A real-time parallel variable frequency pump group flow calculation method

Abstract

The application discloses a kind of real-time parallel variable frequency pump group flow calculation methods, this method can not rely on flowmeter, realize parallel variable frequency pump group main road / each branch flow calculation. Using pump real-time power data, based on the head-flow characteristic and efficiency-flow characteristic of pump containing variable frequency working condition, the possible flow value of each variable frequency pump branch of parallel pump group is calculated and obtained, combined with pump group flow calculation evaluation system including pump group flow matching, head matching and running stability matching, the pump group flow combination scheme with highest reliability is selected, and pump group flow data is output.The application can be used for the flow calculation of different types and different operating states of pump, and has good effect for the flow calculation of pump in different operating conditions, and establishes an important foundation for the development of pump group optimization control.

Description

Technical Field

[0001] This invention relates to the field of public works control and energy-saving optimization, specifically to a method for calculating the flow rate of a real-time parallel variable frequency pump set. Background Technology

[0002] Pumps account for approximately 20% of the nation's total annual electricity consumption, making their improved operating efficiency crucial for achieving energy conservation and emission reduction goals. Improving pump operating efficiency significantly reduces the cost over the entire pump product lifecycle. Data shows that pump motor efficiency can be further increased by 2%–5%, and there is still a 3%–7% potential for improvement in pump design and manufacturing.

[0003] In intelligent pump systems, to ensure efficient and reliable operation, multiple pumps are often connected in parallel. The operating conditions of each pump are adjusted by controlling its speed and valves, ensuring each pump operates within its optimal range. Flow rate is a crucial parameter for regulating the water circulation system of parallel pump sets. However, due to the susceptibility of flow meters to damage in industrial settings, their inherent measurement errors, or unsuitable pipeline conditions for installation / modification / addition of flow meters, obtaining flow data for the main and branch lines in parallel pump set water circulation systems is difficult and inaccurate. This severely impacts the optimization and regulation process of the parallel pump set water circulation system, reducing the feasibility and reliability of optimization solutions.

[0004] Power-flow prediction is a commonly used flow calculation method in industry. It quickly estimates the current flow rate of a pump using pump power data collected on-site. The application of power-flow prediction relies on an accurate pump power-flow characteristic curve. However, industrial pumps often suffer from missing power-flow characteristic parameters or curve shifts due to aging, limiting the application examples and resulting in low accuracy. Furthermore, the nonlinear power-flow characteristic curve increases the error in the prediction results under low-flow operating conditions and when frequency variables are introduced. Therefore, traditional, simple power-flow prediction is unsuitable for the regulation and optimization of parallel pump group water circulation systems with high reliability requirements. Summary of the Invention

[0005] The technical problem solved by this invention is to provide a method for real-time calculation of the flow rate of a variable frequency pump. This method can accurately calculate the pump flow rate even when the pump's basic parameters and factory characteristic curves are lacking.

[0006] This invention discloses a method for real-time calculation of the flow rate of a variable frequency pump, comprising the following steps:

[0007] S1 filters historical data on flow rate, head, and power.

[0008] S2, use the filtered data to perform curve fitting to generate the pump's head-flow characteristic curve and efficiency-flow characteristic curve; or use the filtered data to perform curve calibration to update the existing head-flow characteristic curve and efficiency-flow characteristic curve.

[0009] S3, read the power or current from the meter and then calculate the shaft power;

[0010] S4. Obtain the power-flow characteristic curve through the head-flow characteristic curve and the efficiency-flow characteristic curve, and perform power-flow calculation based on the shaft power.

[0011] S5 evaluates the calculation results by calculating and evaluating the pump group flow rate, and outputs the pump group flow rate data.

[0012] Furthermore, the filtering of historical data on flow rate, head, and power includes:

[0013] S11, perform statistical analysis on the historical operating data of parallel pump sets;

[0014] S12, based on the user-set adjustable threshold, determine outliers in the historical operating data of the parallel pump group after statistical analysis;

[0015] S13, remove or replace outliers;

[0016] S14, extract characteristic data reflecting pump operation.

[0017] Furthermore, the outliers are data points below 1.5 times the interquartile range of the quarter quartile, or data points above 1.5 times the interquartile range of the third quartile.

[0018] Furthermore, the formula for the head-flow characteristic curve is:

[0019] H=α0×q 2 +α1×q×i+α2×i 2

[0020] The formula for the efficiency-flow characteristic curve is:

[0021]

[0022] Where H is the pump head; q is the pump flow rate; i is the variable frequency; η is the pump efficiency; and α0, α1, and α2, as well as β0 and β1 are coefficients.

[0023] Furthermore, the curve fitting involves determining the coefficients α0, α1, α2, β0, and β1; the curve calibration involves updating the coefficients α0, α1, α2, β0, and β1.

[0024] Furthermore, the shaft power calculation includes converting the actual pump power to the pump shaft power based on a power conversion formula, wherein the power conversion formula is...

[0025]

[0026] Among them, P shaft P is the shaft power of the pump. meter R is the power rating of the pump's meter. i Let n be the pump's iron and copper loss rates, n be the pump's rotational speed, and f be the frequency of the pump electrodes. pair The number of magnetic pairs of the motor; the loss rate is determined based on the pump's structure, service life, and operating status.

[0027] Furthermore, the power-flow calculation includes solving the power-flow characteristic curve equation using the bisection method based on efficiency prediction and pump frequency conversion characteristics, and calculating the pump flow rate under different operating conditions, including low flow rate and frequency conversion conditions.

[0028] Furthermore, the pump set flow calculation and evaluation is conducted for several candidate flow solution schemes, using an evaluation system that includes pump set flow matching, head matching, and stability matching, and outputting pump set flow data using a supervised method.

[0029] Furthermore, the fitness function and the penalty functions for flow rate, head, and stability of the evaluation system are respectively:

[0030]

[0031]

[0032]

[0033]

[0034] Where F cal For fitness; p s p is the stability penalty function; q The flow penalty function; p H Here, N is the head penalty function; N is the total number of flow rate data; N p q represents the number of pumps operating in the pump set. i Here is the flow rate data of the pump unit at time i; H j Let q be the head of the j-th pump; ave q represents the average flow rate of the pump unit during this period; req H represents the flow rate demand of the pump system during this period. meter The measured value of the pump head gauge.

[0035] The beneficial effects of this invention are:

[0036] 1. By updating and calibrating the pump's head-flow characteristic curve through pump stage operation data, the method can be used to calculate the flow rate of pumps of different types and operating conditions.

[0037] 2. Based on pump efficiency, the pump power to flow rate is calculated. The influence of pump frequency variation is fully considered in the process, so that the method has a good effect on the flow rate calculation of pumps under different operating conditions.

[0038] 3. The pump set flow calculation process has been realized, forming a pump set flow calculation and evaluation system that includes pump set flow matching, head matching, and operational stability matching. The monitoring method is used to ensure that the output of pump set flow data has high reliability, which lays an important foundation for the development of pump set optimized control. Attached Figure Description

[0039] Figure 1 This is a flowchart of the flow calculation method for parallel variable frequency pump sets.

[0040] Figure 2 This is a comparison chart of the actual flow rate and calculated flow rate of a single pump in Example 1.

[0041] Figure 3 This is a comparison chart of the actual flow rate and calculated flow rate of the dual pumps in Example 2.

[0042] Figure 4 This is a flowchart illustrating the flow calculation method for parallel variable frequency pump sets according to an embodiment of the present invention. Detailed Implementation

[0043] To further understand the present invention, preferred embodiments of the present invention are described below in conjunction with examples. However, it should be understood that these descriptions are only for further illustrating the features and advantages of the present invention, and not for limiting the scope of the claims of the present invention.

[0044] This invention provides a real-time method for calculating the flow rate of a parallel variable frequency pump set, achieving flow rate calculation for the main circuit / each branch circuit of the parallel pump set without relying on a flow meter. For example... Figure 1 and Figure 4 As shown, using the pump's factory head-flow characteristic curve as a benchmark, the head-flow characteristic curve is calibrated by selectively using the pump's operating flow data to obtain the current pump's head-flow characteristic curve and record key parameters of the characteristic curve. The real-time power / current data of the pump collected by the meter is processed to calculate the pump's shaft power data. Pump efficiency-flow characteristic curves and head-flow characteristic curves including a variable frequency term are established, and the power-flow characteristic curve is obtained by conversion. The shaft power data is then substituted, and the current pump flow rate is calculated using a bisection method combined with the supervision of the pump group flow calculation and evaluation method. Specifically, the following steps are included:

[0045] S1 filters historical data on flow rate, head, and power.

[0046] S2, use the filtered data to perform curve fitting to generate the pump's head-flow characteristic curve and efficiency-flow characteristic curve; or curve calibration to update the existing head-flow characteristic curve and efficiency-flow characteristic curve.

[0047] S3, calculate shaft power after reading the meter power or current;

[0048] S4. Obtain the power-flow characteristic curve through the head-flow characteristic curve and the efficiency-flow characteristic curve, and perform power-flow calculation based on the shaft power.

[0049] S5 evaluates the pump group flow rate calculation results, selects the optimal flow rate combination scheme, and outputs the pump group flow rate data.

[0050] The first step S1 includes:

[0051] S11, perform statistical analysis on the historical operating data of parallel pump sets;

[0052] S12, based on the user-set adjustable threshold, determine outliers in the historical operating data of the parallel pump group after statistical analysis;

[0053] S13, remove or replace outliers;

[0054] S14, extract characteristic data reflecting pump operation.

[0055] The head-flow characteristic curve, pump efficiency-flow characteristic curve, and power-flow characteristic curve containing the variable frequency term in this invention are shown in Eq.1-3, respectively. In the formulas, H is the pump head in meters (m); q is the pump flow rate in cubic meters (m³). 3 h -1 ; i is the variable frequency, which is the ratio of the current pump frequency to the power frequency (usually 50Hz); η is the pump efficiency; P is the pump power, in kW; α0, α1 and α2 are the head-flow characteristic curve coefficients; β0 and β1 are the efficiency-flow characteristic curve coefficients.

[0056] H=α0×q 2 +α1×q×i+α2×i 2 Eq.1

[0057]

[0058]

[0059] The key parameters of the head-flow characteristic curve refer to the quadratic coefficient α0, the linear coefficient α1, and the constant α2 of the pump's head-flow characteristic curve, which are fitted with a quadratic function model. The key parameters of the efficiency-flow characteristic curve refer to the quadratic coefficient β0 and the linear coefficient β1 of the pump's efficiency-flow characteristic curve, which are fitted with a quadratic function model passing through the origin.

[0060] To obtain the pump's head-flow characteristic curve, two methods can be used: obtaining it from the pump's manufacturer's instructions or fitting it with historical operating data. Furthermore, it's possible to set up the system to update and calibrate the pump's head-flow characteristic curve using pump stage operating data.

[0061] For pumps with long service lives or those lacking factory instructions, historical operating data can be used to fit the pump's head-flow characteristic curve. When fitting historical data, first read the historical operating data of the parallel pump group, typically including the first 800 hours of historical data. Then, perform outlier analysis on the historical data, set a user-adjustable outlier coefficient, determine the data anomaly judgment range, and remove or replace outliers based on actual needs, combining subjective and objective methods. Preferably, when removing outlier data, calculate the median and upper / lower quartiles of the historical data, and set a certain interquartile range (usually 1.5 * interquartile range) above / below the upper / lower quartiles as needed for outlier data range. Outlier data within this range is removed, or data can be replaced through interpolation. Then, perform feature data extraction. Feature data refers to data that reflects pump characteristics extracted from the pump's historical operating data. In this invention, flow rate data with low correlation to pump frequency variation is preferred.

[0062] Characteristic curve calibration refers to the process of refitting the characteristic curve using the pump's factory characteristic curve data and filtered historical flow and head data, while adhering to the characteristic curve mechanism. Head-flow characteristic curve calibration requires obtaining the pump's flow data during operation. This data refers to the flow data of a single pump during the long-term operation of a parallel pump set, when the flow rate is measurable or calculable. If historical data does not include or cannot be calculated for single-pump operation, the head-flow characteristic curve calibration process will not be performed.

[0063] The formula for calculating shaft power is:

[0064]

[0065] Among them, P shaft The pump's shaft power is expressed in watts (W); P meter R is the power rating of the pump, measured in watts (W). i denoted by , where is the loss rate of the pump's iron and copper losses; 'n' is the pump's rotational speed, in revolutions per minute (rpm). -1 f is the frequency of the pump electrode, in Hz; npair This refers to the number of magnetic pairs in the motor. If the meter collects current data, it needs to be converted to power data based on the pump's power supply type and voltage. Pump power data processing includes deducting copper and iron losses from the power output, and deducting mechanical losses. The deduction percentages for copper and iron losses can be estimated based on the pump's usage time and conditions (approximately 3% to 7%). Mechanical losses are calculated based on the pump's slip.

[0066] In pump power-flow calculation, a method based on pump efficiency prediction and pump frequency conversion characteristics is used. The bisection method is employed to solve complex nonlinear equations, achieving accurate flow calculation under different pump operating conditions. During the bisection method, the minimum and maximum flow points of the pump are used as the basis for setting the boundary values ​​for the bisection. The accuracy requirement for the bisection method is an error value of less than 10%. -3 .

[0067] For pump types with extreme shaft power values, different flow rates corresponding to the power are obtained by setting different bisection boundary conditions, generating different combinations of pump group branch flow rates. When the pump power is within a specific range, multiple different pump flow rates may be obtained, requiring further evaluation using pump group flow rate calculation. Pump group flow rate calculation and evaluation utilizes a system that includes pump group flow rate matching, head matching, and operational stability matching. A supervised method is used to output highly reliable pump group flow rate combination schemes and accurate pump group flow rate data. During the solution process, the solved flow rate is compared with the solved head and the real-time flow rate and head data collected by the pump. Simultaneously, the reliability of the pump operation under various flow rate solution schemes is analyzed. A fitness function (Eq.5) and flow rate, head, and stable operation penalty functions (Eq.6-8) are established to evaluate pump group flow rate matching, head matching, and pump operational stability, selecting the most suitable pump flow rate solution scheme and outputting the calculated flow rate data.

[0068]

[0069]

[0070]

[0071]

[0072] Where F cal fitness for calculating pump flow rate; p s Calculate the stability penalty function for the flow rate; p q Calculate the flow penalty function for the flow; p H Calculate the head penalty function for the flow rate; N is the total number of flow rate data; N p q represents the number of pumps operating in the pump set. i Here is the flow rate data of the pump unit at time i, in m³ / s.3 h -1 H j q represents the head of the j-th pump, in meters (m); ave This represents the average flow rate of the pump unit during this period, in cubic meters per second (m³). 3 h -1 ;q req This represents the flow rate demand of the pump system during this period, in cubic meters per second (m³). 3 h -1 H meter The value is the pump head gauge reading, in meters (m).

[0073] The accuracy of this method will be verified through two examples: single-pump and dual-pump.

[0074] Example 1: Single pump

[0075] For a single-pump operation in a circulating water project, the pump power-flow rate calculation was performed using the method of this invention. The pump model is KQSN800-M14J / 818, and its basic parameters, including a rotational speed of 740 rpm, are specified in the manufacturer's instructions. -1 The motor has a rated power of 800kW and characteristic flow rates and head points of 3143 / 50, 5238 / 43, and 6286 / 38 (m³). 3 h -1 / m).

[0076] Historical operating data of the pump was extracted and analyzed. Quartiles (one-quarter, two-quarter, and three-quarters) were calculated for the pump's flow rate, head, and power data. Interquartile ranges were calculated, and data points below the quarter-quarters and above 1.5 times the interquartile range were defined as outliers. If an outlier was found in the flow rate, head, or power data at any given moment, that moment's data was discarded, and the data was replaced with linear interpolation of data from the preceding and following moments.

[0077] Shaft power calculations were performed on the historical power data of the pump (e.g., Eq. 4). Considering the short service life of the pump, the ratio of copper loss to iron loss was set to 5%, and based on the number of magnetic pairs of the motor (4) and the rotational speed (740 rpm)... -1 The mechanical loss of the pump is set to 1.4% based on the frequency (50Hz).

[0078] The processed data was analyzed and found to have certain characteristics. The pump's head-flow characteristic curve was calibrated using the selected historical flow and head data. The calibrated head-flow characteristic curve is: H = -0.00000045*q^2 + 0.00047110*q + 53.01288327.

[0079] Using historical pump operating flow rate, head, and calculated shaft power data, the pump efficiency was calculated according to Eq.3. Combining the variable frequency, flow rate, and efficiency data with Eq.2, the final efficiency-flow characteristic curve was obtained as: η=-0.0000000196*q^2+0.0002569148*q. Since the pump's manufacturer's instructions did not provide the rated flow rate, we used the efficiency-flow characteristic curve to analyze the pump's performance at 6500m³ / h. 3 h -1 The pump achieves its highest efficiency at a given flow rate, which is then set as the pump's rated flow point.

[0080] By introducing a variable frequency variable, the head-flow characteristic curve and the efficiency-flow characteristic curve are integrated into a power-flow characteristic curve. The power-flow characteristic curve of the pump (with variable frequency variable) is as follows:

[0081] P=(-0.00000045*q^3+0.00047110*q^2*i+53.01288327*q*i^2) / (3600*(-0.0000000196*(q / i)^2+0.0002569148*(q / i))).

[0082] The collected real-time power data of the pump is used as input. The real-time power data is converted into real-time shaft power data using methods for processing historical power data, and then substituted into the pump's power-flow characteristic curve for calculation. Using the pump's rated flow rate as a dividing line, the pump flow rate corresponding to the shaft power is calculated for flow rates higher and lower than the rated flow rate. This embodiment only involves single-pump operation, therefore only two pump flow rate calculation schemes are generated. The calculated head corresponding to the calculated flow rate can be obtained by substituting the power-derived flow rate into the pump's head-flow characteristic.

[0083] Because the power-flow characteristic curve corresponding to the pump model in this single-pump embodiment exhibits nonlinear characteristics, two different pump flow values ​​may be obtained when the pump power is within a specific range. Therefore, during the solution process, the calculated flow rate and the calculated head are compared with the real-time flow rate and head data collected by the pump. Simultaneously, the reliability of the pump operation under the two flow rate solution schemes is analyzed. By establishing a fitness function (Eq.5) and flow rate, head, and stable operation penalty functions (Eq.6-8), the pump set flow rate matching, head matching, and pump operation stability are evaluated. The most suitable pump flow rate solution scheme is selected, and the calculated flow rate data is output.

[0084] The actual flow rate data and the flow rate data obtained through power-flow calculation in this single-pump embodiment are as follows: Figure 2As shown, the pump flow rate calculated using pump power according to this method has an accuracy rate of 81.25% (relative error less than 10%), and the relative deviation of the average pump flow rate within the selected data point time period (240h) is only 0.1%.

[0085] Example 2: Dual Pumps

[0086] For a circulating water project with multiple pumps operating in parallel, the pump power-flow rate is calculated using the flow rate calculation method for parallel variable frequency pump sets. The pump model is KQSN800-M14J / 818. The pump's basic parameters, including a speed of 740 rpm, are specified in the manufacturer's instructions. -1 The motor has a rated power of 800kW and characteristic flow rates and head points of 3143 / 50, 5238 / 43, and 6286 / 38 (m³). 3 h -1 / m).

[0087] Historical operating data of the pump was extracted and analyzed. Quartiles (one-quarter, two-quarter, and three-quarters) were calculated for the pump's flow rate, head, and power data. Interquartile ranges were calculated, and data points below the quarter-quarters and above 1.5 times the interquartile range were defined as outliers. If an outlier was found in the flow rate, head, and power data of different pumps at the same time, that data point was discarded and replaced with linear interpolation of data from preceding and following times.

[0088] Shaft power was calculated based on the pump's historical power data. Considering the pump's short service life, the ratio of copper loss to iron loss was set to 5%, and the mechanical loss of the pump was set to 1.4% based on the motor's magnetic pairs (4), rotational speed (740 rpm), and frequency (50 Hz).

[0089] The processed data, after analysis, exhibits certain characteristics. The head-flow characteristic curve of the pumps was calibrated using the selected historical flow and head data. The calibrated head-flow characteristic curve of pump 1 is as follows:

[0090] H=-0.00000045*q^2+0.00047110*q+53.01288327;

[0091] The head-flow characteristic curve of pump 2 after calibration is as follows:

[0092] H=-0.00000044*q^2+0.00047532*q+52.80976641.

[0093] The pump efficiency is calculated using historical flow rate, head, and calculated shaft power data. Eq.2 is then fitted using variable frequency, flow rate, and efficiency data to obtain the efficiency-flow rate characteristic curve. The efficiency-flow rate characteristic curve for pump 1 is as follows:

[0094] η=-0.0000000196*q^2+0.0002569148*q;

[0095] The efficiency-flow characteristic curve of pump 2 is as follows:

[0096] η=-0.0000000197*q^2+0.0002518003*q.

[0097] Since the pump's rated flow rate was not provided in the manufacturer's specifications, we used the pump's efficiency-flow characteristic curve to analyze the pump's performance at 6500 m³ / h. 3 h -1 The pump achieves its highest efficiency at a given flow rate, which is then set as the pump's rated flow point.

[0098] After introducing a variable frequency variable, the head-flow characteristic curve and the efficiency-flow characteristic curve are integrated into a power-flow characteristic curve. The power-flow characteristic curve (with variable frequency variable) of pump 1 is as follows:

[0099] P=(-0.00000045*q^3+0.00047110*q^2*i+53.01288327*q*i^2) / (3600*(-0.0000000196*(q / i)^2+0.0002569148*(q / i)));

[0100] The power-flow rate characteristic curve (with variable frequency) of pump 2 is as follows:

[0101] P=(-0.00000044*q^3+0.00047532*q^2*i+52.80976641*q*i^2) / (3600*(-0.0000000197*(q / i)^2+0.0002518003*(q / i))).

[0102] The real-time power data of the pump is collected as input, and the real-time power data is converted into real-time shaft power data using the method of processing previous historical power data. This data is then substituted into the pump's power-flow characteristic curve for calculation. Using the pump's rated flow rate as a dividing line, the pump flow rate corresponding to the shaft power is calculated for flow rates higher and lower than the rated flow rate. This embodiment involves dual-pump operation, with four pump flow rate calculation schemes. By substituting the power-derived flow rate into the pump's head-flow characteristic, the corresponding head can be obtained. The calculated flow rate and head are compared with the real-time collected flow rate and head data of the pump, and the reliability of the pump operation under the four flow rate calculation schemes is analyzed. Through evaluation of flow rate matching, head matching, and pump operating stability, the most suitable pump flow rate calculation scheme is selected, and the calculated flow rate data is output.

[0103] The actual flow rate data of the dual pumps and the flow rate data obtained through power-flow calculation in this embodiment are as follows: Figure 3 As shown in the figure. The results show that the pump flow rate calculated using pump power according to this method has an accuracy rate of 94.00% (relative error less than 10%), and the relative deviation of the average pump flow rate within the selected data point time period (720h) is only 0.8%.

[0104] As can be seen from the analysis of Examples 1 and 2, the present invention utilizes real-time pump power data, and calculates the possible flow values ​​of each variable frequency pump branch in the parallel pump group based on the pump's head-flow characteristics and efficiency-flow characteristics including variable frequency operation. Considering a pump group flow calculation and evaluation system that includes pump group flow matching, head matching, and operational stability, the pump group flow combination scheme with the highest reliability is selected, and the pump group flow data is output, achieving a good accuracy.

[0105] The above description of the embodiments is only for the purpose of helping to understand the method and core ideas of the present invention. It should be noted that those skilled in the art can make several improvements and modifications to the present invention without departing from the principles of the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Claims

1. A method for calculating the flow rate of a real-time parallel variable frequency pump set, characterized in that: Includes the following steps: S1 filters historical data on flow rate, head, and power. S2, use the filtered data to perform curve fitting to generate the pump's head-flow characteristic curve and efficiency-flow characteristic curve; or use the filtered data to perform curve calibration to update the existing head-flow characteristic curve and efficiency-flow characteristic curve. S3, read the power or current from the meter and then calculate the shaft power; S4, obtain the power-flow characteristic curve through the head-flow characteristic curve and the efficiency-flow characteristic curve, and perform power-flow calculation based on the shaft power; wherein, the formula for the head-flow characteristic curve is: ； The formula for the efficiency-flow characteristic curve is: ； Where H is the pump head; q is the pump flow rate; i is the variable frequency; η is the pump efficiency; α0, α1, and α2, as well as β0 and β1 are coefficients; The power-flow calculation includes solving the power-flow characteristic curve equation using the bisection method based on efficiency prediction and pump frequency conversion characteristics, and calculating the pump flow rate under different operating conditions, including low flow rate and frequency conversion conditions. S5, evaluate the pump group flow rate calculation results and output the pump group flow rate data. The pump group flow rate calculation evaluation is to evaluate several candidate flow rate solution schemes and use an evaluation system including pump group flow rate matching, head matching and stability matching, and output the pump group flow rate data using a supervised method.

2. The method for calculating the flow rate of a real-time parallel variable frequency pump set according to claim 1, characterized in that: The filtering of historical data on flow rate, head, and power includes: S11, perform statistical analysis on the historical operating data of parallel pump sets; S12, based on the user-set adjustable threshold, determine outliers in the historical operating data of the parallel pump group after statistical analysis; S13, remove or replace outliers; S14, extract characteristic data reflecting pump operation.

3. The method for calculating the flow rate of a real-time parallel variable frequency pump set according to claim 2, characterized in that: The outliers are data points that are below 1.5 times the interquartile range of the quarter quartile or above 1.5 times the interquartile range of the third quartile.

4. The method for calculating the flow rate of a real-time parallel variable frequency pump set according to claim 3, characterized in that: The curve fitting involves determining the coefficients α0, α1, α2, β0, and β1; the curve calibration involves updating the coefficients α0, α1, α2, β0, and β1.

5. The method for calculating the flow rate of a real-time parallel variable frequency pump set according to claim 1, characterized in that: The shaft power calculation includes converting the actual pump power to the pump shaft power based on a power conversion formula, which is: Among them, P shaft P is the shaft power of the pump. meter R is the power rating of the pump's meter. i Let n be the loss rate of the pump's iron and copper losses, n be the pump's rotational speed, and f be the frequency of the pump electrodes. pair The number of magnetic pairs of the motor; the loss rate is determined based on the pump's structure, service life, and operating status.

6. The method for calculating the flow rate of a real-time parallel variable frequency pump set according to claim 1, characterized in that: The fitness function and the flow rate, head, and stability penalty functions of the evaluation system are as follows: ； ； ； Where F cal For fitness; p s p is the stability penalty function; q The flow penalty function; p H Here, N is the head penalty function; N is the total number of flow rate data; N p q represents the number of pumps operating in the pump set. i Here is the flow rate data of the pump unit at time i; H j Let q be the head of the j-th pump; ave q represents the average flow rate of the pump unit during the current period; req H represents the flow rate demand of the pump system during the current time period. meter The measured value of the pump head gauge.

Images