A pipe flow and water quality monitoring device and method
By integrating flow and water quality monitoring modules and using LSTM models for intelligent analysis, the problems of scattered pipeline monitoring systems and inaccurate early warnings have been solved, achieving efficient and economical blockage prediction and early warning.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NANJING BAFANG CONSTR ENG INSPECTION CO LTD
- Filing Date
- 2026-04-13
- Publication Date
- 2026-07-14
AI Technical Summary
In existing technologies, pipeline flow and water quality monitoring systems are scattered, costly, and lack intelligent early warning, resulting in inaccurate blockage predictions and complex systems that make it difficult to achieve comprehensive early warning.
Design an integrated device that combines a flow monitoring module and a water quality detection module, and uses a long short-term memory neural network for data analysis to achieve intelligent prediction and graded early warning.
Through integrated design and intelligent analysis, equipment costs and space occupation are reduced, the accuracy and timeliness of early warnings are improved, false alarms and missed alarms are reduced, and the allocation of maintenance resources is optimized.
Smart Images
Figure CN122384902A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of pipeline monitoring technology, specifically to a monitoring device and method for pipeline flow and water quality. Background Technology
[0002] In numerous fields such as modern industrial production, urban water supply, wastewater treatment, and environmental monitoring, real-time monitoring of the flow rate and water quality of liquids within pipelines is crucial. Flow monitoring can be used for metering, process control, and leak detection; water quality monitoring is related to product quality, environmental protection, and public health. Traditionally, flow monitoring and water quality monitoring are often performed by two separate sets of equipment, installed at different locations within the pipeline. This results in complex systems, high costs, and fragmented data, making it difficult to achieve correlation analysis and comprehensive early warning.
[0003] Furthermore, during long-term operation, pipelines are prone to localized blockages due to factors such as impurity deposition, scaling, biofilm growth, or external pressure. Early blockages are often difficult to detect; by the time flow rate drops significantly or pressure changes abruptly, the blockage is already quite severe, potentially leading to production interruptions, equipment damage, or even safety accidents. Currently, pipeline blockage prediction relies mainly on experience or simple threshold alarms, lacking intelligent prediction methods based on big data and artificial intelligence, resulting in insufficient accuracy and timeliness of early warnings.
[0004] In recent years, with the development of IoT and AI technologies, combining sensor data with machine learning algorithms for predictive maintenance of equipment has become a trend. However, in the field of pipeline flow and water quality monitoring, there is still no integrated and intelligent device that can simultaneously monitor flow and water quality and intelligently predict blockage risks based on historical data.
[0005] Therefore, developing a pipeline monitoring device and method that integrates flow monitoring, water quality testing, and intelligent early warning is of great significance for improving pipeline operation reliability, reducing maintenance costs, and ensuring production and environmental safety. Summary of the Invention
[0006] The purpose of this invention is to provide a device and method for monitoring pipeline flow and water quality, aiming to at least solve the problems of system dispersion, high cost, and lack of intelligent early warning in the aforementioned background technology. By integrating a flow monitoring module and a water quality detection module through an integrated design, and utilizing a long short-term memory neural network to perform in-depth analysis of the monitoring data, intelligent prediction and graded early warning of pipeline blockage probability can be achieved, thereby improving the intelligence level and accuracy of pipeline monitoring and early warning.
[0007] To achieve the above objectives, the present invention provides the following technical solution: A device for monitoring pipeline flow and water quality, comprising: A housing, wherein the housing has a cavity for containing liquid; A flow monitoring module is located at the bottom of the interior of the housing, used to monitor the flow rate of liquid flowing into the housing and output a flow signal; A water quality detection module is disposed inside the upper part of the housing, and is used to detect at least one water quality parameter of the liquid inside the housing and output a water quality signal; The system also includes a processor module electrically connected to the flow monitoring module and the water quality detection module. The processor module includes a signal acquisition unit, a data storage unit, a communication unit, and an intelligent analysis unit. The intelligent analysis unit is configured to generate pipeline status early warning information based on a pre-trained time-series prediction model and the flow signal and water quality signal.
[0008] As a further embodiment of the present invention, the flow monitoring module includes a differential pressure transmitter and at least one inlet pipe; one end of the inlet pipe is connected to an external pipeline, and the other end is connected to the inlet of the differential pressure transmitter; the differential pressure transmitter is used to measure the pressure difference when the liquid flows through and calculate the flow rate accordingly; the flow monitoring module also includes an outlet pipe, one end of which is connected to the outlet of the differential pressure transmitter, and the other end extends to the upper part of the interior of the housing, for introducing the liquid that has undergone flow measurement into the detection area of the water quality detection module.
[0009] As a further aspect of the present invention, the water quality detection module includes: The detector body has an inner partition and an outer cylinder, forming multiple detection chambers; A water level detection sensor is installed at a predetermined height on the detector body to detect whether the liquid level has reached the detection requirements; At least one water quality sensor is disposed in the detection chamber for detecting one or more of the following: pH value, turbidity, conductivity, dissolved oxygen, or concentration of a specific ion.
[0010] As a further aspect of the present invention, a sleeve is fitted onto the lower end of the detector body, and the side wall of the sleeve is provided with a plurality of water passage holes for allowing liquid to enter the detector body uniformly and slowly.
[0011] As a further embodiment of the present invention, the top of the housing is provided with a cover, and the cover is provided with an overflow port; the bottom of the housing is provided with a base, and the base is provided with a plug and a groove for fixing the flow monitoring module; the processor module also includes a vibration sensor and a temperature sensor; and the intelligent analysis unit is further configured to fuse vibration data and temperature data to correct the warning results.
[0012] A method for monitoring pipeline flow and water quality includes the following steps: S1. The flow rate data of the liquid in the pipeline is collected in real time through the flow monitoring module, and the water quality data of the liquid is collected in real time through the water quality detection module. S2. Input the collected flow data and water quality data into the intelligent analysis unit of the processor module; S3. The intelligent analysis unit calls the pre-trained time series prediction model to analyze the real-time data and calculate its deviation from the normal operating condition baseline. S4. If the deviation exceeds the preset threshold, an early warning message is generated and sent to the monitoring center or mobile terminal via the communication unit. As a further embodiment of the present invention, a model training step is included before step S3: T1. Collect historical flow and water quality data under normal operating conditions to construct a training dataset; T2. Preprocess the training dataset, including data cleaning, normalization, and serialization; T3. Construct a long short-term memory neural network model and train it using the training dataset to obtain a baseline model under normal operating conditions.
[0013] As a further aspect of the present invention, the preprocessing also includes: dividing different scenarios according to pipeline characteristics, and training a corresponding baseline model for each scenario.
[0014] As a further embodiment of the present invention, the generation of early warning information in step S4 includes: calculating the probability of congestion and classifying the early warning level according to the degree of deviation; when the deviation exceeds 20% but is less than 40%, it is a general attention level; when the deviation is between 40% and 60%, it is a warning level; when the deviation exceeds 60%, it is an emergency level. It also includes step S5: reviewing and correcting the early warning information by combining pipeline vibration data and temperature data; if both vibration and temperature data are abnormal, the early warning level is increased. The flow data includes instantaneous flow and cumulative flow; the water quality data includes at least one of pH value, turbidity, conductivity, dissolved oxygen or specific ion concentration.
[0015] As a further embodiment of the present invention, the long short-term memory neural network model includes an input layer, at least one LSTM hidden layer, and an output layer; the number of nodes in the input layer is the same as the number of input features, and the output layer outputs the predicted value or deviation value for the next time step.
[0016] Compared with the prior art, the beneficial effects of the present invention are: 1. When using this invention, flow monitoring and water quality detection are integrated into the same device, sharing the same housing, processor and power supply, which reduces the number of devices, lowers installation costs and space occupation, and facilitates maintenance and management. In addition to flow and water quality, it can also integrate data such as vibration and temperature to review and correct the early warning results, further improving the reliability of the early warning and reducing false alarms and missed alarms.
[0017] 2. When using this invention, different early warning levels are divided according to the degree of deviation, and a blockage probability estimate is provided, which provides operation and maintenance personnel with an intuitive basis for decision-making, making it easier to take graded response measures and optimize the allocation of maintenance resources.
[0018] 3. When using this invention, the modular design makes it easy to replace and upgrade sensors; the self-diagnostic function can detect sensor faults in a timely manner, ensuring long-term stable operation of the system; through 4G / 5G and other wireless communication modules, data can be uploaded to the cloud platform or monitoring center in real time, realizing remote unattended monitoring and facilitating centralized management of multiple sites. Attached Figure Description
[0019] Figure 1 This is a three-dimensional structural diagram of a pipeline flow and water quality monitoring device.
[0020] Figure 2 This is a schematic diagram of the exploded structure of a pipeline flow and water quality monitoring device.
[0021] Figure 3 This is a schematic diagram showing the exploded structure of various components of a pipeline flow and water quality monitoring device.
[0022] Figure 4 This is a schematic diagram of the top cover structure of a pipeline flow and water quality monitoring device.
[0023] Figure 5 This is a schematic diagram of the water quality detection module structure of a pipeline flow and water quality monitoring device.
[0024] Figure 6 This is a schematic diagram of the internal structure of the water quality detection module in a pipeline flow and water quality monitoring device.
[0025] Figure 7 This is a schematic diagram of the support structure for a pipeline flow and water quality monitoring device.
[0026] Figure 8 This is a schematic diagram of the internal structure of a pipeline flow and water quality monitoring device.
[0027] Figure 9 This is a schematic diagram of the bottom structure of the casing of a pipeline flow and water quality monitoring device.
[0028] Figure 10This is a schematic diagram of the flow monitoring module structure of a pipeline flow and water quality monitoring device.
[0029] In the diagram: 1. Cover; 101. Support; 102. Overflow port; 2. Water quality detection module; 201. Processor module; 21. Water level sensor; 22. Detection box cover; 23. Detector body; 231. Sealing ring; 232. Inner partition; 234. Outer cylinder; 24. Sleeve; 241. Water passage hole; 3. Shell; 31. Water outlet; 32. Water tank; 33. Connection hole; 34. Placement slot; 4. Flow monitoring module; 41. Inlet pipe; 42. Outlet pipe; 43. Power supply port; 5. Base; 51. Plug; 52. Groove. Detailed Implementation
[0030] To address the problem of [the aforementioned situation], the present invention provides a device for monitoring pipeline flow and water quality.
[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0032] Please see Figure 1-10 This invention provides a device for monitoring pipeline flow and water quality, comprising: The housing 3 has a cavity inside for containing liquid; The flow monitoring module 4 is located at the bottom of the interior of the housing 3 and is used to monitor the flow rate of liquid flowing into the housing 3 and output a flow signal. The water quality detection module 2 is located inside the upper part of the housing 3 and is used to detect at least one water quality parameter of the liquid inside the housing 3 and output a water quality signal. The processor module 201 is electrically connected to the flow monitoring module 4 and the water quality detection module 2. The processor module 201 includes a signal acquisition unit, a data storage unit, a communication unit, and an intelligent analysis unit. The intelligent analysis unit is configured to generate pipeline status early warning information based on a pre-trained time-series prediction model and the flow signal and water quality signal.
[0033] The time-series prediction model is a Long Short-Term Memory (LSTM) neural network model. The intelligent analysis unit is further configured to: input real-time collected flow and water quality data into the LSTM model, calculate the deviation from the normal operating baseline, and trigger an early warning signal when the deviation exceeds a preset threshold. The LSTM model is trained using historical flow and water quality data under normal operating conditions. The historical data spans at least one complete production cycle or at least one year to ensure the model fully learns the fluctuation patterns under normal operating conditions.
[0034] The housing 3 is cylindrical or square in shape and made of rust-resistant material (such as stainless steel or engineering plastic), forming an internal cavity to contain liquid. The top of the housing 3 is open and is sealed to the cover 1 via a support 101. The bottom of the housing 3 is fixedly connected to the base 5, and a support leg or mounting flange can be installed under the base 5 to fix the equipment to the ground or a bracket. The cover 1 is provided with an overflow port 102. When the liquid level inside the housing 3 is too high, excess liquid can be discharged from the overflow port 102 to prevent overflow. The overflow port 102 can be connected to a return pipe or a discharge pipe. The base 5 is provided with a plug 51 and a groove 52. The plug 51 is used to pass laterally through the bottom of the housing 3 to lock the flow monitoring module 4 in a predetermined position; the groove 52 is used to accommodate and position the lower part of the flow monitoring module 4 to prevent it from shaking.
[0035] The flow monitoring module 4 is located at the bottom of the housing 3 and includes an inlet pipe 41, an outlet pipe 42, a power supply port 43, a differential pressure transmitter 404, and a flow calculation unit 405. One end of the inlet pipe 41 extends outside the housing 3 through a connection hole 303 and connects to the pipe 6 to be measured; the other end connects to the inlet of the differential pressure transmitter 404. The differential pressure transmitter 404 has a throttling element (such as an orifice plate or nozzle) or directly measures the pressure difference between upstream and downstream, and calculates the instantaneous flow rate based on the pressure difference. The flow calculation unit 405 (which can be integrated into the differential pressure transmitter or set up independently) converts the pressure difference signal into flow data and outputs it to the processor module 201. One end of the outlet pipe 42 connects to the outlet of the differential pressure transmitter 404, and the other end extends upward to the upper part of the housing 3, near the bottom of the water quality detection module 2, so as to guide the liquid after flow measurement into the water quality detection area. The power supply port 43 is used to provide power (such as 24V DC) to the flow monitoring module 4.
[0036] The differential pressure transmitter measures the pressure difference as a liquid flows through it and calculates the flow rate accordingly. The flow monitoring module also includes an outlet pipe, one end of which is connected to the outlet of the differential pressure transmitter, and the other end extends to the upper part of the housing to guide the liquid that has passed through the flow measurement into the detection area of the water quality detection module.
[0037] The water quality detection module 2 is located inside the upper part of the housing 3, and includes a detector body 204, a detection box cover 203, a water level sensor 202, and multiple water quality sensors (such as a pH sensor 21, a turbidity sensor 22, and a conductivity sensor 23). The detector body 204 is cylindrical, with an inner partition 206 and an outer cylinder 207 dividing the interior into multiple detection chambers, each of which can accommodate one type of sensor. A sleeve 208 is fitted onto the lower end of the detector body 204, and the side wall of the sleeve 208 has multiple water passage holes 209 to allow liquid to slowly and evenly enter the detector body 204, avoiding water flow impact from affecting measurement accuracy. The detection box cover 203 closes to the top of the detector body 204, forming a closed detection space. The water level sensor 202 (such as a float switch or a capacitive level switch) is installed at a predetermined height on the detector body 204. When the liquid level rises to this height, it indicates that the sensor is completely submerged in liquid, and detection can begin. Each water quality sensor converts the detection signal into an electrical signal and outputs it to the processor module 201.
[0038] The detector body has an inner partition and an outer cylinder, forming multiple detection chambers to accommodate different types of sensors. A water level sensor is positioned at a predetermined height within the detector body to detect whether the liquid level has reached the required level, ensuring the sensor is completely submerged during detection. A water quality sensor is located within the detection chamber and can detect one or more of the following: pH value, turbidity, conductivity, dissolved oxygen, or concentration of specific ions.
[0039] The processor module 201 is installed at the bottom of the detector body 204 or in a separately located control box. It includes a signal acquisition unit (such as an A / D converter), a data storage unit (such as an SD card or Flash), a communication unit (such as a 4G / 5G module or a Wi-Fi module), an intelligent analysis unit (such as an ARM processor or an edge computing chip), and a power management unit. The processor module 201 is electrically connected to the flow monitoring module 4, the water quality detection module 2, and possibly additional vibration and temperature sensors. It collects data in real time, performs local analysis, and uploads the data and early warning information to the monitoring center or mobile terminal via the communication unit.
[0040] The top of the housing is equipped with a cover, which has an overflow port to prevent liquid from overflowing. The bottom of the housing is equipped with a base, which has a plug and groove for fixing the flow monitoring module, facilitating module installation and positioning.
[0041] This invention also provides a method for monitoring pipeline flow and water quality, applied to the aforementioned monitoring equipment, comprising the following steps: Step S1: Data Acquisition The flow monitoring module 4 collects instantaneous and cumulative flow at a fixed frequency (e.g., once per minute), while the water quality detection module 2 collects parameters such as pH, turbidity, and conductivity at the same or lower frequency (e.g., once every 5 minutes). All data is timestamped and temporarily stored in the data storage unit. Step S2: Data Preprocessing The processor module 201 preprocesses the raw data, including: Outlier removal: Use the 3σ principle or box plot method to identify and remove outliers that significantly deviate from the normal range; Missing value imputation: For data missing due to temporary sensor failure, linear interpolation or forward imputation method is used to fill in the missing data. Normalization: Scaling data of each dimension to the [0,1] interval to eliminate the influence of dimensions; Serialization: Continuous data is truncated into fixed-length time windows (e.g., each window is 24 hours long with a step size of 1 hour) to construct a sample sequence.
[0042] Step S3: Model Training (Offline Phase) Before the equipment is put into use or during periodic offline phases, train an LSTM model using historical normal operating condition data. The specific steps are as follows: T1. Collect at least one year of historical data, covering normal operating conditions under different seasons and production loads, and construct a training dataset.
[0043] T2. Perform the above preprocessing on the training data.
[0044] T3. Construct the LSTM network structure: The number of nodes in the input layer is equal to the feature dimension (e.g., flow rate + pH + turbidity + conductivity, a total of 4 dimensions); the hidden layer uses two LSTM units, each with 64 neurons, and a Dropout layer is added to prevent overfitting; the output layer is a fully connected layer, and the number of nodes is equal to the prediction step size (e.g., the deviation value for predicting the next 1 hour).
[0045] T4. Training the model: Using the historical sequence as input and the true value of the next time step as output, the mean squared error is used as the loss function, and the Adam optimizer is used for training until the model converges.
[0046] T5. Save the model: Solidify the trained model parameters as the baseline model for normal operating conditions.
[0047] Step S4: Online monitoring and deviation calculation During device operation, real-time data is collected and preprocessed to form the input sequence for the current moment. This sequence is then input into a trained LSTM model, which outputs a predicted value (i.e., the expected normal value). The relative deviation between the current actual value and the predicted value is calculated. Deviation = |Actual value - Predicted value| / Predicted value × 100%; Step S5: Early Warning Judgment Set a threshold (e.g., 20%). If the deviation exceeds the threshold, an alert is triggered. Further classify the alert level based on the magnitude of the deviation: 40% ≤ Deviation < 60%: Warning level (orange alert), indicating a high probability of an anomaly, and it is recommended to arrange an inspection; Deviation ≥ 60%: Emergency level (red warning), indicating a serious abnormality, which may indicate a blockage or an impending malfunction, requiring immediate shutdown and handling.
[0048] Step S6: Multi-source data fusion correction If the equipment is equipped with vibration and temperature sensors, vibration amplitude and temperature data can be collected additionally. When flow rate and water quality trigger an alarm, check if vibration and temperature are also abnormal (e.g., vibration amplitude exceeds twice the normal value, temperature rises by more than 5°C). If both are abnormal, raise the alarm level by one level; if only flow rate and water quality are abnormal while vibration and temperature are normal, maintain the original level, but mark it as "requires manual review". Step S7: Early Warning Output and Remote Communication The processor module 201 sends the warning information (including device ID, time, deviation value, warning level, and congestion probability) to the cloud server or directly pushes it to the mobile terminal 8 of the maintenance personnel via a communication unit (such as a 4G module). Simultaneously, the on-site display unit (such as an LCD screen) displays the current data and warning status in real time. Step S8: Self-diagnosis and maintenance Perform self-diagnostic procedures regularly (e.g., daily): send self-test commands to each sensor to check if the response is normal; monitor the sensor power supply voltage and signal range; if a sensor is found to be offline or the data is abnormal, issue a sensor fault alarm to remind maintenance personnel to replace or repair it.
[0049] To facilitate maintenance, the flow monitoring module 4 and water quality detection module 2 can be designed as drawer-type modules, connected to the housing 3 via guide rails and quick-connect connectors. Replacement does not require disassembling the entire device; simply remove the faulty module and insert the spare.
[0050] For situations requiring simultaneous monitoring of multiple pipelines, multiple sets of flow monitoring modules 4 and water quality detection modules 2 can be installed within the same housing 3, each corresponding to different pipelines, sharing the processor module 201 and communication unit to achieve multi-channel parallel monitoring.
[0051] For equipment used in flammable and explosive environments such as chemical plants, all electrical components (sensors, processors, power supplies) are explosion-proof, and the housing is explosion-proof, meeting the requirements of Ex d or Ex e explosion-proof ratings.
[0052] Furthermore, the early warning algorithm can be optimized: Introducing an attention mechanism into the LSTM model enables the model to automatically focus on the time steps and features most important for the current prediction, thereby improving sensitivity to anomalous starting points. Data patterns before and after historical congestion events are stored as an anomaly pattern library. When real-time data shows a high similarity to patterns in the library, an early warning is issued even if the deviation does not reach a threshold. It calculates not only the deviation for individual features, but also the correlation residuals between multiple features (such as the correlation between flow rate and turbidity). When the correlation changes abruptly, it may indicate blockage or a sudden change in water quality, and can be used as an auxiliary criterion.
[0053] The working principle of this invention is as follows: The liquid to be tested enters the flow monitoring module 4 from pipe 6 through inlet pipe 41. After the differential pressure transmitter 404 measures the flow rate, the liquid flows out from outlet pipe 42 and enters the inner cavity of housing 3. As the liquid continues to flow in, the liquid level gradually rises. When the liquid level exceeds the water passage hole 241 of sleeve 208, the liquid slowly enters the detector body 204 through the hole and continues to rise until it submerges all water quality sensors. At this time, water level detection sensor 202 is triggered, indicating that water quality detection module 2 starts working, and each sensor measures water quality parameters. After the measurement is completed, the liquid can be discharged through outlet hole 301 at the bottom of housing 3 (connected to downstream pipe or discharge port), or through overflow port 102. Throughout the process, processor module 201 continuously collects flow rate and water quality data for storage and analysis.
[0054] The working principle of this invention is based on an intelligent closed loop of "sensing-analysis-early warning". First, through an integrated flow monitoring module and water quality detection module, the physical and chemical parameters of the liquid within the pipeline are sensed in real time. Second, an embedded processor is used for data preprocessing and local intelligent analysis. The core of this model is a time-series prediction model based on a long short-term memory neural network. This model learns from historical normal data to obtain a baseline, and then compares the deviation of the current data with the baseline in real time to identify abnormal trends. When the deviation exceeds a set threshold, the system generates graded early warning information based on the degree of deviation and sends it to a remote monitoring platform and maintenance personnel via a wireless communication network. Simultaneously, by combining multi-source data such as vibration and temperature, the early warning is reviewed and corrected to improve accuracy. Furthermore, the system has self-diagnostic capabilities to ensure long-term reliable operation. This design transforms pipeline monitoring from passive post-event handling to proactive predictive maintenance, significantly improving the safety and economy of pipeline operation.
[0055] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.
Claims
1. A device for monitoring pipeline flow and water quality, characterized in that, include: The housing (3) has a cavity inside for containing liquid; The flow monitoring module (4) is located at the bottom of the inner part of the housing (3); The water quality testing module (2) is located inside the upper part of the housing (3); The processor module (201) is electrically connected to the flow monitoring module (4) and the water quality detection module (2). The processor module (201) includes a signal acquisition unit, a data storage unit, a communication unit and an intelligent analysis unit.
2. The pipeline flow and water quality monitoring device according to claim 1, characterized in that, The flow monitoring module (4) includes a differential pressure transmitter and at least one inlet pipe (41); one end of the inlet pipe (41) is connected to an external pipeline, and the other end is connected to the inlet of the differential pressure transmitter; the differential pressure transmitter is used to measure the pressure difference when the liquid flows through and calculate the flow rate accordingly; the flow monitoring module (4) also includes an outlet pipe (42), one end of the outlet pipe (42) is connected to the outlet of the differential pressure transmitter, and the other end extends to the upper part of the inner part of the housing (3) for introducing the liquid that has undergone flow measurement into the detection area of the water quality detection module (2).
3. The pipeline flow and water quality monitoring device according to claim 1, characterized in that, The water quality testing module (2) includes: The detector body (23) has an inner partition (232) and an outer cylinder (234) inside, forming multiple detection chambers; A water level sensor (21) is installed at a predetermined height on the detector body (23) to detect whether the liquid level meets the detection requirements; At least one water quality sensor is disposed in the detection chamber for detecting one or more of the following: pH value, turbidity, conductivity, dissolved oxygen, or concentration of a specific ion.
4. The pipeline flow and water quality monitoring device according to claim 3, characterized in that, The lower end of the detector body (23) is fitted with a sleeve (24), and the side wall of the sleeve (24) is provided with a plurality of water passage holes (241) to allow liquid to enter the detector body (23) evenly and slowly.
5. The pipeline flow and water quality monitoring device according to claim 1, characterized in that, The top of the housing (3) is provided with a cover (1), and the cover (1) is provided with an overflow port (102); the bottom of the housing (3) is provided with a base (5), and the base (5) is provided with a plug (51) and a groove (52) for fixing the flow monitoring module (4); the processor module (201) also includes a vibration sensor and a temperature sensor.
6. A method for monitoring pipeline flow and water quality, applied to the monitoring equipment described in any one of claims 1 to 5, characterized in that, Includes the following steps: S1. The flow rate data of the liquid in the pipeline is collected in real time through the flow monitoring module (4), and the water quality data of the liquid is collected in real time through the water quality detection module (2). S2. Input the collected flow data and water quality data into the intelligent analysis unit of the processor module (201); S3. The intelligent analysis unit calls the pre-trained time series prediction model to analyze the real-time data and calculate its deviation from the normal operating condition baseline. S4. If the deviation exceeds the preset threshold, an early warning message is generated and sent to the monitoring center or mobile terminal through the communication unit.
7. The method for monitoring pipeline flow and water quality according to claim 6, characterized in that, Step S3 is preceded by a model training step: T1. Collect historical flow and water quality data under normal operating conditions to construct a training dataset; T2. Preprocess the training dataset, including data cleaning, normalization, and serialization; T3. Construct a long short-term memory neural network model and train it using the training dataset to obtain a baseline model under normal operating conditions.
8. The method for monitoring pipeline flow and water quality according to claim 7, characterized in that, The preprocessing also includes: dividing different scenarios according to pipeline characteristics, and training a corresponding baseline model for each scenario.
9. The pipeline flow and water quality monitoring device according to claim 6, characterized in that, The generation of early warning information in step S4 includes: calculating the probability of congestion and classifying the early warning level according to the degree of deviation; when the deviation exceeds 20% but is less than 40%, it is a general attention level; when the deviation is between 40% and 60%, it is a warning level; when the deviation exceeds 60%, it is an emergency level. It also includes step S5: reviewing and correcting the early warning information by combining pipeline vibration data and temperature data; if both vibration and temperature data are abnormal, the early warning level is increased. The flow data includes instantaneous flow and cumulative flow; the water quality data includes at least one of pH value, turbidity, conductivity, dissolved oxygen or specific ion concentration.
10. The method for monitoring pipeline flow and water quality according to claim 6, characterized in that, The long short-term memory neural network model includes an input layer, at least one LSTM hidden layer, and an output layer; the number of nodes in the input layer is the same as the number of input features, and the output layer outputs the predicted value or deviation value for the next time step.