Water environment monitoring method and system based on Internet of Things, and storage medium

Abstract

The invention relates to the technical field of Internet of Things monitoring, and discloses a water environment monitoring method and system based on Internet of Things, and a storage medium. The method comprises the following steps: deploying a plurality of sensor nodes at different water area monitoring points for collecting monitoring signals of a water environment, obtaining attribute information of the sensor nodes, establishing a transmission mode based on the attribute information, and transmitting the monitoring signals to a cloud platform based on the transmission mode; the cloud platform checks whether errors exist in the monitoring signals, if yes, the monitoring signals are corrected, standard data are generated, and if not, the monitoring signals are set as the standard data; summarizing all the standard data to generate a sample set based on the collection categories of the sensor nodes, establishing a water quality prediction model, outputting a water quality dynamic evaluation result by the water quality prediction model based on the sample set, and outputting pollution early warning information based on the water quality dynamic evaluation result; and the cloud platform sends the pollution early warning information to the management end to complete monitoring of the water environment. The water environment monitoring efficiency and accuracy are improved.

Description

Technical Field

[0001] This application relates to the technical field of Internet of Things monitoring, and particularly to a water environment monitoring method, system and storage medium based on the Internet of Things. Background Technique

[0002] The water environment monitoring methods mainly include manual monitoring, sensor monitoring and traditional hydrological monitoring. Manual monitoring mainly involves on-site sampling and analysis by humans, which cannot achieve long-term unattended operation and is restricted by manpower, material resources, financial resources, environmental conditions, etc. Sensor monitoring uses sensor elements such as dissolved oxygen sensors, nitrate nitrogen sensors, and chlorophyll sensors. The detected data has high accuracy and precision. However, since a single node can only monitor the data of one point, the coverage of the monitoring network is not comprehensive enough and the data is discrete. Traditional hydrological monitoring can measure various hydrological elements more accurately. The measurement methods are relatively standardized and normalized, and the data is representative, systematic and complete. However, most of them require the deployment of fixed collection equipment, with high monitoring costs, long data sampling periods, and cannot achieve real-time monitoring. At the same time, data error detection is of great significance for improving the quality of water quality monitoring, achieving comprehensive monitoring and effective management of the water environment. The traditional water quality monitoring system mainly uses wired transmission, and the error detection of data is mainly carried out when the data is transmitted to the server. Once an error is detected, the transmission fails and the data cannot be transmitted successfully. Such defects limit the transmission efficiency and affect the performance of the water environment monitoring system and the data quality.

[0003] Water quality prediction is one of the key objectives of water environment monitoring. The water environment is complex and variable and is affected by various aspects. How to construct an accurate and effective water quality prediction model directly affects the effectiveness of water environment management. The quality of sample data directly affects the performance of the water quality prediction model. There are many influencing factors and it is easily interfered by complex environments. The collected data is difficult to reflect the water quality information in real time and the monitoring error is large. At the same time, a single node cannot represent the overall water quality state of the corresponding basin, which also affects the evaluation effectiveness of the water quality prediction model for the overall water quality state, resulting in problems such as low efficiency and unguaranteed quality of water environment monitoring, and poor water environment monitoring effects.

[0004] Similar prior art includes a Chinese patent application with publication number CN118798649A, which discloses a floating water environment monitoring and early warning method and device, the method comprising the following steps: using a floating water environment monitoring and early warning device to obtain initial water environment monitoring information, optimizing the initial water environment monitoring information, and obtaining water environment monitoring target information; establishing a water environment monitoring system based on the water environment monitoring target information; establishing a water environment pH prediction model based on the target information and the monitoring system to obtain pH prediction results; constructing a water environment turbidity analysis model based on the target information and the monitoring system to obtain turbidity analysis results; obtaining a comprehensive monitoring report of the water environment through the pH prediction results, turbidity analysis results, and the evaluation results of the monitoring system, and matching the environmental early warning plan. The invention constructs a mathematical model and a monitoring system to conduct a comprehensive monitoring and analysis of the water environment, and matches appropriate early warning measures and emergency protection plans, thereby ensuring the healthy and sustainable development of the water environment. There is also a Chinese patent application with publication number CN119269761A, which discloses a water environment monitoring system and method based on big data analysis. The invention relates to the field of environmental monitoring technology, solves the technical problem of focusing only on the monitoring of water quality parameters, while ignoring the impact of biological factors, human activity factors, etc. on the water environment, and reducing the accuracy of the overall monitoring and evaluation of the water environment. The invention can collect a wide range of water environment-related data through the water environment data acquisition module, including water quality, hydrology, biology, population and pollution sources, etc., to achieve comprehensive monitoring of the water environment, and comprehensively consider the impact of biological factors and human activity factors on the water environment. The calculation of the biodiversity index and the biological pollution index reflects the structure and health of the biological community. The water quality monitoring and early warning module can monitor and evaluate in real time according to the water environment value, and timely detect abnormal water quality and generate early warning information or evaluation reports by comparing with the threshold and matching with the water quality grade interval.

[0005] The shortcomings of the existing technology are mainly reflected in the fact that only pH monitoring of the water environment is carried out, which reduces the diversity of water environment monitoring. When conducting comprehensive monitoring of the water environment, there is a lack of evaluation of the stability and timeliness of monitoring data transmission, which reduces the reliability of the evaluation report. In actual situations, it is necessary to ensure the accuracy of data transmission, and at the same time, it is necessary to integrate multiple monitoring signals to conduct comprehensive monitoring of the water environment and water quality prediction and analysis. Summary of the invention

[0006] The present application provides a water environment monitoring method, system and storage medium based on the Internet of Things, which are used to improve the efficiency and accuracy of water environment monitoring based on the Internet of Things.

[0007] In a first aspect, the present application provides a water environment monitoring method based on the Internet of Things, and the water environment monitoring method based on the Internet of Things includes:

[0008] Deploy a variety of sensor nodes at different water monitoring points to collect monitoring signals of the water environment, obtain attribute information of the sensor nodes, establish a transmission mode based on the attribute information, and transmit the monitoring signals to a cloud platform based on the transmission mode;

[0009] The cloud platform checks whether there is an error in the monitoring signal, and if so, corrects the monitoring signal and generates standard data, otherwise, sets the monitoring signal as the standard data;

[0010] Based on the collection category of the sensor node, all the standard data are aggregated to generate a sample set, and a water quality prediction model is established. The water quality prediction model outputs a water quality dynamic assessment result based on the sample set, and outputs pollution warning information based on the water quality dynamic assessment result;

[0011] The cloud platform sends the pollution warning information to the management end to complete the monitoring of the water environment.

[0012] In combination with the first aspect, establishing a transmission mode based on the attribute information includes:

[0013] Acquire the location information of the sensor node based on the attribute information, aggregate all the sensor nodes to generate multiple subsets based on the location information, and configure a wireless analysis device for any of the subsets;

[0014] Set any of the wireless analysis devices as a node, set all of the sensor nodes included in any of the subsets as child nodes, connect the child nodes with the corresponding nodes to generate a block connection diagram, connect the block connection diagrams corresponding to all of the nodes in turn based on the transmission distance to generate a transmission network diagram, and the transmission network diagram generates the transmission mode based on the data transmission between the nodes.

[0015] In combination with the first aspect, transmitting the monitoring signal to the cloud platform based on the transmission mode includes:

[0016] The wireless analysis device receives the monitoring signal based on the communication interface, converts the monitoring signal into a baseband signal, extracts real data and imaginary data from the baseband signal, and aggregates the real data and the imaginary data based on the timestamp to generate sample data;

[0017] Extract the signal strength corresponding to the sample data, set the sample data whose signal strength is greater than or equal to a first preset value as envelope data, demodulate the sample data to generate header data, combine the envelope data with the header data to generate data to be transmitted, and transmit the data to be transmitted to the cloud platform based on the transmission mode.

[0018] In combination with the first aspect, the cloud platform checks whether there is an error in the monitoring signal, including:

[0019] The signal strength change rate of the data to be transmitted is obtained based on the envelope data, and the integrity of the data to be transmitted is obtained based on the header data. If the absolute value of the signal strength change rate is less than a second preset value or the integrity is less than a third preset value, it is determined that there is an error in the monitoring signal, and the monitoring signal is set as an error signal.

[0020] In combination with the first aspect, correcting the monitoring signal and generating standard data includes:

[0021] The wireless analysis device identifies the sensor node corresponding to the error signal and sends a request signal to the sensor node. The sensor node receives the request signal and reads the sample data from the historical record based on the timestamp, regenerates the envelope data corresponding to the sample data based on a first mode and sets it as the first envelope data, transmits the first envelope data to the wireless analysis device, and the wireless analysis device regenerates the standard data based on the first envelope data and transmits it to the cloud platform.

[0022] In combination with the first aspect, the establishment of a water quality prediction model includes:

[0023] Constructing a neural ordinary differential equation, using the neural ordinary differential equation as a prediction model, initializing parameters of the neural ordinary differential equation, extracting sub-sample data from the sample set based on a monitoring target, using an ordinary differential solver to perform time prediction on the sub-sample data, and generating a prediction result;

[0024] Calculating an error between the prediction result and an actual monitoring value in the sub-sample data, deducing backward in time based on the ordinary differential solver, calculating a derivative of an error function with respect to a parameter of the prediction model based on the error, and updating the parameter of the prediction model based on the derivative;

[0025] Repeat this step until the error is less than or equal to a fourth preset value, and set the updated prediction model as the water quality prediction model.

[0026] In combination with the first aspect, constructing the neural ordinary differential equation includes:

[0027] The neural ordinary differential equation is constructed based on a first formula, wherein the first formula is: Where x(t) represents the water quality parameter at a certain time point t in the subsample data, f[x(t), t, w] represents the function generated by neural network learning, w is the parameter of the prediction model, t, t 0 and t 1are different time points, x(t 1 ) is the predicted future time point t 1 The corresponding water quality parameter, x(t 0 ) is the time point t in the sub-sample data 0 Corresponding water quality parameters;

[0028] Setting a parameter category corresponding to the water quality parameter based on the monitoring target, and acquiring the water quality parameter in the sub-sample data based on the parameter category;

[0029] A future prediction time is set, and the prediction result corresponding to the future prediction time is output based on the first formula, and the prediction result is set as the water quality dynamic assessment result.

[0030] In combination with the first aspect, the outputting of pollution warning information based on the water quality dynamic assessment result includes:

[0031] The dynamic change rate of the water quality dynamic assessment result is obtained based on the future prediction time, a threshold interval is set, the risk coefficient of the dynamic change rate is determined based on the threshold interval, all the monitoring targets are summarized based on the risk coefficient, and the pollution warning information is generated.

[0032] In a second aspect, the present application provides a water environment monitoring system based on the Internet of Things, and the water environment monitoring system based on the Internet of Things includes:

[0033] A collection module, used to deploy a variety of sensor nodes at different water monitoring points to collect monitoring signals of the water environment, obtain attribute information of the sensor nodes, establish a transmission mode based on the attribute information, and transmit the monitoring signals to the cloud platform based on the transmission mode;

[0034] A verification module, used for the cloud platform to verify whether there is an error in the monitoring signal, and if so, correct the monitoring signal and generate standard data; otherwise, set the monitoring signal as the standard data;

[0035] An analysis module, for aggregating all the standard data based on the collection category of the sensor node to generate a sample set, and establishing a water quality prediction model, wherein the water quality prediction model outputs a water quality dynamic assessment result based on the sample set, and outputs pollution warning information based on the water quality dynamic assessment result;

[0036] The management module is used for the cloud platform to send the pollution warning information to the management end to complete the monitoring of the water environment.

[0037] A third aspect of the present application provides a computer-readable storage medium, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer executes the above-mentioned water environment monitoring method based on the Internet of Things.

[0038] In the technical solution provided by this application, first, this application deploys various types of sensor nodes at monitoring points in different waters, and realizes comprehensive and refined monitoring of the water environment through multi-level and multi-type sensor deployment, covering different areas and different types of water quality parameters. Then, based on the attribute information of the sensor nodes, the nodes are divided into multiple subsets, and a wireless analysis device is configured for each subset. The wireless analysis device acts as a node, and the sensor nodes in the subset are connected as sub-nodes to generate a block connection diagram and construct a transmission network diagram, which can realize efficient data transmission, reduce data transmission delay and packet loss rate, take into account the transmission distance and node load, ensure the high reliability and stability of the network, and the transmission mode and network structure are highly scalable and can be dynamically adjusted according to changes in the monitoring area and the number of sensor nodes. Finally, a water quality prediction model is established based on the sample set, and the Neural Ordinary Differential Equation is used as the prediction model. The dynamic evaluation results of water quality are generated through time prediction, and pollution warning information is output and sent to the management end. It can process massive monitoring data in real time to ensure the timeliness and accuracy of the data, capture the complex dynamic changes of water quality parameters, and achieve high-precision time prediction and dynamic evaluation. Based on the results of intelligent analysis, pollution warning information is automatically generated to achieve real-time monitoring and warning of the water environment and improve emergency response capabilities.

[0039] This application also uses the cloud platform to perform error detection on the received monitoring signals, including signal strength change rate and data integrity assessment, to ensure the accuracy and integrity of the monitoring data, improve data quality, and recover through retransmission and error correction mechanisms to ensure the stable operation of the monitoring system. Summarize the risk factors of all monitoring targets, generate pollution warning information, achieve accurate warning of water environment changes, and promptly discover potential pollution risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0041] Figure 1 This is a schematic diagram of an embodiment of a water environment monitoring method based on the Internet of Things in an embodiment of the present application;

[0042] Figure 2This is a schematic diagram of an embodiment of a transmission network diagram in an embodiment of the present application;

[0043] Figure 3 A flowchart of an embodiment of generating pollution warning information in an embodiment of the present application;

[0044] Figure 4 This is a schematic diagram of an embodiment of a water environment monitoring system based on the Internet of Things in an embodiment of the present application. DETAILED DESCRIPTION

[0045] The present application embodiment provides a kind of water environment monitoring method, system and storage medium based on the Internet of Things. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or precedence. It should be understood that the data used in this way can be interchangeable in appropriate circumstances, so that the embodiments described here can be implemented in a sequence other than the content illustrated or described here. In addition, the term "including" or "having" and any variation thereof are intended to cover non-exclusive inclusions, for example, the process, method, system, product or equipment comprising 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 equipment.

[0046] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiment of the present application, an embodiment of the water environment monitoring method based on the Internet of Things includes:

[0047] Step S101, deploying a variety of sensor nodes at different water monitoring points to collect monitoring signals of the water environment, obtaining attribute information of the sensor nodes, establishing a transmission mode based on the attribute information, and transmitting the monitoring signals to the cloud platform based on the transmission mode.

[0048] It is understandable that the execution subject of the present application can be a water environment monitoring device based on the Internet of Things, or a terminal or a server, which is not limited here. The present application embodiment is described by taking the server as the execution subject as an example.

[0049] Specifically, representative monitoring points are selected according to factors such as the geographical characteristics of the water area, the direction of water flow, and the distribution of pollution sources. For example, water area monitoring points can be set up in the upper reaches, middle reaches, lower reaches of the river, and at the confluence of the main tributaries, and multiple water area monitoring points can be selected to cover different areas. A variety of sensor nodes are deployed at each monitoring point, including but not limited to water quality sensors, meteorological sensors, flow rate sensors, and pollutant sensors, etc., wherein the water quality sensor is used to monitor water quality parameters such as water temperature, pH value, dissolved oxygen (DO), conductivity, turbidity, chemical oxygen demand (COD), biochemical oxygen demand (BOD), etc., the meteorological sensor is used to monitor meteorological parameters such as air temperature, humidity, wind speed, wind direction, and rainfall, the flow rate sensor is used to monitor water flow speed and water level changes, and the pollutant sensor is used to monitor specific pollutants, such as ammonia nitrogen, total phosphorus, heavy metals, pesticide residues, etc. Attribute information means that each sensor node has a unique identifier, including but not limited to recording its geographical location, deployment time, sensor type, measurement range, accuracy, etc. Information. According to factors such as the geographical location, signal strength, and network coverage of the sensor node, an optimized data transmission path, i.e., a transmission mode, is established. The collected monitoring signals are transmitted to the cloud platform in real time through the above-mentioned transmission mode. The cloud platform refers to a platform built on cloud computing technology that provides multiple services such as computing, storage, networking, and data processing. It provides scalable, on-demand resources and services through the Internet, enabling users to flexibly access and manage these resources without having to build and maintain complex physical infrastructure themselves.

[0050] Step S102: The cloud platform checks whether there is any error in the monitoring signal. If so, the monitoring signal is corrected and standard data is generated. Otherwise, the monitoring signal is set as standard data.

[0051] Specifically, after receiving the monitoring signal transmitted by the sensor node, the cloud platform first performs data verification. If data errors are found, error correction is performed, and the corrected data is set as standard data. The specific verification method will be described later.

[0052] Step S103: Based on the collection category of the sensor node, all standard data are aggregated to generate a sample set, and a water quality prediction model is established. The water quality prediction model outputs a dynamic water quality assessment result based on the sample set, and outputs pollution warning information based on the dynamic water quality assessment result.

[0053] Specifically, all standard data are summarized according to the collection category of the sensor node to generate a sample set, which contains water quality parameters at different time points and different monitoring points. The water quality prediction model refers to a model used to analyze and predict water quality parameters. The model input is current and historical water quality parameter data, and the output is the prediction result. The dynamic water quality assessment result is generated based on the prediction result. The water quality prediction model outputs the dynamic water quality assessment results in real time, such as water quality level (excellent, good, light pollution, moderate pollution, heavy pollution). Based on the assessment results, the cloud platform generates pollution warning information. For example, when the water quality level drops to light pollution, an early warning is triggered.

[0054] Step S104: The cloud platform sends the pollution warning information to the management end to complete the monitoring of the water environment.

[0055] Specifically, the pollution warning information includes the warning level (low, medium, high), warning time, warning area, pollution type and other information. The pollution warning information can be sent to the management end through SMS, email, APP push and other methods. The management end can be the management system of relevant departments such as the environmental protection department and the water conservancy department. The received warning information can trigger corresponding emergency response measures, such as launching emergency plans, increasing monitoring frequency, taking governance measures, etc. Through the above specific implementation methods, the water environment monitoring method based on the Internet of Things can realize real-time monitoring, dynamic evaluation and pollution warning of the water environment, and provide scientific basis and technical support for water environment management.

[0056] In a specific embodiment, establishing a transmission mode based on attribute information includes:

[0057] (1) Acquire location information of sensor nodes based on attribute information, aggregate all sensor nodes to generate multiple subsets based on the location information, and configure a wireless analysis device for any subset.

[0058] (2) Any wireless analysis device is set as a node, and all sensor nodes included in any subset are set as child nodes. The child nodes are connected to corresponding nodes to generate a block connection diagram. The block connection diagrams corresponding to all nodes are connected in sequence based on the transmission distance to generate a transmission network diagram. The transmission network diagram generates a transmission mode based on data transmission between nodes.

[0059] Specifically, the sensor node has a built-in GPS module, which automatically obtains and sends its geographic location information to the cloud platform through attribute information. According to the location information of the sensor node, all sensor nodes are divided into multiple subsets. For example, according to the geographic location, the sensor nodes located in the same geographical area (such as the same river section, the same lake area) are divided into a subset, and according to the density of the sensor nodes, the densely distributed nodes are divided into a subset to ensure that the number of nodes in each subset is moderate, or, a comprehensive division is performed in combination with the geographical area and density factors. For example, in a water environment monitoring system of a large lake, the lake can be divided into four areas: east, west, south, and north, and the sensor nodes in each area constitute a subset. The wireless analysis device refers to a device used to monitor and analyze the radio quality when the sensor node performs wireless transmission, such as an industrial gateway, a wireless base station, etc. The functions that the wireless analysis device should have include data reception and forwarding, data processing and storage, and network connection. Among them, data reception and forwarding refers to the ability to receive data sent by the child node (sensor node) and forward it to the transmission network. Data processing and storage refers to having a certain data processing and storage capability, and being able to perform preliminary data processing and caching. Network connectivity means being able to connect to the Internet and transfer data to the cloud platform.

[0060] Each subset is configured with a wireless analysis device, and it is set as the node of the subset to configure network parameter settings, security settings and data processing rules. All sensor nodes (sub-nodes) in each subset are connected to the corresponding wireless analysis device (node), and the connection method can adopt wireless communication technology (such as Wi-Fi, Zigbee, LoRa, etc.). Each subset and its connected wireless analysis device constitute a block, and the relationship between nodes and sub-nodes in each block is represented as a block connection graph. Based on the transmission distance, the block connection graphs corresponding to all nodes are connected in turn. The connection method can be wireless connection. The block connection graphs corresponding to all wireless analysis devices are connected in turn through wireless networks (such as 4G / 5G, satellite communications) to generate a transmission network graph. All block connection graphs are connected through nodes to generate a complete transmission network graph. The connection relationship between nodes represents the data transmission path. Such as Figure 2 As shown, in the schematic diagram of constructing the transmission network diagram, the sensor nodes A1, A2 and A3 correspond to the wireless analysis device A, the sensor nodes B1, B2 and B3 correspond to the wireless analysis device B, and the sensor nodes C1, C2 and C3 correspond to the wireless analysis device C, wherein → represents the data transmission direction, and the block connection diagram constructed by the wireless analysis devices A, B and C is transmitted hierarchically by transmission distance and finally transmitted to the cloud platform D.

[0061] Based on the transmission network diagram, a transmission mode for data transmission is generated. The transmission mode defines the transmission path and transmission method of data from the sensor node to the cloud platform. Through the above specific implementation method, the transmission mode is established based on the attribute information, which can realize efficient and reliable data transmission between the sensor node and the cloud platform, providing a stable data foundation for the water environment monitoring system.

[0062] In a specific embodiment, transmitting the monitoring signal to the cloud platform based on the transmission mode includes:

[0063] (1) The wireless analysis device receives the monitoring signal based on the communication interface, converts the monitoring signal into a baseband signal, extracts real data and imaginary data from the baseband signal, and generates sample data by aggregating the real data and the imaginary data based on the timestamp.

[0064] (2) extracting the signal strength corresponding to the sample data, setting the sample data whose signal strength is greater than or equal to a first preset value as envelope data, demodulating the sample data to generate header data, combining the envelope data with the header data to generate data to be transmitted, and transmitting the data to be transmitted to the cloud platform based on the transmission mode.

[0065] Specifically, the wireless analysis device is equipped with a variety of communication interfaces, which can support a variety of communication protocols and interface types, such as Ethernet, Wi-Fi, 4G / 5G, etc., to adapt to different network environments and terminal devices. The communication protocol and parameters (such as frequency, bandwidth, modulation mode) between each sensor node and the wireless analysis device need to be pre-configured to ensure that the signal can be correctly received. The wireless analysis device receives monitoring signals from sensor nodes through its communication interface. The monitoring signals may contain various types of data, such as water quality data (pH value, dissolved oxygen, conductivity, etc.), meteorological data (temperature, humidity, wind speed, etc.), pollutant concentration, etc. The baseband signal refers to the original signal that has not been modulated or carrier modulated, and usually contains real and imaginary data. The real data usually contains amplitude information, and the imaginary data contains phase information. For example, digital signal processing (DSP) technology can be used to decompose the signal into real and imaginary parts through fast Fourier transform (FFT). The timestamp refers to the time point used to identify data collection to ensure the timing of the data. The aggregated data forms sample data, each of which contains real and imaginary part information and a corresponding timestamp.

[0066] The signal strength can be calculated by the signal-to-noise ratio (SNR) or the received signal strength indicator (RSSI). The sample data with a signal strength greater than or equal to the first preset value is set as the envelope data. The envelope data usually contains the main information of the signal for subsequent demodulation processing. The envelope data is demodulated to restore the original baseband data. The demodulation process depends on the modulation method between the sensor node and the wireless analysis device, such as amplitude shift keying (ASK), frequency shift keying (FSK) and phase shift keying (PSK). The demodulated data includes header data and payload. The header data includes identification information, length information, check information, etc. of the sample data, which is used for identification and verification of the sample data. The header data may include information such as data length and check code. The envelope data is combined with the header data to generate the data to be transmitted, and transmitted to the cloud platform based on the pre-established transmission mode. Through the above-mentioned specific implementation method, the monitoring signal is transmitted to the cloud platform based on the transmission mode, so that efficient and reliable data transmission can be achieved, ensuring the real-time and accuracy of water environment monitoring data.

[0067] In a specific embodiment, the cloud platform checks whether there is an error in the monitoring signal, including:

[0068] The signal strength change rate of the data to be transmitted is obtained based on the envelope data, and the integrity of the data to be transmitted is obtained based on the header data. If the absolute value of the signal strength change rate is less than the second preset value or the integrity is less than the third preset value, it is determined that there is an error in the monitoring signal, and the monitoring signal is set as an error signal.

[0069] Specifically, the envelope data contains actual water environment monitoring data and signal strength information (such as RSSI, SNR, etc.). Based on the envelope data, the signal strength change rate is calculated, and the signal strength change rate = (current signal strength-previous signal strength) / time interval. For example, if the current signal strength is -80dBm, the previous signal strength is -70dBm, and the time interval is 1 second, the change rate is -10dBm / s. The second preset value can be set according to the specific communication environment and application scenario. If the absolute value of the signal strength change rate is less than the second preset value, it may indicate that the signal strength change is abnormal, and there is signal attenuation or interference.

[0070] The header data is parsed from the received data to be transmitted, and the header data contains the identification information, length information, check information, etc. of the data packet. Based on the check information (such as check code, CRC check, etc.) in the header data, the integrity of the data is evaluated. For example, the check code is used to check the data packet to verify the correctness of the data. The data packet is checked using a cyclic redundancy check (CRC) to detect whether an error occurs during data transmission. The data length indicated in the header is compared with the actual received data length to verify the integrity of the data packet. The third preset value can be set according to the specific data packet structure and application requirements. If the integrity is less than the third preset value, it indicates that the data packet may be damaged or lost. If any condition is met, that is, the absolute value of the signal strength change rate is less than the second preset value or the integrity is less than the third preset value, it is determined that the monitoring signal has an error. Through the above specific implementation method, the cloud platform can effectively check whether there is an error in the monitoring signal, and take corresponding processing measures to ensure the accuracy and reliability of water environment monitoring data.

[0071] In a specific embodiment, correcting the monitoring signal and generating standard data includes:

[0072] The wireless analysis device identifies the sensor node corresponding to the error signal and sends a request signal to the sensor node. The sensor node receives the request signal and reads the sample data from the historical record based on the timestamp, and regenerates the envelope data corresponding to the sample data based on the first mode and sets it as the first envelope data. The first envelope data is transmitted to the wireless analysis device. The wireless analysis device regenerates the standard data based on the first envelope data and transmits it to the cloud platform.

[0073] Specifically, after the wireless analysis device receives the error signal marked by the cloud platform, it first parses the relevant information (sensor node ID, timestamp, error type, etc.) in the error signal, and then locates the corresponding sensor node. Positioning can be achieved by querying the sensor node information table, which contains the ID, geographical location, deployment time and other information of each sensor node. The wireless analysis device generates a request signal for requesting the sensor node to resend the data. After receiving the request signal, the sensor node first parses the signal, confirms the operation to be performed, determines the time point of the data to be resent, and confirms the target sensor node of the request signal. Based on the timestamp obtained by the analysis, the sensor node reads the corresponding sample data from its internal historical records. The historical records are usually stored in the local memory of the sensor node and contain sample data collected in the past period of time.

[0074] The first mode refers to obtaining the corresponding signal strength in the sample data one by one for screening. Specifically, the sensor node calculates the signal strength of each IQ sample (I: in-phase component, Q: orthogonal component), where: The calculated signal strength data are arranged in chronological order to construct the first envelope data. The generated first envelope data is transmitted to the wireless analysis device through the wireless communication interface. After receiving the first envelope data, the wireless analysis device performs data analysis and verification. The verification process can adopt the same verification method as the sensor node to ensure the correctness of the data. The standard data generation process can be achieved through data processing, data format conversion and data verification. The regenerated standard data is transmitted to the cloud platform through the transmission mode. Through the above specific implementation method, errors in the monitoring signal can be effectively corrected, and accurate standard data can be generated to ensure the reliability and continuity of water environment monitoring data.

[0075] In a specific embodiment, establishing a water quality prediction model includes:

[0076] (1) Construct a neural ordinary differential equation, use the neural ordinary differential equation as a prediction model, initialize the parameters of the neural ordinary differential equation, extract sub-sample data from the sample set based on the monitoring target, use the ordinary differential solver to perform time prediction on the sub-sample data, and generate prediction results.

[0077] (2) Calculate the error between the prediction result and the actual monitoring value in the sub-sample data, deduce the time backward based on the ordinary differential solver, calculate the derivative of the error function with respect to the parameters of the prediction model based on the error, and update the parameters of the prediction model based on the derivative.

[0078] (3) Repeat this step until the error is less than or equal to a fourth preset value, and set the updated prediction model as the water quality prediction model.

[0079] Specifically, the neural ordinary differential equation is a new type of deep learning model that combines neural networks with ordinary differential equations, which can effectively process continuous-time data and complex dynamic systems. To initialize the parameters of the neural ordinary differential equation, common initialization methods such as Xavier initialization and He initialization can be used to ensure the stability of model training. The monitoring target refers to the type of water quality monitored by the water area monitoring point. For example, water quality parameters such as pH value, dissolved oxygen, and conductivity at a specific monitoring point can be selected as sub-sample data. The extracted sub-sample data is pre-processed (normalization, missing value processing, etc.). Using an ordinary differential equation solver (such as the Euler method and the Runge-Kutta method) to perform time prediction on the sub-sample data and generate prediction results can avoid reducing the accuracy of the prediction results when there are missing values ​​in the sub-sample data.

[0080] Error calculation can use indicators such as mean square error (MSE) and mean absolute error (MAE), which is the error function. Based on the ordinary differential solver, the time is deduced backward to perform error back propagation and parameter update. The gradient descent method (such as Adam, SGD and other optimization algorithms) is used to update the parameters of the prediction model based on the derivative. For example, the gradient of the error to the model parameters is calculated through the back propagation algorithm, and the model parameters are updated according to the gradient direction and step size.

[0081] Repeat the above time prediction, error calculation and parameter update steps until the error is less than or equal to the fourth preset value, and the fourth preset value can be set according to the specific application scenario and accuracy requirements. Set the updated neural ordinary differential equation model as the final water quality prediction model, and evaluate the model, for example, using a test set for verification, and calculate the prediction accuracy and generalization ability of the model. Through the above specific implementation method, an efficient and accurate water quality prediction model can be constructed to achieve dynamic evaluation of the water environment and pollution warning.

[0082] In a specific embodiment, constructing a neural ordinary differential equation includes:

[0083] (1) Construct the Neural Ordinary Differential Equation based on the first formula, which is: Among them, x(t) represents the water quality parameter at a certain time point t in the subsample data, f[x(t), t, w] represents the function generated by neural network learning, w is the parameter of the prediction model, t, t 0 and t 1 are different time points, x(t 1 ) is the predicted future time point t 1 The corresponding water quality parameter, x(t 0 ) is the time point t in the subsample data 0 Corresponding water quality parameters.

[0084] (2) Parameter categories corresponding to water quality parameters are set based on monitoring objectives, and water quality parameters are obtained from subsample data based on parameter categories.

[0085] (3) Setting a future prediction time, outputting a prediction result corresponding to the future prediction time based on the first formula, and setting the prediction result as a water quality dynamic assessment result.

[0086] Specifically, in the first formula, f[x(t), t, w] is implemented by a neural network, which can adopt a multi-layer perceptron (MLP), convolutional neural network (CNN), recurrent neural network (RNN) and other architectures. The specific choice depends on the data characteristics and application requirements. This formula describes how the rate of change of the water quality parameter x(t) at time t is determined by the current state x(t), time t and model parameter w. By solving this ordinary differential equation, the future time point t can be predicted. 1Water quality parameter x(t 1 ).

[0087] Based on the monitoring objectives (such as water quality type and monitoring points), the parameter categories corresponding to the water quality parameters are set, for example, water temperature, turbidity, dissolved oxygen, etc. corresponding to physical parameters, pH value, conductivity, chemical oxygen demand (COD), biochemical oxygen demand (BOD), etc. corresponding to chemical parameters, and algae density, total bacteria count, etc. corresponding to biological parameters. Then, the water quality parameters of the corresponding parameter categories are obtained from the subsample data as historical data.

[0088] Set the future time point to be predicted, that is, the future prediction time. Through the time integration method, the water quality parameters at future time points are gradually predicted to generate prediction results. For example, the water quality dynamic assessment results in the following format can be generated: the temperature at the future prediction time t3 is 25℃, the pH value is 7.2, and the dissolved oxygen is 8.5mg / L.

[0089] In a specific embodiment, outputting pollution warning information based on the water quality dynamic assessment result includes:

[0090] The dynamic change rate of the dynamic water quality assessment results is obtained based on the future prediction time, and the threshold interval is set. The risk coefficient of the dynamic change rate is determined based on the threshold interval. All monitoring targets are summarized based on the risk coefficient to generate pollution warning information.

[0091] Specifically, the dynamic change rate refers to the rate of change of water quality parameters over time. A threshold range (Threshold Range) is set for each water quality parameter to determine the risk level of its change rate. The threshold range can be set according to water quality standards, historical data and environmental factors, or it can be dynamically adjusted according to factors such as season, geographical location, and water body type. The risk factor can be quantified using the following method. Level division refers to dividing the risk into three levels: low, medium, and high, corresponding to different change rates and threshold ranges. Numerical scoring refers to assigning a numerical value to each risk level, for example, low risk is 1, medium risk is 2, and high risk is 3. The risk factors of all monitoring targets (water quality parameters) are summarized and arranged. The pollution warning information may include warning level, warning time, warning area, water quality parameters of major risks, recommended measures and other information. Through the above-mentioned specific implementation methods, pollution warning information can be effectively output based on the results of dynamic water quality assessment to achieve real-time monitoring and early warning of the water environment, such as Figure 3 As shown, a specific flow chart for generating pollution warning information.

[0092] The above describes the water environment monitoring method based on the Internet of Things in the embodiment of the present application. The following describes the water environment monitoring system based on the Internet of Things in the embodiment of the present application. Figure 4In the embodiment of the present application, an embodiment of the water environment monitoring system based on the Internet of Things includes:

[0093] The acquisition module is used to deploy various sensor nodes at different water monitoring points to collect monitoring signals of the water environment, obtain the attribute information of the sensor nodes, establish a transmission mode based on the attribute information, and transmit the monitoring signals to the cloud platform based on the transmission mode.

[0094] The verification module is used by the cloud platform to verify whether there are errors in the monitoring signal. If so, the monitoring signal is corrected and standard data is generated. Otherwise, the monitoring signal is set as standard data.

[0095] The analysis module is used to generate a sample set by aggregating all standard data based on the collection category of the sensor node, and to establish a water quality prediction model. The water quality prediction model outputs a dynamic water quality assessment result based on the sample set, and outputs pollution warning information based on the dynamic water quality assessment result.

[0096] The management module is used by the cloud platform to send pollution warning information to the management end to complete the monitoring of the water environment.

[0097] Through the synergy of the above components, first of all, this application deploys various types of sensor nodes at different water monitoring points, and realizes comprehensive and refined monitoring of the water environment through multi-level and multi-type sensor deployment, covering different regions and different types of water quality parameters. Then, based on the attribute information of the sensor nodes, the nodes are divided into multiple subsets, and a wireless analysis device is configured for each subset. The wireless analysis device acts as a node, and the sensor nodes in the subset are connected as sub-nodes to generate a block connection diagram and construct a transmission network diagram, which can realize efficient data transmission, reduce data transmission delay and packet loss rate, take into account the transmission distance and node load, ensure the high reliability and stability of the network, and the transmission mode and network structure are highly scalable and can be dynamically adjusted according to changes in the monitoring area and the number of sensor nodes. Finally, a water quality prediction model is established based on the sample set, and the Neural Ordinary Differential Equation is used as the prediction model. The dynamic evaluation results of water quality are generated through time prediction, and pollution warning information is output and sent to the management end. It can process massive monitoring data in real time to ensure the timeliness and accuracy of the data, capture the complex dynamic changes of water quality parameters, and achieve high-precision time prediction and dynamic evaluation. Based on the results of intelligent analysis, pollution warning information is automatically generated to achieve real-time monitoring and warning of the water environment and improve emergency response capabilities.

[0098] This application also performs error detection on the monitored signals received through the cloud platform, including the rate of change of signal strength and the evaluation of data integrity, to ensure the accuracy and integrity of the monitored data, improve data quality, and can also be restored through retransmission and error correction mechanisms to ensure the stable operation of the monitoring system. The risk coefficients of all monitored targets are aggregated to generate pollution warning information, achieving precise warning of water environment changes and timely discovery of potential pollution risks.

[0099] This application also provides a computer-readable storage medium. The computer-readable storage medium can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions run on a computer, the computer is made to execute the steps of the water environment monitoring method based on the Internet of Things.

[0100] Those skilled in the art can 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 foregoing method embodiments and will not be elaborated herein.

[0101] 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 this application, in essence, or the part that contributes to the prior art, or all or part of this 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 to cause a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0102] As described above, the above embodiments are only used to illustrate the technical solutions of this application and are not intended to limit them; although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of this application.

Claims

1. A water environment monitoring method based on the Internet of Things, characterized in that: The water environment monitoring method based on the Internet of Things includes: Deploy a variety of sensor nodes at different water monitoring points to collect monitoring signals of the water environment, obtain attribute information of the sensor nodes, establish a transmission mode based on the attribute information, and transmit the monitoring signals to a cloud platform based on the transmission mode; The cloud platform checks whether there is an error in the monitoring signal, and if so, corrects the monitoring signal and generates standard data, otherwise, sets the monitoring signal as the standard data; Based on the collection category of the sensor node, all the standard data are aggregated to generate a sample set, and a water quality prediction model is established. The water quality prediction model outputs a water quality dynamic assessment result based on the sample set, and outputs pollution warning information based on the water quality dynamic assessment result; The cloud platform sends the pollution warning information to the management end to complete the monitoring of the water environment.

2. The water environment monitoring method based on the Internet of Things according to claim 1 is characterized in that: Establishing a transmission mode based on the attribute information includes: Acquire the location information of the sensor node based on the attribute information, aggregate all the sensor nodes to generate multiple subsets based on the location information, and configure a wireless analysis device for any of the subsets; Set any of the wireless analysis devices as a node, set all of the sensor nodes included in any of the subsets as child nodes, connect the child nodes with the corresponding nodes to generate a block connection diagram, connect the block connection diagrams corresponding to all of the nodes in turn based on the transmission distance to generate a transmission network diagram, and the transmission network diagram generates the transmission mode based on the data transmission between the nodes.

3. The water environment monitoring method based on the Internet of Things according to claim 2 is characterized in that: Transmitting the monitoring signal to the cloud platform based on the transmission mode includes: The wireless analysis device receives the monitoring signal based on the communication interface, converts the monitoring signal into a baseband signal, extracts real data and imaginary data from the baseband signal, and aggregates the real data and the imaginary data based on the timestamp to generate sample data; Extract the signal strength corresponding to the sample data, set the sample data whose signal strength is greater than or equal to a first preset value as envelope data, demodulate the sample data to generate header data, combine the envelope data with the header data to generate data to be transmitted, and transmit the data to be transmitted to the cloud platform based on the transmission mode.

4. The water environment monitoring method based on the Internet of Things according to claim 3 is characterized in that: The cloud platform checks whether there is an error in the monitoring signal, including: The signal strength change rate of the data to be transmitted is obtained based on the envelope data, and the integrity of the data to be transmitted is obtained based on the header data. If the absolute value of the signal strength change rate is less than a second preset value or the integrity is less than a third preset value, it is determined that there is an error in the monitoring signal, and the monitoring signal is set as an error signal.

5. The water environment monitoring method based on the Internet of Things according to claim 4 is characterized in that: Correcting the monitoring signal and generating standard data, including: The wireless analysis device identifies the sensor node corresponding to the error signal and sends a request signal to the sensor node. The sensor node receives the request signal and reads the sample data from the historical record based on the timestamp, regenerates the envelope data corresponding to the sample data based on a first mode and sets it as the first envelope data, transmits the first envelope data to the wireless analysis device, and the wireless analysis device regenerates the standard data based on the first envelope data and transmits it to the cloud platform.

6. The water environment monitoring method based on the Internet of Things according to claim 1 is characterized in that: The water quality prediction model is established, comprising: Constructing a neural ordinary differential equation, using the neural ordinary differential equation as a prediction model, initializing parameters of the neural ordinary differential equation, extracting sub-sample data from the sample set based on a monitoring target, using an ordinary differential solver to perform time prediction on the sub-sample data, and generating a prediction result; Calculating an error between the prediction result and an actual monitoring value in the sub-sample data, deducing backward in time based on the ordinary differential solver, calculating a derivative of an error function with respect to a parameter of the prediction model based on the error, and updating the parameter of the prediction model based on the derivative; This step is repeated until the error is less than or equal to a fourth preset value, and the updated prediction model is set as the water quality prediction model.

7. The water environment monitoring method based on Internet of Things according to claim 6 is characterized in that: The method of constructing a neural ordinary differential equation comprises: The neural ordinary differential equation is constructed based on a first formula, wherein the first formula is: Wherein, x(t) represents the water quality parameter at a certain time point t in the subsample data, f[x(t), t, w] represents the function generated by neural network learning, w is the parameter of the prediction model, t, t0 and t1 are all different time points, x(t1) is the water quality parameter corresponding to the predicted future time point t1, and x(t0) is the water quality parameter corresponding to the time point t0 in the subsample data; Setting a parameter category corresponding to the water quality parameter based on the monitoring target, and acquiring the water quality parameter in the sub-sample data based on the parameter category; A future prediction time is set, and the prediction result corresponding to the future prediction time is output based on the first formula, and the prediction result is set as the water quality dynamic assessment result.

8. The water environment monitoring method based on Internet of Things according to claim 7 is characterized in that: The outputting of pollution warning information based on the water quality dynamic assessment result includes: The dynamic change rate of the water quality dynamic assessment result is obtained based on the future prediction time, a threshold interval is set, the risk coefficient of the dynamic change rate is determined based on the threshold interval, all the monitoring targets are summarized based on the risk coefficient, and the pollution warning information is generated.

9. A water environment monitoring system based on the Internet of Things, characterized in that: The water environment monitoring system based on the Internet of Things includes: A collection module, used to deploy a variety of sensor nodes at different water monitoring points to collect monitoring signals of the water environment, obtain attribute information of the sensor nodes, establish a transmission mode based on the attribute information, and transmit the monitoring signals to the cloud platform based on the transmission mode; A verification module, used for the cloud platform to verify whether there is an error in the monitoring signal, and if so, correct the monitoring signal and generate standard data; otherwise, set the monitoring signal as the standard data; An analysis module, for aggregating all the standard data based on the collection category of the sensor node to generate a sample set, and establishing a water quality prediction model, wherein the water quality prediction model outputs a water quality dynamic assessment result based on the sample set, and outputs pollution warning information based on the water quality dynamic assessment result; The management module is used for the cloud platform to send the pollution warning information to the management terminal to complete the monitoring of the water environment.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the water environment monitoring method based on the Internet of Things as described in any one of claims 1-8 is implemented.

Images