Water pool accumulated water monitoring method and system based on overflow port

Abstract

The invention relates to the technical field of pool accumulated water processing, and discloses a pool accumulated water monitoring method and system based on an overflow gap, and the method comprises the steps: comparing a liquid level value digital signal with a threshold value set by a historical liquid level value digital signal, and outputting an overflow monitoring signal if the liquid level value digital signal is higher than the threshold value; receiving an overflow monitoring signal, collecting real-time data of an overflow port pool, constructing an overflow port simulation model based on a historical overflow port pool signal, and inputting the real-time data into the overflow port simulation model for simulation; judging the state of the overflow port pool according to the simulation result, formulating an adjustment scheme according to the state of the overflow port pool, inputting the adjustment scheme into the overflow port simulation model for simulation, and performing iteration in sequence until the data of the overflow port pool is qualified. The system comprises an overflow signal acquisition module, an overflow simulation module and a scheme making module. Dynamic adjustment of the state of the overflow port pool is achieved through the simulation model, and it is ensured that the pool runs within a safe range.

Description

Technical Field

[0001] The present invention relates to the technical field of water pond ponding treatment, and particularly relates to a water pond monitoring method and system based on an overflow outlet. Background Art

[0002] The ponding area of the water pond may cause the ground to be slippery, increasing the risk of falls and slips, especially for children and the elderly who are more likely to be injured. At the same time, it may become a breeding ground for mosquitoes and other pests, increasing the risk of disease transmission, such as dengue fever, malaria, etc. Long-term ponding may lead to water quality deterioration and produce odors, affecting the comfort of the surrounding environment. The ponding may seep into the foundation of the building, causing the foundation to be unstable and affecting the structural safety of the building. The water pond ponding situation needs to be taken seriously. By measures such as cleaning the drainage system, improving the design, strengthening the maintenance, and introducing a rainwater collection system, the water pond ponding problem can be effectively solved to ensure people's life safety and physical health.

[0003] The first prior art, a Chinese patent with the patent number: 202410327282.X, relates to a method for calculating the water level of a mining area reservoir based on a multi-rotor unmanned aerial vehicle, including: setting characteristic points at preset positions of the target mining area reservoir, where the characteristic points include control points and marked water level measurement points; according to the positions of the characteristic points, using the multi-rotor unmanned aerial vehicle to perform orthophoto aerial photography on the target mining area reservoir to obtain aerial images, where the aerial images include infrared aerial images and visible light aerial images; performing enhancement processing on the aerial images and extracting information to obtain water level measurement points; calculating the water level of the target mining area reservoir according to the water level measurement points, control points, and phase points taken by the multi-rotor unmanned aerial vehicle to obtain the water level of the target mining area reservoir. Although it realizes the intelligent inspection of automatic recording of the water level of the mining area reservoir and improves the flexibility of the monitoring range of the reservoir; however, the real-time performance and automation degree are relatively low, resulting in difficulty in timely discovering and handling problems.

[0004] Prior Art II, a Chinese patent with the patent number 202411551119.8 relates to the technical field of flood and waterlogging monitoring, and discloses a urban flood and waterlogging monitoring device with a protection function, including a basic vertical pole. At the upper end of the basic vertical pole, a rain gauge is fixedly installed. On one side of the middle part of the basic vertical pole, an electric control box is fixedly installed. On one side of the electric control box, a box door is rotatably installed. On both sides of the electric control box, a number of first through holes are vertically arranged. In the middle of the interior of the electric control box, a fixed plate is fixedly installed. The interior of the electric control box is divided into a first installation cavity and a second installation cavity by the fixed plate. A protection component is arranged inside the first installation cavity; although, the two first movable plates can be mutually close to each other, so as to reduce the space for the protection component to protect a number of electronic components, and reduce the situation of accumulated water contacting the number of electronic components; and then, the two second movable plates are mutually close to each other, further reducing the space for the protection component to protect a number of electronic components, and avoiding the accumulated water contacting the number of electronic components; however, the technical limitations and environmental adaptability are relatively poor, resulting in unstable performance of the monitoring system in extreme environments.

[0005] Prior Art III, a Chinese patent with the patent number 202411305577.3 discloses a buried dynamic waterlogging monitoring method based on ultrasonic anti-interference technology. First, install equipment at the monitoring point. The equipment consists of a buried sensor and an ultrasonic water level sensor. According to the liquid-medium ultrasonic sensor, by measuring the time interval from the emission to the reception of the sound wave, and combining the compensated sound velocity of the sound wave, the distance of the sound wave transmission and the distance between the ultrasonic probe and the interface are obtained. Although, by using the ultrasonic wave echo to the water surface of the installed equipment, the current liquid level height can be obtained, and a system for real-time monitoring of waterlogging conditions at key locations such as low-lying areas, tunnels, and underpasses in cities, automatically collects the water level data of the waterlogging by installing sensors underground, and transmits the data to a remote monitoring center or a cloud platform through wireless communication technology for processing and analysis, so as to achieve timely early warning and auxiliary decision-making; however, the technical limitations and environmental adaptability are relatively poor, resulting in unstable performance of the monitoring system in extreme environments.

[0006] Currently, Prior Art I, Prior Art II, and Prior Art III have problems such as relatively poor technical limitations and environmental adaptability, as well as relatively low real-time performance and automation degree. To solve the above problems, the present invention provides a pool monitoring method and system based on an overflow port. Summary of the Invention

[0007] The main purpose of the present invention is to provide a pool monitoring method and system based on an overflow port to solve the problems of relatively poor limitations and environmental adaptability, as well as relatively low real-time performance and automation degree in the prior art.

[0008] To achieve the above purpose, the present invention provides the following technical solutions:

[0009] A method for monitoring pool water accumulation based on an overflow outlet. The method for monitoring pool water accumulation based on an overflow outlet includes:

[0010] Real-time monitoring of the water level in the pool through a high-precision liquid level sensor, collecting digital signals of the liquid level value, comparing the digital signals of the liquid level value with the threshold set by the historical digital signals of the liquid level value. If the digital signals of the liquid level value are higher than the threshold, an overflow monitoring signal is output.

[0011] Upon receiving the overflow monitoring signal, collect real-time data of the overflow outlet pool and preprocess the real-time data to obtain a data set. Based on the historical signals of the overflow outlet pool, construct an overflow outlet simulation model, and input the real-time data into the overflow outlet simulation model for simulation.

[0012] Judge the state of the overflow outlet pool according to the simulation results, formulate an adjustment plan according to the state of the overflow outlet pool, and input the adjustment plan into the overflow outlet simulation model for simulation, and iterate in turn until the data of the overflow outlet pool is qualified.

[0013] As a further improvement of the present invention, the process of comparing the digital signals of the liquid level value with the threshold set by the historical digital signals of the liquid level value includes the following steps:

[0014] Connect the high-precision liquid level sensor to the data acquisition system, perform data cleaning, smoothing and normalization preprocessing operations on the collected original data, quantize the preprocessed signal, and then encode the quantized data to convert the preprocessed electrical signal into a digital signal of the liquid level value.

[0015] Input the digital signals of the liquid level value into the intelligent recognition program of the overflow outlet pool, sort based on the time period, perform statistical analysis on the collected historical digital signals of the liquid level value, and calculate statistical quantities such as its mean and standard deviation.

[0016] Set the pressure threshold according to the analysis results of the historical digital signals of the liquid level value, compare the real-time collected liquid level value signal with the set liquid level value threshold. If the collected liquid level value signal is within the range, continue to monitor. If it is higher than the liquid level value threshold, it will be processed.

[0017] As a further improvement of the present invention, the process of inputting the real-time data into the overflow outlet simulation model for simulation includes the following steps:

[0018] Collect signal data such as the liquid level, flow rate and time series of the historical overflow outlet pool; perform preprocessing operations on the signal data to obtain a data set; expand the data set through time series translation, adding random noise and generative adversarial networks.

[0019] Analyze the periodic characteristics and trend terms through time series analysis, convert real-time data into features with preset meaningful values, and calculate the water level value; fuse the monitoring values of multiple sensor data at different positions; divide the dataset into a training set, a validation set, and a test set;

[0020] Construct an overflow outlet simulation model based on historical overflow outlet pool signals, and use the training set to train the overflow outlet simulation model; after training is completed, input the real-time data into the overflow outlet simulation model to generate the overflow outlet simulation result.

[0021] As a further improvement of the present invention, the process of augmenting the dataset through time series translation, adding random noise, and generative adversarial networks includes the following steps:

[0022] Translate the data of the original time series forward and backward along the time axis to generate new data samples; translate the water level data as a whole by several hours to simulate the water level change patterns at different time periods; add random perturbations to the original data to simulate the measurement errors or signal interferences existing in the actual environment;

[0023] Access the random noise input and output the synthesized time series data; distinguish the actual data from the generated data and provide a feedback signal; through iterative optimization until reaching the preset realistic data threshold, add the data reaching the realistic data threshold to the dataset;

[0024] Add the data obtained by time series translation, adding random noise, and reaching the realistic data threshold to the dataset and integrate them to obtain the final dataset.

[0025] As a further improvement of the present invention, the process of fusing the monitoring values of multiple sensor data at different positions includes the following steps:

[0026] Analyze the periodic characteristics and trend terms through time series analysis, convert real-time data into features with preset meaningful values, and calculate the water level value; perform feature complementarity by dynamically adjusting the weights of sensors at different positions; adaptively allocate the contribution degrees of data at each position according to factors such as sensor accuracy and environmental interference;

[0027] Classify the monitoring values of sensors in the same area or with similar functions to form a fusion unit; compare the real-time data with the historical overflow outlet pool data, and at the same time, the sensor data within the same cluster needs to reach the preset spatial correlation. If an anomaly occurs, mark it and trace back to the data source;

[0028] Combine different modal data such as water level and flow velocity, extract the features of each modality through an encoder, and fuse them in the latent space; convert the multivariate time series data into a unified representation, and then generate the fused monitoring values through a decoder.

[0029] As a further improvement of the present invention, the process of converting real-time data into features with preset meaningful values includes the following steps:

[0030] Extract periodic features and trend terms through time series analysis; decompose real-time data into trend, periodic, and residual components; calculate mean and variance statistics within a time window; extract frequency domain features;

[0031] Identify the starting time according to preset parameters; extract statistical features of real-time data, where the statistical features include mean, maximum value, minimum value, and standard deviation, to quantify the volatility and stability of real-time data;

[0032] Combine the preset high pool duration to divide the time range and discard data of non-timekeepers; combine the trend term and slope to predict future water level changes.

[0033] As a further improvement of the present invention, the process of training the overflow outlet simulation model using a training set includes the following steps:

[0034] Divide the training set into multiple small batches, where each small batch has 64 samples; input one batch at a time for forward propagation, and the input data calculates the predicted value through the network; calculate the loss between the predicted value and the true value, and use the backpropagation algorithm to adjust the weights;

[0035] After each cycle, use the validation set to evaluate the model performance, and adjust the learning rate, batch size, and network parameters, etc. according to the validation results; if the validation loss does not improve after reaching the preset number of rounds, terminate the training in advance;

[0036] Use an independent test set to evaluate the mean square error, accuracy, and POC curve of the overflow outlet simulation model; and integrate the trained overflow outlet simulation model into the overflow outlet monitoring system to process sensor data in real time and output simulation results.

[0037] As a further improvement of the present invention, the process of judging the state of the overflow outlet pool according to the simulation results includes the following steps:

[0038] Compare the real-time liquid level, flow rate, and other parameters output by the simulation with the preset safety thresholds; if the liquid level exceeds the first-level threshold, trigger a warning; if it reaches the second-level threshold, it is determined that the overflow risk is high, and different treatment plans are immediately formulated;

[0039] Identify abnormal patterns through statistical calculations and trend analysis; check whether the simulation results are consistent with the historical data. If the deviation between the simulated predicted liquid level and the time monitoring value exceeds the preset deviation threshold, adjust the data of the overflow outlet simulation model and iterate until it is qualified;

[0040] Perform comprehensive scoring by combining multi-dimensional data; generate an alarm by determining whether the multi-dimensional data is within a preset numerical range; if the simulation results simultaneously meet the conditions that the liquid level is lower than the threshold, the flow rate is stable, and the equipment is operating normally, then determine that the status is qualified; otherwise, formulate an adjustment plan and re-conduct simulation verification.

[0041] As a further improvement of the present invention, the process of formulating an adjustment plan and re-conducting simulation verification includes the following steps:

[0042] When the liquid level, flow rate, or equipment parameters exceed the preset range, determine the source of abnormal data through statistical calculation, and use a contribution graph to analyze the contribution degree of each variable to the fault to identify the key adjustment target; if the liquid level deviation is caused by flow rate fluctuations, then preferentially adjust the flow control parameters;

[0043] Backtrack and test according to historical data to dynamically adjust the threshold; according to the liquid level deviation value and the preset reference interval, retrieve the preliminary adjustment value from the overflow water pool database; if the deviation still exceeds the preset threshold after adjustment, then trigger a stuck alarm;

[0044] Re-conduct the simulation with the modified parameters, monitor the residual convergence, time step, and number of iterations; form an adjustment plan and verify the multi-adjustment plan through a multi-dimensional scoring mechanism; input the modified plan and the modified overflow water pool data into the database for storage.

[0045] To achieve the above object, the present invention also provides the following technical solution:

[0046] A water pool water accumulation monitoring system based on an overflow outlet, which is applied to the water pool water accumulation monitoring method based on the overflow outlet. The water pool water accumulation monitoring system based on the overflow outlet includes:

[0047] An overflow signal acquisition module, which is used to monitor the water level of the water pool in real time through a high-precision liquid level sensor, acquire the digital signal of the liquid level value, compare the digital signal of the liquid level value with the threshold set by the historical digital signal of the liquid level value, and if the digital signal of the liquid level value is higher than the threshold, then output an overflow monitoring signal;

[0048] An overflow simulation module, which is used to receive the overflow monitoring signal, acquire the real-time data of the overflow water pool, and perform preprocessing on the real-time data to obtain a data set; construct an overflow outlet simulation model based on the historical overflow water pool signal, and input the real-time data into the overflow outlet simulation model for simulation;

[0049] A plan formulation module, which is used to judge the status of the overflow water pool according to the simulation results, formulate an adjustment plan according to the status of the overflow water pool, and input the adjustment plan into the overflow outlet simulation model for simulation, and iterate in turn until the overflow water pool data is qualified.

[0050] The present invention collects digital signals of the water level in the water tank in real time through a high-precision liquid level sensor to ensure the accuracy and real-time nature of the data; compares the collected digital signals of the liquid level value with the threshold values set for the historical liquid level value digital signals to determine whether the set overflow threshold is exceeded; when the liquid level value is higher than the threshold, the system outputs an overflow monitoring signal to trigger subsequent processing procedures; realizes real-time monitoring of the water level through a high-precision sensor to ensure timely detection of abnormal situations; improves the accuracy of liquid level monitoring by comparing with historical data to avoid false alarms or missed alarms; when the liquid level exceeds the threshold, the system can promptly issue an overflow monitoring signal to provide a basis for subsequent processing. After receiving the overflow monitoring signal, real-time data of the overflow water tank is collected, and the data is preprocessed to generate a data set; a simulation model is constructed based on historical signals of the overflow water tank, and the real-time data is input into the model for simulation; the plan is adjusted according to the simulation results, and the adjusted plan is input into the model for simulation, iterating in turn until the qualified standard is met; through data preprocessing and feature extraction, the data quality of the input model is ensured, and the reliability of the simulation results is improved; through iterative simulation, the adjustment plan is continuously optimized to ensure that the final state of the overflow water tank meets the requirements; the simulation model is used to realize dynamic prediction and adjustment of the state of the overflow water tank, improving management efficiency. Judge the state of the overflow water tank according to the simulation results, and analyze whether there is an overflow risk or areas that need to be adjusted; formulate an adjustment plan according to the state of the overflow water tank, and input the plan into the simulation model for verification; verify the effectiveness of the adjustment plan through the simulation model, and continue to optimize the plan if necessary until the qualified standard is reached; realize dynamic adjustment of the state of the overflow water tank through the simulation model to ensure that the water tank operates within a safe range; reduce manual intervention and improve management efficiency through iterative optimization of the adjustment plan; effectively avoid overflow risks and ensure the safety of the water tank operation through scientific adjustment plans and simulation verification. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a schematic flow chart of the steps of an embodiment of the water tank monitoring method based on the overflow outlet of the present invention;

[0052] Figure 2 It is a schematic flow chart of the steps of comparing the digital signal of the liquid level value with the threshold value set for the historical digital signal of the liquid level value in an embodiment of the water tank monitoring method based on the overflow outlet of the present invention;

[0053] Figure 3 It is a schematic flow chart of the steps of inputting real-time data into the overflow outlet simulation model for simulation in an embodiment of the water tank monitoring method based on the overflow outlet of the present invention;

[0054] Figure 4 It is a schematic flow chart of the steps of expanding the data set by time series translation, adding random noise, and generative adversarial network in an embodiment of the water tank monitoring method based on the overflow outlet of the present invention;

[0055] Figure 5 Schematic diagram of the step process for fusing the monitoring values of multiple sensor data at different positions in an embodiment of the pool monitoring method based on the overflow outlet of the present invention;

[0056] Figure 6 Schematic diagram of the step process for converting real-time data into features with preset meaningful values in an embodiment of the pool monitoring method based on the overflow outlet of the present invention;

[0057] Figure 7 Schematic diagram of the step process for training the overflow outlet simulation model using the training set in an embodiment of the pool monitoring method based on the overflow outlet of the present invention;

[0058] Figure 8 Schematic diagram of the step process for judging the state of the overflow outlet pool according to the simulation results in an embodiment of the pool monitoring method based on the overflow outlet of the present invention;

[0059] Figure 9 Schematic diagram of the step process for formulating an adjustment plan and re-performing simulation verification in an embodiment of the pool monitoring method based on the overflow outlet of the present invention;

[0060] Figure 10 Schematic diagram of the functional modules in an embodiment of the pool monitoring system based on the overflow outlet of the present invention;

[0061] Figure 11 Schematic diagram of the structure of an embodiment of the electronic device of the present invention;

[0062] Figure 12 Schematic diagram of the structure of an embodiment of the storage medium of the present invention. Detailed implementation manners

[0063] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0064] The terms "first", "second", and "third" in the present invention are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first", "second", and "third" may explicitly or implicitly include at least one such feature. In the description of the present invention, the meaning of "a plurality" is at least two, such as two, three, etc., unless otherwise specifically defined. In the embodiments of the present invention, all directional indications (such as up, down, left, right, front, back...) are only used to explain the relative positional relationship and movement conditions between components in a specific posture (as shown in the drawings). If the specific posture changes, the directional indications will also change accordingly. In addition, the terms "comprise" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally further include unlisted steps or units, or may optionally further include other steps or units inherent to these processes, methods, products, or devices.

[0065] Reference to "embodiment" in this context means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the present invention. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0066] As Figure 1 shown, this embodiment provides an embodiment of a pool monitoring method based on an overflow outlet. In this embodiment, the pool monitoring method based on the overflow outlet specifically includes the following steps:

[0067] Step S1: Real-time monitor the water level of the pool through a high-precision liquid level sensor, collect the digital signal of the liquid level value, compare the digital signal of the liquid level value with the threshold set by the historical digital signal of the liquid level value. If the digital signal of the liquid level value is higher than the threshold, output an overflow monitoring signal;

[0068] Step S2: Upon receiving the overflow monitoring signal, collect the real-time data of the pool at the overflow outlet and preprocess the real-time data to obtain a data set; construct an overflow outlet simulation model based on the historical pool signal at the overflow outlet, and input the real-time data into the overflow outlet simulation model for simulation;

[0069] Step S3: Judge the state of the pool at the overflow outlet according to the simulation result, formulate an adjustment plan according to the state of the pool at the overflow outlet, and input the adjustment plan into the overflow outlet simulation model for simulation, and iterate in turn until the data of the pool at the overflow outlet is qualified.

[0070] Preferably, in step S1 of this embodiment, the digital signal of the water level in the water tank is collected in real time by a high-precision liquid level sensor to ensure the accuracy and real-time nature of the data; the digital signal of the collected liquid level value is compared with the threshold value set by the historical liquid level value digital signal to determine whether the set overflow threshold is exceeded; when the liquid level value is higher than the threshold value, the system outputs an overflow monitoring signal to trigger the subsequent processing flow; the real-time monitoring of the water level is realized through a high-precision sensor to ensure the timely discovery of abnormal situations; the accuracy of liquid level monitoring is improved by comparing with historical data to avoid false alarms or missed reports; when the liquid level exceeds the threshold value, the system can promptly send out an overflow monitoring signal to provide a basis for subsequent processing. After receiving the overflow monitoring signal in step S2, the real-time data of the overflow water tank is collected, and the data is preprocessed to generate a data set; a simulation model is constructed based on the historical signals of the overflow water tank, and the real-time data is input into the model for simulation; the scheme is adjusted according to the simulation results, and the adjusted scheme is input into the model for simulation, and the iteration is carried out in turn until the qualified standard is met; the data quality of the input model is ensured through data preprocessing and feature extraction, and the reliability of the simulation results is improved; the adjusted scheme is continuously optimized through iterative simulation to ensure that the final state of the overflow water tank meets the requirements; the dynamic prediction and adjustment of the state of the overflow water tank are realized by using the simulation model to improve the management efficiency. In step S3, the state of the overflow water tank is judged according to the simulation results, and it is analyzed whether there is an overflow risk or a place that needs to be adjusted; an adjustment scheme is formulated according to the state of the overflow water tank, and the scheme is input into the simulation model for verification; the effectiveness of the adjustment scheme is verified through the simulation model, and the scheme is continuously optimized if necessary until the qualified standard is reached; the dynamic adjustment of the state of the overflow water tank is realized through the simulation model to ensure that the water tank operates within a safe range; the manual intervention is reduced through iterative optimization of the adjustment scheme to improve the management efficiency; through a scientific adjustment scheme and simulation verification, the overflow risk is effectively avoided, and the safety of the water tank operation is guaranteed.

[0071] Further, as Figure 2 shown, the process of comparing the digital signal of the liquid level value with the threshold value set by the historical liquid level value digital signal in step S1 specifically includes the following steps:

[0072] Step S11: Connect the high-precision liquid level sensor to the data acquisition system, perform preprocessing operations such as data cleaning, smoothing, and normalization on the collected original data, quantize the preprocessed signal, and then encode the quantized data to convert the preprocessed electrical signal into a digital signal of the liquid level value;

[0073] Step S12: Input the digital signal of the liquid level value into the intelligent recognition program of the overflow water tank, sort it based on the time period, perform statistical analysis on the collected historical liquid level value digital signals, and calculate statistical quantities such as its mean value and standard deviation;

[0074] Step S13: Set the pressure threshold according to the analysis result of the historical liquid level value digital signal, compare the real-time collected liquid level value signal with the set liquid level value threshold. If the collected liquid level value signal is within the range, continue to monitor. If it is higher than the liquid level value threshold, process it.

[0075] Preferably, in step S11 of this embodiment, the liquid level data is collected by a high-precision liquid level sensor, and the original data is preprocessed, such as cleaning, smoothing, and normalizing, by combining with a data acquisition system. Then, the signal is quantized and encoded, and finally the electrical signal is converted into a liquid level value digital signal. The use of high-precision sensors and data preprocessing technology ensures the accuracy and stability of the data, improves the reliability and availability of the liquid level data, and provides high-quality basic data for subsequent analysis. In step S12, the liquid level value digital signal is input into an intelligent recognition program, the historical liquid level data is sorted based on time periods, and statistical quantities such as the mean and standard deviation are calculated. The law of liquid level change is revealed through statistical analysis, providing a basis for real-time monitoring and early warning. In step S13, the pressure threshold is set according to the analysis result of the historical liquid level data, and the real-time collected liquid level value signal is compared with the threshold. If the liquid level value is within the range, continue to monitor. If it is higher than the threshold, process it. Combining real-time monitoring and threshold judgment mechanisms, dynamic monitoring and early warning are realized, abnormal liquid level conditions are detected in a timely manner, and the safe operation of the system is ensured.

[0076] Further, as Figure 3 shown, the process of inputting real-time data into the overflow port simulation model for simulation in step S2 specifically includes the following steps:

[0077] Step S21: Collect signal data such as the liquid level, flow rate, and time series of the historical overflow port water pool; perform preprocessing operations on the signal data, such as removing duplicate values, filling missing values, and removing abnormal values, to obtain a data set; expand the data set through time series translation, adding random noise, and generative adversarial networks.

[0078] Step S22: Analyze the periodic characteristics and trend terms through time series, convert the real-time data into features with preset meaningful values, and calculate the water level value; fuse the monitoring values at different positions of multiple sensor data; divide the data set into a training set, a validation set, and a test set.

[0079] Step S23: Build an overflow port simulation model based on the historical overflow port water pool signals, and use the training set to train the overflow port simulation model; after training is completed, input the real-time data into the overflow port simulation model to generate an overflow port simulation result.

[0080] Preferably, in step S21 of this embodiment, signal data such as the liquid level and flow rate of the historical overflow water pool are collected and preprocessed (such as de-duplication, filling missing values, and removing outliers). At the same time, the data set is expanded by time series translation, adding random noise, and generative adversarial networks; improving the quality and diversity of the data, providing reliable basic data support for subsequent modeling. In step S22, periodic features and trend terms are extracted through time series analysis, real-time data is converted into features with preset meanings, and the water level value is calculated. In addition, the monitoring values at different positions are integrated through multi-sensor data fusion technology, and the data set is divided into a training set, a validation set, and a test set; improving the input quality and generalization ability of the model, ensuring the applicability of the model in different scenarios. In step S23, an overflow port simulation model is constructed based on historical signals and trained using the training set. After training, real-time data is input into the model to generate simulation results; realizing the dynamic prediction of the operation state of the overflow port and improving the intelligent level of the overflow port management.

[0081] Further, as Figure 4 shown, the process of expanding the data set by time series translation, adding random noise, and generative adversarial networks in step S21 specifically includes the following steps:

[0082] Step S211: Translate the data of the original time series forward and backward along the time axis to generate new data samples; translate the water level data as a whole by several hours to simulate the water level change patterns at different times; add random perturbations to the original data to simulate measurement errors or signal interferences existing in the actual environment;

[0083] Step S212: Access the random noise input and output the synthesized time series data; distinguish the actual data from the generated data and provide a feedback signal; through iterative optimization until the preset realistic data threshold is reached, and add the data reaching the realistic data threshold to the data set;

[0084] Step S213: Add the data obtained by time series translation, adding random noise, and reaching the realistic data threshold to the data set and integrate them to obtain the final data set.

[0085] Preferably, in step S211 of this embodiment, new data samples are generated by performing time-axis translation and random perturbation on the original time series data to simulate the water level change patterns and measurement errors in different time periods; the diversity and complexity of the data are enhanced, and the time offset and noise interference in the actual environment are simulated to provide richer data support for subsequent model training. In step S212, through random noise input and iterative optimization, the actual data and the generated data are distinguished, and the fidelity of the data is gradually improved until a preset threshold is reached; through the feedback signal and iterative optimization, the authenticity and reliability of the data are improved to ensure that the generated data can effectively reflect the actual scenario, thereby enhancing the generalization ability of the model. In step S213, the data that has been translated, added with noise, and verified for fidelity is integrated into the data set to form the final training data set; by integrating diverse data samples, a high-quality data set is constructed to provide a comprehensive training basis for the deep learning model and enhance the performance and prediction ability of the model.

[0086] Further, as Figure 5 shown, the process of fusing the monitoring values of multiple sensor data at different positions in step S22 specifically includes the following steps:

[0087] Step S221: Convert the real-time data into features with preset meaningful values by analyzing the periodic features and trend terms of the time series, and calculate the water level value; perform feature complementarity by dynamically adjusting the weights of sensors at different positions; adaptively allocate the contribution degrees of data at each position according to factors such as sensor accuracy and environmental interference;

[0088] Step S222: Classify the monitoring values of sensors in the same area or with similar functions to form a fusion unit; compare the real-time data with the historical data of the overflow water pool, and at the same time, the sensor data within the same cluster needs to reach the preset spatial correlation. If an anomaly occurs, it is marked and the data source is traced back;

[0089] Step S223: Combine different modal data such as water level and flow velocity, extract the features of each modality through an encoder, and then fuse them in the latent space; convert the multivariate time series data into a unified representation, and then generate the fused monitoring value through a decoder.

[0090] Among them, step S221 represents the multi-sensor dynamic weight fusion formula:

[0091]

[0092] In the formula, W i (t) represents the fusion weight of the i-th sensor at time t; represents the signal energy density of sensor i (defined as the integral of the absolute value of the second derivative of the liquid level change rate within a sliding window); denotes the signal energy density of sensor j (defined as the integral of the absolute value of the second derivative of the liquid level change rate within a sliding window); denotes the phase offset of sensor i (based on the statistical variance of the zero-crossing distribution of the time-domain waveform); denotes the phase offset of sensor j (based on the statistical variance of the zero-crossing distribution of the time-domain waveform) denotes the weighted median of the phase offsets of all sensors; ζ, η denote the environmental coupling coefficients (dynamically generated from the normalized entropy values of the temperature gradient and mechanical vibration spectrum); N denotes the number of sensors; a dummy variable for the summation operation, k ∈ [1, N] traverses all sensors; integrating the vorticity conservation principle in fluid mechanics, mapping the sensor accuracy to the energy density weight; introducing the concept of electromagnetic field phase synchronization, and correcting the influence of environmental interference through the phase offset.

[0093] Step S222 represents the spatial correlation clustering and anomaly backtracking formula:

[0094]

[0095] In the formula, C m denotes the data set of the m-th fusion unit; denotes the mean vector of the m-th cluster in the historical data; x i denotes the observation vector of the i-th sensor in the state space; x j denotes the observation vector of the j-th sensor in the state space; denotes the standard deviation matrix of the historical cluster; denotes the spatial correlation attenuation function between sensors i and j (based on the spectral radius of the signal covariance matrix); λ denotes the time attenuation factor (inversely proportional to the sensor sampling frequency); Γ denotes the preset correlation tolerance threshold (dynamically adjusted by the maximum deviation of historical abnormal events); t0 denotes the initial calibration time; integrating the principle of entropy increase in thermodynamics, quantifying the irreversibility of data drift through the time integral term; introducing the concept of manifold distance in geometric topology, and dynamically correcting the cluster boundary.

[0096] Step S223 represents the multi-modal hidden space fusion formula:

[0097]

[0098] In the formula, F fused denotes the fused multi-modal monitoring value; τ k denotes the high-order tensor transformation operator of the k-th mode (mapping the time-series data to the frequency-domain - spatial-domain joint basis function); denotes the hidden space projection of the i-th sensor in the k-th mode (generated through a non-linear kernel function); denotes the tensor contraction operation (based on the Einstein summation convention); ωk represents the modal weight (jointly calculated from the mutual information entropy between modes and the energy ratio); D represents the decoding operator (mapping the latent space tensor to the physical dimension output); K represents the number of independent information channels in multimodal fusion; M represents the total number of clusters formed by spatial correlation clustering; m represents the cluster index variable; drawing on the concept of entangled states in quantum mechanics, cross-modal information entanglement is achieved through tensor contraction; introducing the principle of optical interference, the feature resolution is enhanced by the phase superposition of the projection operator.

[0099] Preferably, in step S221 of this embodiment, the periodic features and trend terms are extracted through time series analysis, the real-time data is converted into features with preset meanings, and the weights are dynamically adjusted in combination with factors such as sensor accuracy and environmental interference to achieve feature complementarity and adaptive allocation of data contribution degrees; it can accurately extract key features such as water level, improve the accuracy and reliability of monitoring data, and at the same time optimize the utilization efficiency of sensor resources. In step S222, the monitoring values of sensors in the same area or with similar functions are classified into fusion units, compared with historical data in combination with real-time data, and abnormal data is detected through spatial correlation, marked and traced back to the data source; through classification and correlation analysis, the data integration ability and abnormal detection ability are improved, ensuring the stability and reliability of the monitoring system. In step S223, in combination with multimodal data such as water level and flow rate, the features of each mode are extracted through an encoder, fused in the latent space, and a unified representation of the monitoring value is generated through a decoder; realizing the unified representation and efficient fusion of multivariate time series data, and improving the comprehensive analysis ability and prediction accuracy of the monitoring system.

[0100] Furthermore, as Figure 6 shown, the process of converting real-time data into features with preset meaningful values in step S221 specifically includes the following steps:

[0101] Step S2211: Extract periodic features and trend terms through time series analysis; decompose the real-time data into trend, periodic, and residual components; calculate statistical quantities such as mean and variance within the time window; extract frequency domain features;

[0102] Step S2212: Identify the starting point of time according to preset parameters; extract the statistical features of the real-time data; where the statistical features include mean, maximum value, minimum value, and standard deviation, etc., quantifying the volatility and stability of the real-time data;

[0103] Step S2213: Combine the preset high pool duration, divide the time range of time, and discard the data of the timer; combine the trend term and slope to predict the future water level change.

[0104] Preferably, in step S2211 of this embodiment, periodic features and trend terms are extracted through time series analysis, and real-time data is decomposed into trend, periodic, and residual components; statistics such as mean and variance within a time window are calculated; frequency domain features are extracted; key features can be extracted from complex time series data, including trend, periodic, and random components, and at the same time, the periodic change law of the data is revealed through frequency domain analysis, providing a basis for subsequent prediction. In step S2212, the starting time is identified according to preset parameters, and statistical features of real-time data (such as mean, maximum value, minimum value, and standard deviation) are extracted to quantify the volatility and stability of the data; through the extraction of statistical features, the dynamic characteristics of real-time data can be quickly evaluated, providing a basis for trend prediction and anomaly detection, and at the same time helping to identify the stability and fluctuation range of the data. In step S2213, the time range is divided in combination with the preset high pool duration, and data of the waste timer is discarded; the future water level change is predicted by combining the trend term and the slope; through the combination of the trend term and the slope, the future water level change trend can be effectively predicted, and at the same time, through the division of the time range and data discarding, the calculation efficiency is optimized, and the accuracy and real-time performance of the prediction are improved.

[0105] Further, the process of extracting the statistical features of the real-time data in step S2212 specifically includes the following steps:

[0106] Step S22121: Establish a dynamic association rule between the environmental coupling factor (such as temperature gradient, mechanical vibration spectrum) in the preset parameters and the transient energy density of the real-time data stream (the integral of the absolute value of the second derivative of the liquid level change per unit time); when the gradient mutation consistency index of the signal of the real-time data (the continuous matching degree of the liquid level change direction in adjacent time windows) and the preset energy trigger threshold meet the following conditions, the determination of the starting time is triggered;

[0107] Energy density accumulation window: within the sliding time kernel, if the accumulated value of the acceleration of the liquid level change exceeds the dynamic threshold corrected by the environmental coupling factor, it is determined as the starting point of an effective event;

[0108] Phase synchronization verification: By comparing the signal phase offset of sensors at different positions at the same moment (based on the zero-crossing distribution of the time-domain waveform), false triggers caused by local interference are excluded;

[0109] Step S22122: Within the calibrated time interval, a two-way feature extraction channel is constructed, and the mean and extreme values of the statistics are reconstructed into a three-dimensional feature vector;

[0110] Forward channel: Map the original liquid level sequence to the process stability space, and calculate the inertia feature (the deviation degree of the mean change rate from the historical trend) and the anti-interference degree (the inverse relationship between the standard deviation and the environmental interference intensity) of the data within the window;

[0111] Reverse channel: By means of frequency-domain and time-domain hybrid sampling, oscillation mode features (proportion of high-frequency disturbance energy) and trend continuity index (residual ratio of the current window mean value to the predicted value of the previous window) are extracted;

[0112] Statistics (mean value, extreme value) are reconstructed into three-dimensional feature vectors. Energy dimension: integral of the fluctuating energy normalized based on the signal power spectral density; Time dimension: KL divergence between the data point distribution density and the preset ideal distribution; Spatial dimension: spatial correlation attenuation coefficient of multi-sensor data within a sliding window;

[0113] Step S22123: For liquid level oscillations that continuously exceed the threshold, a wide time window is used to calculate the stability parameter, and a memory decay coefficient of the historical stable period parameter is introduced to suppress overshoot; for high-frequency and low-amplitude noise, invalid components are dynamically stripped through time-domain band-pass filtering and spatial consistency verification.

[0114] Preferably, this embodiment breaks through the traditional fixed threshold or single statistic mode, and establishes a new analysis framework of "environmental coupling trigger - multi-dimensional feature space - dual-channel stability"; transforms the energy conservation principle of the fluid system into a dynamic threshold generation rule in the field of signal processing, realizing cross-domain technology migration; solves the inherent problem of mutual interference between high-frequency noise and low-frequency trend in traditional methods through the frequency-domain and time-domain hybrid sampling mechanism of the reverse channel; forms a technical solution through an innovative analysis method of the underlying signal dynamics characteristics.

[0115] Furthermore, as Figure 7 shown, the process of training the overflow port simulation model using the training set in step S23 specifically includes the following steps:

[0116] Step S231: Divide the training set into multiple small batches, where each small batch has 64 samples; input one batch each time for forward propagation, and calculate the predicted value of the input data through the network; calculate the loss between the predicted value and the true value, and use the backpropagation algorithm to adjust the weights;

[0117] Step S232: After each epoch, use the validation set to evaluate the model performance, and adjust the learning rate, batch size, and network parameters, etc. according to the validation results; if the validation loss does not improve after reaching the preset number of rounds, terminate the training in advance;

[0118] Step S233: Use an independent test set to evaluate the mean square error, accuracy, and POC curve of the overflow port simulation model; and integrate the trained overflow port simulation model into the overflow port monitoring system to process sensor data in real time and output simulation results.

[0119] Among them, the dynamic loss function and weight update formula in step S231:

[0120]

[0121] where, L (b) represents the dynamic loss value of the b-th batch; α represents the amplitude penalty factor (positively correlated with the cumulative gradient of the prediction error); β represents the non-linear exponent (adaptively adjusted by the historical loss curvature); γ represents the temporal consistency weight (dynamically generated according to the derivative correlation between the predicted value and the true value); h(·) represents the saturation function used to constrain the product of gradients; y i′ represents the true liquid level value of the i'-th sampling point; represents the predicted output value of the i'-th sampling point; The Lyapunov stability condition in the fusion control theory enhances the temporal consistency through derivative correlation; The fractal geometry idea is introduced to capture the multi-scale characteristics of errors through the non-linear exponent.

[0122] Adaptive learning rate adjustment formula in step S232:

[0123]

[0124] where, η e represents the learning rate of the e-th cycle; represents the gradient of the validation loss of the k'-th cycle; ∈ represents the smoothing factor (a very small constant to prevent the denominator from being zero); η0 represents the initial learning rate (inversely proportional to the sample entropy value of the training set); Based on the second law of thermodynamics, the irreversible dissipation in the training process is quantified through the gradient entropy value; The cosmological inflation model is introduced to simulate the "cooling" process of the learning rate through exponential decay;

[0125] Multi-dimensional performance evaluation formula in step S233

[0126]

[0127] where, M 评估 represents the comprehensive multi-dimensional performance index; K' represents the temporal synchronization gain coefficient (generated by the derivative correlation between the prediction and the true signal); S 预测 , S 真实 represents the phase space trajectory of the prediction and the true signal (constructed through the Hilbert transform); ⊙ represents the tensor Hadamard product, accumulating element-wise multiplications; Λ i″ represents the energy density weight of the i''-th time slice (positively correlated with the kurtosis of the signal power spectrum; Δ represents the non-linear error amplification exponent (dynamically adjusted by the statistical quantile of the historical maximum error; Φ represents the contribution degree of the hidden space features (normalized value based on the mutual information entropy between modes); Ψ j′ represents the gradient sensitivity factor of the j'-th latent variable (related to the condition number of the Jacobian matrix of the decoder output; represents the spectral norm of the gradient covariance matrix of the j'-th hidden space feature; Z represents the environmental noise suppression factor (generated by weighting the Fourier coefficients of the sensor signal-to-noise ratio); P动态 represents the dynamic stability parameter (L2 norm of the second derivative of the trend term); Q 稳态 represents the steady-state consistency parameter (JS divergence between the data distribution within the window and the ideal model); represents the Kronecker product (used to couple dynamic and steady-state characteristics); Ω represents the multi-physics coupling index (calculated from the eigenvalues of the covariance matrix of multi-source data such as temperature and vibration); τ1, τ2 represent the boundaries of the time window; T represents the total number of sampling points within the evaluation window; the prediction residual at the i″-th time slice; represents the inherent error of the reference sensor system at the i″-th time slice. By introducing the path integral idea in quantum field theory and performing integral operations on the phase space trajectory, the non-linear synchronization characteristics of the prediction and the real signal in the time-frequency domain are captured, replacing the static calculation of the traditional mean square error; the generation of the dynamic gain coefficient draws on the principle of optical coherence, and the correlation of the signal derivative is used to simulate the phase matching process of light wave interference. The denominator part of the formula incorporates the particle collision energy model in high-energy physics, combines the energy density of the prediction error with the historical statistical characteristics, and realizes the scale-invariant evaluation of the error through a non-linear exponent. The design of the hidden space gradient sensitivity factor is inspired by the biological neural synapse plasticity mechanism, and the contribution intensity of the feature to the model output is quantified through gradient covariance. The construction of the environmental noise suppression factor draws on the active acoustic noise reduction technology, and frequency-domain selective filtering is achieved by weighting the Fourier coefficients; the generation of the coupling index introduces the plate stress coupling model in geophysics, and the system complexity is mapped through the eigenvalues of the multi-source data covariance. The Kronecker product operation of the dynamic parameter and the steady-state parameter simulates the rate balance equation in chemical reaction kinetics, forcing the model to achieve an optimal trade-off between fast response and long-term stability.

[0128] Preferably, in step S231 of this embodiment, the training set is divided into small batches (64 samples per batch), the predicted values are calculated through forward propagation, and the weights are adjusted using the backpropagation algorithm; small batch training and the backpropagation algorithm are adopted, and the predicted values are calculated by combining forward propagation; by processing the data in batches, the memory occupancy is reduced and the training efficiency is improved; the backpropagation algorithm can efficiently update the weights and optimize the model performance. In step S232, the model performance is evaluated using the validation set after each epoch, and hyperparameters such as the learning rate and batch size are adjusted according to the validation results; if the validation loss does not improve, the training is terminated early; the hyperparameters are dynamically adjusted, and the model performance is monitored through the validation set to avoid overfitting and save computing resources at the same time. In step S233, the mean square error, accuracy, and POC curve of the model are evaluated using an independent test set, and the model is integrated into the monitoring system to process sensor data in real time and output simulation results; the independent test set evaluates the generalization ability of the model and combines the real-time data processing ability; it ensures that the model performs well on unseen data and realizes fast response and high-precision simulation in practical applications.

[0129] Furthermore, as Figure 8 shown, the process of judging the state of the overflow water pool according to the simulation results in step S3 specifically includes the following steps:

[0130] Step S31: Compare the real-time liquid level, flow rate and other parameters output by the simulation with the preset safety thresholds; if the liquid level exceeds the first-level threshold, trigger an early warning; if the second-level threshold is reached, it is determined that the overflow risk is high, and different treatment plans are immediately formulated;

[0131] Step S32: Identify abnormal patterns through statistical calculation and trend analysis; check whether the simulation results are consistent with the historical data. If the deviation between the simulated predicted liquid level and the time monitoring value exceeds the preset deviation threshold, adjust the data of the overflow outlet simulation model and iterate until qualified;

[0132] Step S33: Conduct a comprehensive score by combining multi-dimensional data; generate an alarm by judging whether the multi-dimensional data is within the preset numerical range; if the simulation results simultaneously meet the conditions that the liquid level is lower than the threshold, the flow rate is stable, and the equipment is operating normally, it is determined that the state is qualified; otherwise, formulate an adjustment plan and conduct a simulation verification again.

[0133] Preferably, in step S31 of this embodiment, the real-time liquid level, flow rate and other parameters output by the simulation are compared with the preset safety thresholds. If the liquid level exceeds the first-level threshold, an early warning is triggered. If the second-level threshold is reached, it is determined that the overflow risk is high and a treatment plan is formulated; through real-time monitoring and threshold comparison, abnormal liquid levels can be detected in a timely manner and early warning or emergency measures can be taken to reduce the overflow risk and ensure the safe operation of the system. In step S32, abnormal patterns are identified through statistical calculation and trend analysis, and whether the simulation results are consistent with the historical data is checked. If the deviation exceeds the preset threshold, the data of the overflow outlet simulation model is adjusted and iterated until qualified; by using statistical analysis and iterative optimization, the accuracy and reliability of the simulation model are ensured, the prediction accuracy is improved, and a scientific basis is provided for subsequent decision-making. In step S33, a comprehensive score is conducted by combining multi-dimensional data, and it is judged whether the multi-dimensional data is within the preset numerical range to generate an alarm; if the simulation results meet the conditions that the liquid level is lower than the threshold, the flow rate is stable, and the equipment is operating normally, it is determined that the state is qualified; otherwise, an adjustment plan is formulated and a simulation verification is conducted again; through multi-dimensional data analysis and scoring mechanism, the system state is comprehensively evaluated to ensure the stability of the liquid level, flow rate and equipment operation, and the overall safety and reliability of the system are improved.

[0134] Furthermore, as Figure 9 shown, the process of formulating an adjustment plan and conducting a simulation verification again in step S33 specifically includes the following steps:

[0135] Step S331: When the liquid level, flow rate, or equipment parameters exceed the preset range, determine the source of abnormal data through statistical calculation, analyze the contribution degree of each variable to the fault using a contribution diagram, and identify the key adjustment targets; if the liquid level deviation is caused by flow rate fluctuations, then preferentially adjust the flow rate control parameters.

[0136] Step S332: Conduct a backtest based on historical data to dynamically adjust the threshold; according to the liquid level deviation value and the preset reference interval, retrieve the preliminary adjustment value from the overflow water pool database; if the deviation still exceeds the preset threshold after adjustment, then trigger a jamming alarm.

[0137] Step S333: Re - simulate the modified parameters, monitor the residual convergence, time step, and number of iterations; form an adjustment plan, and verify multiple adjustment plans through a multi - dimensional scoring mechanism; input the modified plan and the modified overflow water pool data into the database for storage.

[0138] Preferably, in step S331 of this implementation, determine the source of abnormal data through statistical calculation, analyze the contribution degree of each variable to the fault using a contribution diagram, and identify the key adjustment targets; if the liquid level deviation is caused by flow rate fluctuations, then preferentially adjust the flow rate control parameters; this can accurately locate the abnormal cause and preferentially handle key problems, improving the fault handling efficiency and system stability. In step S332, conduct a backtest based on historical data to dynamically adjust the threshold; according to the liquid level deviation value and the preset reference interval, retrieve the preliminary adjustment value from the overflow water pool database; if the deviation still exceeds the preset threshold after adjustment, then trigger a jamming alarm; through dynamically adjusting the threshold and validating with historical data, ensure the scientificity and accuracy of the adjustment plan, and at the same time, respond to abnormal situations in a timely manner through the alarm mechanism. In step S333, re - simulate the modified parameters, monitor the residual convergence, time step, and number of iterations; form an adjustment plan, and verify multiple adjustment plans through a multi - dimensional scoring mechanism; input the modified plan and the modified overflow water pool data into the database for storage; through simulation verification and the multi - plan scoring mechanism, ensure the optimality and reliability of the adjustment plan, and at the same time, realize the real - time update and storage of data, improving the intelligent level of the system.

[0139] As Figure 10 shown, this embodiment also provides an embodiment of the water pool water accumulation monitoring system based on the overflow outlet. In this embodiment, the water pool water accumulation monitoring system based on the overflow outlet is applied to the water pool water accumulation monitoring method based on the overflow outlet in the above - mentioned embodiment. This water pool water accumulation monitoring system based on the overflow outlet includes:

[0140] An overflow signal acquisition module 1, which is used to continuously monitor the water level of the water pool through a high - precision liquid level sensor, acquire the digital signal of the liquid level value, compare the digital signal of the liquid level value with the threshold set by the historical liquid level value digital signal, and if the digital signal of the liquid level value is higher than the threshold, then output an overflow monitoring signal.

[0141] An overflow simulation module 2, configured to receive an overflow monitoring signal, collect real-time data of the overflow outlet pool, preprocess the real-time data to obtain a data set; construct an overflow outlet simulation model based on historical overflow outlet pool signals, and input the real-time data into the overflow outlet simulation model for simulation;

[0142] A scheme formulation module 3, configured to judge the state of the overflow outlet pool according to the simulation result, formulate an adjustment scheme according to the state of the overflow outlet pool, and input the adjustment scheme into the overflow outlet simulation model for simulation, and iterate in turn until the data of the overflow outlet pool is qualified.

[0143] Preferably, in this embodiment, the overflow signal acquisition module 1 monitors the water level of the pool in real time through a high-precision liquid level sensor, collects the digital signal of the liquid level value, and compares it with the threshold set by the historical liquid level value digital signal; when the liquid level value is higher than the threshold, an overflow monitoring signal is output to achieve the functions of rapid response and early warning for abnormal water levels. The overflow simulation module 2 receives the overflow monitoring signal, collects the real-time data of the overflow outlet pool, preprocesses the data, constructs an overflow outlet simulation model based on historical data, and inputs the real-time data into the model for simulation; dynamically analyzes the state of the overflow outlet pool through the simulation model, provides a scientific basis for formulating an adjustment scheme, and improves the intelligent level of overflow management. The scheme formulation module 3 judges the state of the overflow outlet pool according to the simulation result, formulates an adjustment scheme, and inputs the scheme into the simulation model for iterative simulation until the data of the overflow outlet pool is qualified; through iterative optimization of the adjustment scheme, it ensures that the state of the overflow outlet pool reaches the expected goal, and improves the accuracy and efficiency of overflow management.

[0144] As Figure 11 shown, this embodiment provides an embodiment of an electronic device. In this embodiment, the electronic device 4 includes a processor 41 and a memory 42 coupled to the processor 41.

[0145] The memory 42 stores program instructions for implementing the layout method of the chemical pump body processing equipment according to any one of the above embodiments.

[0146] The processor 41 is configured to execute the program instructions stored in the memory 42 to perform the layout of the chemical pump body processing equipment.

[0147] Among them, the processor 41 can also be referred to as a CPU (Central Processing Unit). The processor 41 may be an integrated circuit chip with signal processing capabilities. The processor 41 can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0148] Furthermore, Figure 12 FIG. 5 is a schematic structural diagram of a storage medium according to an embodiment of the present application. The storage medium 5 of the embodiment of the present application stores program instructions 51 that can implement all of the above methods. Among them, the program instructions 51 can be stored in the above storage medium in the form of a software product, including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the methods described in various embodiments of the present application. The foregoing storage medium includes: various media that can store program codes such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disc, or a terminal device such as a computer, a server, a mobile phone, or a tablet.

[0149] In several embodiments provided by the present invention, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed mutual coupling or direct coupling or communication connection can be through some interfaces, and the indirect coupling or communication connection of devices or units can be in an electrical, mechanical, or other form.

[0150] In addition, each functional unit in various embodiments of the present invention can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit. The above is only the embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structural or equivalent process transformation made by using the specification and drawings of the present invention, or directly or indirectly applied to other related technical fields, shall be equally included in the patent protection scope of the present invention.

[0151] The specific embodiments of the invention have been described in detail above, but they are only examples, and the invention is not limited to the specific embodiments described above. For those skilled in the art, any equivalent modifications or substitutions to the invention are also within the scope of the invention. Therefore, all equivalent transformations, modifications, improvements, etc. made without departing from the spirit and principles of the invention should be covered within the scope of the invention.

Claims

1. A method for monitoring pond water accumulation based on an overflow outlet, characterized in that The pool water accumulation monitoring method based on the overflow outlet includes: Real-time monitoring of the pool water level through a high-precision liquid level sensor, collecting the digital signal of the liquid level value, comparing the digital signal of the liquid level value with the threshold set by the historical liquid level value digital signal. If the digital signal of the liquid level value is higher than the threshold, an overflow monitoring signal is output; Upon receiving the overflow monitoring signal, collecting the real-time data of the pool at the overflow outlet and preprocessing the real-time data to obtain a data set; constructing an overflow outlet simulation model based on the historical pool signal at the overflow outlet, and inputting the real-time data into the overflow outlet simulation model for simulation; Judging the state of the pool at the overflow outlet according to the simulation result, formulating an adjustment plan according to the state of the pool at the overflow outlet, and inputting the adjustment plan into the overflow outlet simulation model for simulation, and iterating in turn until the data of the pool at the overflow outlet is qualified.

2. The method for monitoring pond water accumulation based on an overflow port according to claim 1, wherein, The process of comparing the digital signal of the liquid level value with the threshold set by the historical liquid level value digital signal includes the following steps: Connect the high-precision liquid level sensor to the data acquisition system, perform data cleaning, smoothing and normalization preprocessing operations on the collected original data, quantize the preprocessed signal, and then encode the quantized data to convert the preprocessed electrical signal into a digital signal of the liquid level value; Input the digital signal of the liquid level value into the intelligent recognition program of the pool at the overflow outlet, sort it based on the time period, and perform statistical analysis on the collected historical digital signal of the liquid level value to calculate its mean value and standard deviation statistic; Set the pressure threshold according to the analysis result of the historical digital signal of the liquid level value, compare the real-time collected liquid level value signal with the set liquid level value threshold. If the collected liquid level value signal is within the range, continue to monitor. If it is higher than the liquid level value threshold, process it.

3. The method for monitoring pond water accumulation based on an overflow outlet according to claim 1, wherein The process of inputting the real-time data into the overflow outlet simulation model for simulation includes the following steps: Collect the liquid level, flow rate and time series signal data of the historical pool at the overflow outlet; perform preprocessing operations on the signal data to obtain a data set; expand the data set through time series translation, adding random noise and generative adversarial network; Analyze the periodic characteristics and trend terms through time series, convert the real-time data into features with preset meaningful values, and calculate the water level value; fuse the monitoring values at different positions of multiple sensor data; divide the data set into a training set, a validation set and a test set; Construct an overflow outlet simulation model based on the historical pool signal at the overflow outlet, and use the training set to train the overflow outlet simulation model; after training is completed, input the real-time data into the overflow outlet simulation model to generate an overflow outlet simulation result.

4. The method for monitoring pond water accumulation based on an overflow port according to claim 3, wherein, The process of expanding the data set through time series translation, adding random noise and generative adversarial network includes the following steps: Translate the data eye time axis of the original time series forward and backward to generate new data samples; translate the water level data as a whole by several hours to simulate the water level change patterns at different time periods; add random perturbations to the original data to simulate the measurement errors or signal interferences existing in the actual environment; Access the random noise input and output the synthesized time series data; distinguish the actual data from the generated data and provide a feedback signal; through iterative optimization until the preset realistic data threshold is reached, add the data that reaches the realistic data threshold to the dataset; Translate the time series, add random noise, and add the data that reaches the realistic data threshold to the dataset for integration to obtain the final dataset.

5. The method for monitoring pond water accumulation based on an overflow port according to claim 3, characterized in that, The process of fusing the monitoring values at different positions of multiple sensor data includes the following steps: Analyze the periodic characteristics and trend terms through time series analysis, convert the real-time data into features with preset meaningful values, and calculate the water level value; dynamically adjust the weights of sensors at different positions for feature complementarity; adaptively allocate the contribution degrees of data at each position according to the sensor accuracy and environmental interference factors; Classify the monitoring values of sensors in the same area or with similar functions to form a fusion unit; compare the real-time data with the historical data of the overflow water pool, and at the same time, the sensor data within the same cluster needs to reach the preset spatial correlation. If an anomaly occurs, mark it and trace back to the data source; Combine different modal data such as water level and flow velocity, extract the features of each modality through an encoder, and perform fusion in the latent space; convert the multivariate time series data into a unified representation, and then generate the fused monitoring value through a decoder.

6. The method for monitoring pond water accumulation based on an overflow outlet according to claim 5, characterized in that, The process of converting real-time data into features with preset meaningful values includes the following steps: Extract the periodic characteristics and trend terms through time series analysis; decompose the real-time data into trend, periodic, and residual components; calculate the mean and variance statistics within the time window; extract the frequency domain features; Identify the starting point of time according to the preset parameters; extract the statistical features of the real-time data; where the statistical features include the mean, maximum value, minimum value, and standard deviation, to quantify the volatility and stability of the real-time data; Combine the preset high pool duration to divide the time range and discard the data of the non-timekeeper; combine the trend term and slope to predict the future water level change.

7. The method for monitoring the pond water accumulation based on the overflow port according to claim 3, wherein The process of training the overflow water simulation model using the training set includes the following steps: Divide the training set into multiple small batches, where each small batch has 64 samples; perform forward propagation by inputting one batch each time, and calculate the predicted value of the input data through the network; calculate the loss between the predicted value and the true value, and use the backpropagation algorithm to adjust the weights; After each epoch, use the validation set to evaluate the model performance, and adjust the learning rate, batch size, and network parameters according to the validation results; if the validation loss does not improve after reaching the preset number of rounds, terminate the training in advance; Use the independent test set to evaluate the mean square error, accuracy, and POC curve of the overflow water simulation model; and integrate the trained overflow water simulation model into the overflow water monitoring system to process the sensor data in real time and output the simulation results.

8. The method for monitoring pond water accumulation based on an overflow outlet according to claim 1, wherein The process of judging the state of the overflow water pool according to the simulation results includes the following steps: Compare the real-time liquid level and flow parameters output by the simulation with the preset safety thresholds; if the liquid level exceeds the first-level threshold, trigger an early warning; if it reaches the second-level threshold, it is determined that the overflow risk is high, and different treatment plans are immediately formulated; Identify abnormal patterns through statistical calculations and trend analysis; check whether the simulation results are consistent with historical data. If the deviation between the simulated predicted liquid level and the time monitoring value exceeds the preset deviation threshold, adjust the data of the overflow port simulation model and iterate until qualified. Conduct a comprehensive score by combining multi-dimensional data; generate an alarm by determining whether the multi-dimensional data is within the preset numerical range; if the simulation results simultaneously meet the conditions that the liquid level is lower than the threshold, the flow rate is stable, and the equipment is operating normally, then determine that the status is qualified; otherwise, formulate an adjustment plan and conduct a simulation verification again.

9. The method for monitoring pond water accumulation based on an overflow port according to claim 8, wherein The process of formulating an adjustment plan and conducting a simulation verification again includes the following steps: When the liquid level, flow rate, or equipment parameters exceed the preset range, determine the source of abnormal data through statistical calculations, and use a contribution diagram to analyze the contribution degree of each variable to the failure, and identify the key adjustment target; if the liquid level deviation is caused by a flow rate fluctuation, then give priority to adjusting the flow control parameters. Conduct a backtest based on historical data to dynamically adjust the threshold; retrieve the preliminary adjustment value from the overflow port pool database according to the liquid level deviation value and the preset reference interval; if the deviation still exceeds the preset threshold after adjustment, trigger a stuck alarm. Re-run the simulation with the modified parameters, monitor the residual convergence, time step, and number of iterations; form an adjustment plan and verify the multi-adjustment plan through a multi-dimensional scoring mechanism; input the modified plan and the modified overflow port pool data into the database for storage.

10. A pool water accumulation monitoring system based on an overflow outlet, which is applied to the pool water accumulation monitoring method based on an overflow outlet as described in any one of claims 1 to 9, and is characterized in that, The pool water accumulation monitoring system based on the overflow port includes: An overflow signal acquisition module, which is used to monitor the water level of the pool in real time through a high-precision liquid level sensor, collect the digital signal of the liquid level value, compare the digital signal of the liquid level value with the threshold set by the historical liquid level value digital signal, and if the digital signal of the liquid level value is higher than the threshold, output an overflow monitoring signal. An overflow simulation module, which is used to receive the overflow monitoring signal, collect the real-time data of the overflow port pool, and preprocess the real-time data to obtain a data set; construct an overflow port simulation model based on the historical overflow port pool signal, and input the real-time data into the overflow port simulation model for simulation. A plan formulation module, which is used to judge the status of the overflow port pool according to the simulation results, formulate an adjustment plan according to the status of the overflow port pool, and input the adjustment plan into the overflow port simulation model for simulation, and iterate until the data of the overflow port pool is qualified.

Images