Water tank liquid level feedback monitoring flow self-adaptive adjusting method

Abstract

The invention discloses a water tank liquid level feedback monitoring flow self-adaptive adjusting method, and belongs to the technical field of fluid measurement. A water tank liquid level feedback monitoring flow self-adaptive adjusting method comprises the steps that liquid level data are collected, sectional area calculation is selected according to the type of a water tank, and a PID parameter set is dynamically switched to adjust the opening degree of a valve; the number of flowmeters installed in a pipeline can be reduced, the problem of installing the flowmeters in a complex environment is solved, the equipment cost is reduced, maintenance tasks are reduced, and meanwhile safe and reliable high-precision flow data are provided.

Description

Technical Field

[0001] The present invention relates to the technical field of fluid measurement, and specifically to a method for adaptively adjusting the flow rate by monitoring the water tank liquid level feedback, which is applicable to occasions such as water supply systems, industrial circulating water systems, and fire fighting water tanks that require precise flow rate monitoring but are inconvenient to install traditional flow meters. Background Art

[0002] In industries such as water supply, chemical engineering, and environmental protection, it is usually necessary to monitor the liquid flow rate in pipelines or water tanks. Traditional methods mainly use electromagnetic flow meters, ultrasonic flow meters, mechanical flow meters, etc. These devices have high costs, complex installations, mechanical wear, require regular maintenance, and are easily affected by pipeline conditions.

[0003] The technology for monitoring the water tank liquid level is relatively mature, but existing methods are usually only used for liquid level alarms or controls, and do not fully utilize the liquid level data to calculate the flow rate. Especially in the application scenario of the pump house water tank, due to the large volume of the water tank and obvious liquid level changes, it is very suitable to calculate the flow rate through the liquid level changes. Therefore, the present invention proposes a method for adaptively adjusting the flow rate by monitoring the water tank liquid level feedback, which uses the data of the liquid level sensor and combines the geometric parameters of the water tank to achieve the calculation of flow rate data without installing instruments on complex pipelines, reduce equipment costs, and improve system reliability. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method and system for accurately calculating the water flow rate only by monitoring the change of the water tank liquid level without installing a flow meter on the pipeline, so as to overcome the problems in the prior art such as difficult installation of the flow meter, high maintenance cost, and limited measurement accuracy, and to control the valve opening in real time according to the calculated flow rate data.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions: A method for adaptively adjusting the flow rate by monitoring the water tank liquid level feedback, comprising: S1. Collect liquid level data at a frequency of 1 - 10 Hz , with an accuracy ≤ ±1 mm; S2. Select the cross-sectional area calculation according to the type of water tank : Regular water tank: Calculate or , Irregular water tank: Fit the boundary curve by the cubic spline interpolation method under natural boundary conditions, and integrate to obtain ; S3. Calculate using the difference method, combine and leakage compensation , and output: ; S4. According to the flow rate deviation , dynamically switch PID parameter group to adjust valve opening; in, is the radius of the circular water tank, is the length of the rectangular water tank, Width of rectangular water tank, is the boundary curve function on the water tank section, is the lower boundary curve function of the water tank section, is the water inlet flow at time t, is the water flow rate at time t, is the volume integral corresponding to the liquid level change, is the sampling time interval, Set a value for the target flow rate.

[0006] Preferably, in step S2: For irregular water tanks, the discrete points of the cross section are measured in sections according to the liquid level height; Constructing a cubic spline function Smoothness conditions are enforced when : and , in, is the j-th cubic spline function, is the first-order derivative at the connection point, is the second-order derivative at the connection point.

[0007] Preferably, after step S1, the following steps are performed: S1.1. Apply Kalman filtering to the liquid level data, and the state equation is defined as: ; The process noise covariance Dynamic adjustment according to traffic fluctuations; is the state vector at the moment, is the process noise.

[0008] Preferably, in step S3: S3.1. Obtain water temperature T in real time and calculate density compensation coefficient: ; S3.2, press Corrected volume calculation; in, is the density after temperature compensation, is the density at the reference temperature, is the coefficient of thermal expansion, is the current water temperature, is the reference temperature, is the volume after temperature compensation, is the instantaneous volume change before temperature compensation.

[0009] Preferably, the is obtained through night static testing: Close the inlet valve and record the liquid level change rate ; Calculate ; is the leakage compensation flow rate.

[0010] Preferably, in step S4: When , adopt ; When , adopt ; When , adopt ; is the parameter group vector.

[0011] Preferably, in step S4: S4.1. Calculate the step size ; S4.2. Constrain the single opening change ; is the opening adjustment step size, is the sensitivity coefficient, is the actual opening change, is the proportional gain coefficient.

[0012] Preferably, integrate to obtain the cumulative flow rate ; When , trigger recalibration; is the reference cumulative flow rate.

[0013] Preferably, when applied to a parallel water tank group: Calculate the total flow rate deviation ; Distribute the inlet valve opening increment according to the volume ratio of each water tank; is the sum of the flow rates of each water tank.

[0014] Preferably, it further includes: Build a digital twin model of the water tank in the cloud and input historical liquid level and flow rate data; Predict future flow rate fluctuations through the LSTM network ; Embed the predicted value into the feedforward compensation term of the edge device PID controller.

[0015] Compared with the prior art, the present invention provides a method for adaptive regulation of water tank liquid level feedback monitoring flow rate, which has the following beneficial effects: 1. This method for adaptive regulation of water tank liquid level feedback monitoring flow rate breaks through the rigid constraints of traditional reliance on pipeline flow meters, and first creates a technical chain of "liquid level change rate → cross-sectional area integration → flow rate calculation → valve control", solves the problem of instrument installation in complex environments, completely eliminates pipeline flow meters, and realizes the integration of flow rate monitoring and control through liquid level feedback closed-loop, with an error ≤ 3%.

[0016] 2. This method for adaptive regulation of water tank liquid level feedback monitoring flow rate integrates segmented PID parameter switching, anti-saturation step size constraint and cloud-edge collaborative prediction to achieve fast and stable tracking in scenarios of sudden flow rate changes.

[0017] In the method for adaptive regulation of water tank liquid level feedback monitoring flow rate, the parts not involved are the same as or can be implemented by the prior art. The present invention can reduce the number of flow meters installed in pipelines, solve the problem of installing flow meters in complex environments, reduce equipment costs, reduce maintenance tasks, and at the same time provide safe, reliable and high-precision flow rate data. Description of the Drawings

[0018] Figure 1 is a flowchart of a method for adaptive regulation of water tank liquid level feedback monitoring flow rate proposed by the present invention; Figure 2 is the overall network topology diagram of a method for adaptive regulation of water tank liquid level feedback monitoring flow rate proposed by the present invention; Figure 3 is a comparison diagram of the calculation data and actual data effects of a method for adaptive regulation of water tank liquid level feedback monitoring flow rate proposed by the present invention. Detailed Embodiments

[0019] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the 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 of the embodiments.

[0020] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention.

[0021] Reference Figures 1 - 3 , a method for adaptively adjusting the flow rate by monitoring the water tank liquid level feedback, comprising: S1. Collect liquid level data at a frequency of 1-10 Hz , with an accuracy of ≤ ±1 mm; S2. Select the cross-sectional area calculation according to the water tank type : Regular water tank: Calculate or , Irregular water tank: Fit the boundary curve under natural boundary conditions by the cubic spline interpolation method, and integrate to obtain ; S3. Use the difference method to calculate , combined with and leakage compensation , and output: ; S4. According to the flow deviation , dynamically switch the PID parameter group to adjust the valve opening; Wherein, is the radius of the circular water tank, is the length of the rectangular water tank, the width of the rectangular water tank, is the upper boundary curve function of the water tank cross-section, is the lower boundary curve function of the water tank cross-section, is the influent flow rate at time t, is the effluent flow rate at time t, is the volume integral corresponding to the liquid level change, is the sampling time interval, is the target flow rate set value.

[0022] In the present invention, non-invasive flow monitoring is adopted, getting rid of the dependence on pipeline flow meters in the traditional scheme (complicated installation and high cost), and only inverting the flow rate through the liquid level change, canceling the effluent flow meter; Compared with the traditional PID with fixed parameters, which has the disadvantages of large overshoot and slow response, this scheme uses a dynamically switched PID parameter group to achieve: fast convergence for large deviations and precise steady state for small deviations

[0023] Furthermore, for high-precision liquid level data acquisition: an ultrasonic liquid level gauge or a radar liquid level sensor is used, the sampling frequency is set to 1-10 Hz, the measurement accuracy reaches ±1 mm, and the liquid level data h(t) is recorded in real time; Accurate configuration of water tank parameters: divided into rectangular water tanks, circular water tanks, and irregular water tanks.

[0024] Circular water tank: accurately measure the radius r and calculate its cross-sectional area:

[0025] Rectangular water tank: Accurately measure the length L and width W, and calculate its cross-sectional area:

[0026] Irregular water tank: The sectional measurement method can be adopted, and the cross-sectional area is measured every 10 cm in height. Establish a liquid level - volume correspondence table, and use the cubic spline interpolation method to calculate the cross-sectional area at any liquid level. Specifically: Obtain the coordinates of discrete points on the cross-section boundary of the water tank (such as through measurement or sensor acquisition), and each point contains two coordinate values (such as ).

[0027] Sort the data points in ascending order according to the abscissa (or ordinate) to form an ordered sequence. For example, after sorting by the x coordinate, the data points are .

[0028] Then construct a cubic spline interpolation function and construct a cubic polynomial between adjacent data points , satisfying the following conditions: Continuity: ; Smoothness: The first derivative , the second derivative ; Boundary conditions: Usually, natural boundary conditions or fixed boundary conditions (known derivative values at the endpoints) are adopted.

[0029] Calculation of cross-sectional area: Calculate the area between the two boundary curves through integration. If the cross-section of the water tank consists of two upper and lower curves and , then the cross-sectional area is:

[0030] Specific steps: Fit the upper and lower boundaries respectively: Perform cubic spline interpolation on the discrete points of the upper and lower boundaries to obtain the functions and .

[0031] Numerical integration: The integral or trapz function can be used to calculate the definite integral.

[0032] Accurate calculation of liquid level change rate: The difference method is used to calculate the liquid level change rate.

[0033]

[0034] Among them, is the sampling interval time.

[0035] According to the law of conservation of mass, the change in the volume of water in the water tank is equal to the cumulative value of the discharge flow. Let: be the liquid level at the cross-sectional area of the water tank (m²); be the time within the liquid level change (m); be the instantaneous water discharge flow (m³ / s); Then the flow calculation formula is:

[0036] If the water tank is a regular cylindrical water tank, , the formula can be simplified to:

[0037] The actual scenario needs to consider the following interference factors: Superposition of influent flow: If the water tank is refilled simultaneously, the influent flow needs to be deducted :

[0038] Liquid level fluctuation filtering: A first-order low-pass filter is used to eliminate high-frequency noise (α = 0.1 is the filtering coefficient):

[0039] Leakage compensation: Calibrate the leakage flow through night static testing , and the correction formula is:

[0040] Treatment of irregular water tanks: For conical or stepped water tanks, the cross-sectional area needs to be calculated in segments. For example, the of a conical water tank can be expressed as (k is the taper coefficient):

[0041] Data optimization processing: Adopt dynamic compensation filtering method. In terms of temperature compensation, if the liquid density is affected by temperature (such as in a hot water system), a temperature sensor is used for correction. In terms of liquid level fluctuation noise, moving average filtering or Kalman filtering is used to eliminate liquid level fluctuations. Moving average filtering suppresses noise interference by maintaining a fixed-length queue and averaging the continuously sampled liquid level data. Its core idea is: using the average value of the current moment and the previous N - 1 moment's sampling values as the current filtering result.

[0042] Filtering result

[0043] where is the current and previous N - 1 sampling values.

[0044] Kalman filtering is a recursive estimation algorithm based on the state - space model. By combining the system dynamic model and the observed data, it realizes the optimal estimation of the state of a dynamic system. Its core idea is to balance the model prediction and the actual observation through two steps of prediction and update, thereby reducing noise interference and improving data accuracy. Its application in the processing of the liquid - level data of the pump house water tank is an efficient and real - time solution, which can effectively suppress noise interference and dynamically track the liquid - level changes.

[0045] 1. Define the state and observation models: liquid - level height It can be extended to the liquid - level change rate to capture dynamic characteristics. Then the change of the liquid - level over time is:

[0046] where, is the inflow and outflow, is the process noise (reflecting model error).

[0047] Observation equation: the relationship between the sensor measurement value and the true liquid - level:

[0048] where, is the observation noise (such as sensor error).

[0049] 2. Initialize the parameters Initial state: Take the first measurement value as the initial estimate).

[0050] Initial covariance: Reflects the uncertainty of the initial estimate (usually set to a relatively large value).

[0051] Noise covariance: The process noise Q characterizes the model error (such as flow rate fluctuations) and needs to be adjusted according to the system stability. The observation noise R reflects the sensor accuracy (such as the error of an ultrasonic liquid - level gauge).

[0052] 3. Prediction and update Prediction step:

[0053] In the above formula, A is the state - transition matrix, B is the control matrix, is the input flow rate.

[0054] Update step:

[0055] where is the Kalman gain, which balances the weights of prediction and observation.

[0056] 4. Parameter Optimization Adjust Q and R: If the liquid level changes frequently (such as frequent start and stop of the water pump), increase Q to enhance the model adaptability; if the sensor has high accuracy, decrease R to trust the observed value more.

[0057] Handle sudden interference: When the liquid level suddenly changes (such as pipeline leakage), the limited amplitude filtering or switching to the strong tracking filtering algorithm can be combined.

[0058] Cumulative flow calculation: Integrate the instantaneous flow to obtain the cumulative flow V:

[0059] Valve opening control: Adjust the proportional (P), integral (I), and derivative (D) parameters of the PID controller inside the PLC to obtain the approximate valve opening:

[0060] Among them 、 、 are the PID parameters respectively. Respond quickly to the current error:

[0061] Among them, SP is the target flow value, and PV is the actual flow value collected by the sensor.

[0062] When the error value is large, the segmented step size algorithm can be used for optimization. Dynamically adjust the step size according to the error signal, and introduce a segmented function (such as the sigmoid-sinh function) to further optimize the step size adjustment rule: If the error is large, increase the step size to accelerate convergence; If the error is small, decrease the step size to stabilize the output.

[0063] Refer to Figures 1 - 3 In step S2: Measure the discrete points of the cross-section by segment according to the liquid level height for the irregular water tank; Construct a cubic spline function and force it to meet the smoothness condition when and ; Among them, is the cubic spline function of the jth segment, is the first derivative at the connection point, is the second derivative at the connection point.

[0064] Execute after step S1: S1.1. Apply Kalman filtering to the liquid level data, and the state equation is defined as: [[ID=Q8]]

[0065] where the process noise covariance is dynamically adjusted according to the flow rate fluctuation amplitude; is the state vector at time, is the process noise.

[0066] In step S3: S3.1. Obtain the water temperature T in real time and calculate the density compensation coefficient:

[0067] S3.2. Correct the volume calculation according to ; where, is the density after temperature compensation, is the density at the reference temperature, is the coefficient of thermal expansion, is the current water temperature, is the reference temperature, is the volume after temperature compensation, is the instantaneous volume change before temperature compensation.

[0068] is obtained through night static testing: Close the inlet valve and record the liquid level change rate ; Calculate ; is the leakage compensation flow rate.

[0069] Referring to Figures 1 - 3 in step S4: When , adopt ; When , adopt ; When , adopt ; is the parameter group vector.

[0070] In step S4: S4.1. Calculate the step size ; S4.2. Constrain the single opening change ; is the opening adjustment step size, is the sensitivity coefficient, is the actual opening change amount, is the proportional gain coefficient.

[0071] Pair Integrate to obtain the cumulative flow rate ; When occurs, trigger recalibration; is the reference cumulative flow rate.

[0072] To address the problem of frequent valve actuation, shortened lifespan, and water hammer effect caused by sudden PID output changes, the following measures are taken: Step size smoothing: Use the hyperbolic tangent function to constrain the step size change rate Hard limit: Limit the maximum change in opening per single time.

[0073] Refer to Figures 1 - 3 When applied to a parallel water tank group: Calculate the total flow deviation ; Allocate the opening increment of the inlet valve according to the volume ratio of each water tank; is the sum of the flow rates of each water tank.

[0074] It also includes: Build a digital twin model of the water tank in the cloud and input historical liquid level and flow data; Predict future flow fluctuations through the LSTM network ; Embed the predicted value into the feedforward compensation term of the edge device PID controller.

[0075] When applied to a parallel water tank group, first calculate the total flow deviation, and then allocate the opening increment according to the volume ratio:

[0076]

[0077] Cloud training of the LSTM model: Input: Historical liquid level, temperature, valve opening (sampling interval 1 min); Output: Future 5-minute flow prediction ; Feedforward compensation: The edge controller receives the predicted value and corrects the PID output: .

[0078] In the example of the present invention, the secondary water supply pump room in an old community in a certain community was built in 2010. The original electromagnetic flowmeter was used to monitor the water outlet flow. In 2022, equipment aging led to a measurement error of 15%, and the maintenance cost was too high, and the pipeline environment was complex. At present, this community has become a peak-shifting water supply pilot project, which needs to adjust the inlet electric control valve in real time according to the inlet flowmeter and valve at the front end of the water tank, the outlet flowmeter of the water tank, and according to the algorithm instructions to achieve peak-shifting water supply, balance pressure, reduce fluctuations, and save energy and reduce carbon emissions. In order to reduce the transformation cost and improve the monitoring reliability, it is planned to use the existing rectangular water tank (size 9.5m×6m×2m) to calculate the water outlet flow through the change of the water tank liquid level, and then realize the control of the valve.

[0079] Referring to Figure 1 In the steps of, for the hardware construction, a high-precision ultrasonic liquid level gauge with a measuring range of 0.1 - 5m and an accuracy of ±0.5mm is installed on the top of the water tank, emitting signals vertically downward; an electric control valve with an adjustable range of 0 - 100 degrees is installed on the water inlet pipe of the water tank, and a high-precision electromagnetic flowmeter is installed to calculate the water inlet flow ; The data acquisition module is to add a PLC module, and upload the data to the cloud server for storage through the MQTT network protocol.

[0080] Referring to Figure 1 In the steps of, the specific method of data acquisition and analysis is as follows: The on-site PLC uses the industrial bus communication mosbus rtu mode to collect the water tank liquid level data and the water inlet flowmeter data, and uses the analog signal to collect the valve opening signal and control the valve. During the data acquisition process, the liquid level data will be affected by noise fluctuations, data loss, data anomalies, etc. caused by the water inlet of the water tank and electromagnetic interference. For Figure 1 In the data analysis steps of, the Kalman filtering method is adopted, and the liquid level gauge data and the flow model are fused through the Kalman filtering algorithm to achieve a more accurate liquid level estimation.

[0081] The implementation of the Kalman filter is as follows: First, initialize the liquid level and flow states: (Liquid level 1.5m, flow rate 0.005m³ / s) Initial covariance:

[0082] Prediction step:

[0083] Update step: Kalman gain:

[0084] State update:

[0085] The experimental data and filtering effect are shown in Table 1 below: Table 1. Water tank level processed by Kalman filter

[0086] For the missing data, the average value of the previous and subsequent data is taken by the difference method and inserted. The outliers are removed according to the change rate.

[0087] Refer to Figure 1 In the steps of, because the influent flow rate is included and the water tank is rectangular, after processing the liquid level data and collecting the influent flow rate data, combined with the flow rate calculation formula, the specific method for the effluent flow rate is as follows: -

[0088] The data is calculated every 5 minutes. Substituting the measured liquid level and influent flow rate into the above formula, the effluent flow rate for a certain period of time is as shown in Table 2 below: Table 2. Flow rate data calculated according to the liquid level

[0089] Refer to Figure 1 In the steps of, upload the processed liquid level data, flow rate and other data to the server side, using the MQTT protocol. This protocol requires configuration of address, port, topic, and data. The server side receives the data, parses and stores it.

[0090] Refer to Figure 1 In the steps of, after the data upload is completed, the algorithm runs and calculates the required influent flow rate of the water tank , and send it to the edge computing gateway (flow control) device. The built-in PID control algorithm adjusts the valve opening. Among them, P = 0.5, I = 6, and D = 0 are set. The overall adjustment is 5 minutes, and the flow rate error value oscillates around ±5~8m³, which is relatively large and does not converge. At this time, the segmented step size algorithm is used for fine adjustment. This method sets the flow deviation partition, which is divided into large, medium, and small deviations. The large deviation , the medium deviation , the small deviation , where B1 and B2 are thresholds. Different intervals use different step size formulas. Among them, the large deviation interval: ( is the large deviation proportionality coefficient) The medium deviation interval: , and an integral term is added to compensate for the steady-state error.

[0091] The small deviation interval: , and a differential term is introduced to suppress overshoot.

[0092] Then adjust the opening according to the step size and limit the maximum adjustment amount per single time. Combine the PID output with the step size constraint to ensure smooth change of the opening. And monitor the flow rate in real time through the sensor, update the deviation and repeat the above steps.

[0093] The key lies in setting different PID parameters in different intervals, such as a high proportional gain in the large deviation interval , a low integral gain , a low derivative gain , reduce in the small deviation interval , increase and to suppress oscillations. The following parameters are obtained through testing; Table 3. Parameter settings in different intervals

[0094] Measurement and simulation: Combined with the formulated plan and the processed data, mainly conduct constant flow experiments, dynamic flow tests, and overall flow tests.

[0095] Under the constant flow test, set the opening of the water tank inlet valve to 0, that is, the water inlet volume is 0, set the frequency of the water pump to a fixed 30 Hz, and install a clamp-on ultrasonic flowmeter on the water tank outlet pipe (for convenient installation and comparison of experimental data). The theoretical outlet flow rate is 10 m³ / h.

[0096] By recording the experimental data, as shown in Table 4 below; Table 4. Constant flow experiment

[0097] It can be seen from Table 4 that the error under the constant flow condition is stable within 3%.

[0098] Under the dynamic flow test, set the opening of the inlet valve to 0, and switch the outlet mechanical valve of the water tank every 10 minutes (5 m³ - 10 m³ - 5 m³ cycle). Record the data as shown in Table 5 below; Table 5. Dynamic flow test

[0099] It can be seen from Table 5 that the calculation errors of both the flow peak value and the flow valley value are within 6%, and most of them are concentrated in 3%. When the dynamic response time is less than 3 seconds, the error is controllable. [[ID=Q8]]

[0100] Under the overall flow test, including the leakage flow , the water tank inlet flow , according to the calculated outlet flow rate of the inlet valve, the algorithm calculates the water tank inlet flow value and sends it to the PLC for PID adjustment of the valve. Among them, set the leakage flow It is 0.2 m³ / h. The algorithm flow rate instruction is calculated based on the predicted flow rate, and the recorded data is shown in Table 6 below; Table 6. Overall Flow Rate Test

[0101] As can be seen from Table 6, except for the error within 0 - 10 minutes being about 14%, the rest are stable around 4%. The overall error is within the controllable range, enabling accurate calculation of the effluent flow rate and precise PID control of the inlet valve.

[0102] Table 7. Cost Comparison

[0103] As can be seen from Table 7, taking this community as an example, in the case of 1 pump house and 3 pressurization zones, the cost savings are as high as 92.4%, showing very high economic benefits.

[0104] Judging from the above experimental results, this invention has the following significant advantages compared with the existing technologies: 1. Completely non-invasive measurement, without the need to drill holes or install any devices on the pipeline; 2. It has significant cost advantages, saving the procurement cost of flow meters, reducing installation costs, and lowering operation and maintenance costs; 3. Excellent measurement performance with small calculation errors; 4. Strong applicability, not restricted by pipeline conditions, and applicable to various water qualities.

[0105] The above is only the preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and all should be covered within the protection scope of the present invention.

Claims

1. A method for adaptively adjusting the flow rate by monitoring the water tank liquid level feedback, characterized in that, Including: S1. Collect liquid level data at a frequency of 1 - 10 Hz , with an accuracy of ≤ ±1 mm; S2. Calculate according to the cross-sectional area selected based on the water tank type : Regular water tank: calculation or , Irregular water tank: Fit the boundary curve under natural boundary conditions by cubic spline interpolation method, and integrate to obtain ; S3. Calculate using the difference method , combined with and leakage compensation , output: ; S4. According to the flow deviation , dynamically switch the PID parameter group to adjust the valve opening; Among them, is the radius of the circular water tank, is the length of the rectangular water tank, is the width of the rectangular water tank, is the function of the upper boundary curve of the water tank cross-section, is the function of the lower boundary curve of the water tank cross-section, is the water inflow rate at time t, is the water outflow rate at time t, is the volume integral corresponding to the liquid level change, is the sampling time interval, is the target flow rate setting value.

2. The method for adaptively adjusting the flow rate by monitoring the water tank liquid level feedback according to claim 1, wherein In step S2: Measure the discrete points of the cross-section of the irregular water tank by dividing it according to the liquid level height; Constructing a cubic spline function Smoothness conditions are enforced when : and , Among them, is the cubic spline function of the j-th segment, is the first derivative at the connection point, is the second derivative at the connection point.

3. A method for adaptively adjusting the flow rate by monitoring the water tank liquid level feedback according to claim 1, characterized in that, Execute after step S1: S1.

1. Apply Kalman filtering to the liquid level data, and the state equation is defined as: ； wherein the process noise covariance is dynamically adjusted according to the amplitude of the flow rate fluctuation; is the moment state vector, is the process noise.

4. A method for adaptively adjusting the flow rate by monitoring the water tank liquid level feedback according to claim 1, characterized in that, In step S3: S3.

1. Obtain the water temperature T in real time and calculate the density compensation coefficient: ； S3.

2. Calculate according to the corrected volume; Among them, is the density after temperature compensation, is the density at the reference temperature, is the coefficient of thermal expansion, is the current water temperature, is the reference temperature, is the volume after temperature compensation, is the instantaneous volume change before temperature compensation.

5. A method for adaptively adjusting the flow rate by monitoring the water tank liquid level feedback according to claim 1, characterized in that, described Obtained through nightly static testing: Close the inlet valve and record the liquid level change rate ; Calculation ; Compensate flow for leaks.

6. The flow rate adaptive regulation method for water tank liquid level feedback monitoring according to claim 1, wherein, In step S4: When is adopted ; When is the case, adopt ; When is adopted ; is a parameter group vector.

7. A method for adaptively adjusting the flow rate by monitoring the water tank liquid level feedback according to claim 1, characterized in that In step S4: S4.

1. Calculate the step size ; S4.

2. Restrict the change of the single opening degree ; is the opening adjustment step size, is the sensitivity coefficient, is the actual opening change amount, is the proportional gain coefficient.

8. A method for adaptively adjusting the flow rate by monitoring the water tank liquid level feedback according to claim 1, characterized in that, For Integrating gives the cumulative flow rate ; When is triggered recalibration; is the reference cumulative flow rate.

9. A method for adaptively adjusting the flow rate by monitoring the water tank liquid level feedback according to claim 1, characterized in that, When applied to a parallel water tank group: Calculate the total flow deviation ; Distribute the increment of the opening of the inlet valve according to the volume ratio of each water tank; is the sum of the flows of each water tank.

10. A method for adaptively adjusting the flow rate by monitoring the water tank liquid level feedback according to claim 1, characterized in that, Also including: Build a digital twin model of the water tank in the cloud and input historical liquid level and flow data; Predicting Future Traffic Fluctuations through LSTM Networks ; Embed the predicted value into the feed-forward compensation term of the edge device PID controller.

Images