data types, sensor states, and urgency levels; Initializing a DNN neural network, using each of the multiple data features as an input node to perform multi-level training on the DNN neural network, and obtaining a data training result; Tuning the data training results by tuning hyperparameters until the model classification accuracy is greater than a preset classification accuracy, thereby obtaining a remote transmission channel discriminator; Transmitting the first type of monitoring data to a first edge device based on the first remote transmission channel, the method comprising: Downloading a pre-trained anomaly detection model and embedding the pre-trained anomaly detection model in the first edge device; Performing anomaly detection on the first type of monitoring data based on the pre-trained anomaly detection model to obtain anomaly detection data; Obtaining a first warning signal according to the abnormality detection data, and the first edge device processing the distributed water meter network according to the first warning signal; After obtaining the anomaly detection data, the method further includes: sending the anomaly detection data to the second edge device, wherein the first edge device and the second edge device are communicatively connected; performing a correlation analysis on the received abnormality detection data according to the second edge device to obtain associated detection data of the abnormality detection data; updating the abnormality detection data according to the associated detection data to obtain updated abnormality detection data; Performing a correlation analysis on the received anomaly detection data according to the second edge device includes: Obtaining abnormal electric energy meter location, abnormal event label and abnormal time sequence information of the abnormal detection data; Performing correlation analysis based on the abnormal electric energy meter location, abnormal event label, and abnormal time series information, respectively, to obtain detection data of neighboring electric energy meters corresponding to the abnormal electric energy meter location, detection data of accompanying events associated with the abnormal event label, and detection data of the same time series corresponding to the abnormal time series information; outputting associated detection data based on the neighborhood electric energy meter detection data, the accompanying event detection data, and the same time series detection data; The first edge device and the second edge device are connected to a cloud processor, and the method further includes: Obtain a first preset expiration time limit and a second preset expiration time limit, wherein the first preset expiration time limit is greater than the second preset expiration time limit; According to the first preset expiration time limit and the second preset expiration time limit, transmission time limit management is performed on the cloud processor and the first edge device and the second edge device respectively.

2. The remote monitoring method for smart water meters integrated with NB-IoT according to claim 1, characterized in that: According to the first preset expiration time limit and the second preset expiration time limit, the method includes: detecting whether the first edge device includes a first failure identification dataset, and if so, sending the first failure identification dataset to the cloud processor for storage and invocation, wherein the first failure identification dataset is a dataset for which the current data entry duration is greater than the first preset failure time limit; Detect whether the second edge device includes a second failure identification data set. If it includes the second failure identification data set, send the second failure identification data set to the cloud processor for storage and call, wherein the second failure identification data set is a data set whose current data entry time is greater than the second preset failure time limit.

3. The remote monitoring method for smart water meters integrated with NB-IoT according to claim 1, characterized in that: The central processing unit is connected to the first NB-IoT communication module and the second NB-IoT communication module through an SPI communication bus.

4. Smart water meter remote monitoring device integrated with NB-IoT, characterized by: The device is used to execute the remote monitoring method of the smart water meter integrated with NB-IoT according to any one of claims 1 to 3, comprising: Communication module integration component: integrating a first NB-IoT communication module and a second NB-IoT communication module, wherein the first NB-IoT communication module corresponds to a first remote transmission channel, and the second NB-IoT communication module corresponds to a second remote transmission channel; Discriminator training component: obtaining a distributed water meter network, extracting a historical monitoring data set of the distributed water meter network, training the historical monitoring data set, and obtaining a remote transmission channel discriminator; Real-time monitoring component: performs real-time monitoring on the distributed water meter network and outputs a real-time monitoring data set; Data judgment component: judge the real-time monitoring data set according to the remote transmission channel discriminator to obtain first-category monitoring data and second-category monitoring data, wherein the first-category monitoring data is data with a degree of urgency greater than or equal to a preset degree of urgency, and the second-category monitoring data is data with a degree of urgency less than the preset degree of urgency; Data transmission component: transmits the first type of monitoring data to a first edge device based on the first remote transmission channel, and transmits the second type of monitoring data to a second edge device based on the second remote transmission channel.

Images