Pump and pipe network system high-fidelity identification method and system based on active resistance modulation

Abstract

The invention discloses a high-fidelity identification method and system for a pump and a pipe network system based on active resistance modulation, and the method comprises the steps: firstly, manually setting at least two pipeline resistance states with obvious differences through adjusting devices such as valves, carrying out the frequency conversion operation of a pump in each state, and collecting a plurality of groups of pressure and frequency data; then, an integrated system physical model is constructed, and the model forcibly uses uniquely shared pump performance parameters and pipe network static lift parameters to synchronously fit all data clusters in different resistance states; and finally, by solving the global optimization problem of the model, identifying a real performance curve of the pump, the static lift of the pipe network and the independent resistance coefficient in each state at one time with high precision. According to the method, a flow sensor is not needed, the ill-conditioned problem of a passive data analysis method is solved, on-site in-situ calibration comparable with laboratory precision is achieved, and a reliable basis is provided for construction of a digital twin model and a virtual flowmeter and system health assessment.

Description

Technical Field

[0001] This invention relates to the fields of fluid machinery testing, system identification, and digital twin technology, and particularly to a high-fidelity identification method and system for pump and pipeline systems based on active resistance modulation. Background Technology

[0002] Variable frequency pump units are core equipment in modern fluid transport systems, and their energy efficiency and operational reliability are crucial to the overall system's economy and safety. To achieve optimized control and predictive maintenance of the pump units, accurately understanding the actual performance curves of the pumps over time due to wear, as well as the dynamic resistance characteristics of the system's piping network under different loads, is an indispensable prerequisite. However, in many engineering practices, due to limitations in cost, installation conditions, or subsequent maintenance, flow sensors are not installed in the systems, making traditional system modeling methods that rely on (Q,H) data pairs unusable.

[0003] To address this challenge, the primary approach currently available is the traditional laboratory testing method. This method yields highly accurate data by conducting comprehensive performance tests on the pump on a precisely controlled test bench. However, this method is costly, cannot be performed online, and fails to capture the "in-situ" characteristics of the pump's interaction with the system after installation in the actual piping network, as well as the characteristics of the piping network itself. Summary of the Invention

[0004] According to a first aspect of the present invention, a high-fidelity identification method for pump and pipeline systems based on active resistance modulation is provided, comprising the following steps: Actively set at least two differentiated pipeline stability resistance states; under each resistance state, control the pump to run at different frequencies and synchronously record the pump's outlet pressure and inlet pressure, thereby obtaining at least two structured data clusters corresponding to the at least two resistance states respectively; A system identification model is constructed. The definition of the system identification model is based on a unique, shared set of physical parameters P_params of the system to be identified. The system physical parameter set P_params includes: a set of pump performance parameters P_pump used to describe the common performance characteristics of the pump under all resistance states, and a set of pipeline model parameters P_sys used to describe the characteristics of the pipeline network. A global objective function is constructed with the system physical parameter set P_params as the optimization variable. The objective function is solved by numerical optimization algorithm, and all parameters in the system physical parameter set P_params are identified at one time. The constraint condition of the objective function is that all structured data clusters must be fitted simultaneously using the same set of pump performance parameter set P_pump and the same reference static head parameter H_static_ref.

[0005] Further, the mathematical structure of the pump performance parameter set P_pump is parameterized based on a physical information model that nonlinearly corrects the ideal affine law; the pump performance parameter set P_pump specifically includes: a set of reference pump parameters used to define the performance curve of the pump at a certain reference operating state; and a set of correction parameters used to parameterize one or more nonlinear functions, thereby describing how the performance curve evolves as the operating frequency changes.

[0006] Further, the system model parameter set P_sys includes a unique, shared reference static head parameter H_static_ref and an independent pipe dynamic resistance coefficient corresponding to each resistance state.

[0007] Further, a global objective function is constructed with the system physical parameter set P_params as the optimization variable, and the objective function is solved by a numerical optimization algorithm, specifically including: Based on the physical balance relationship, an algebraic expression about the unknown flow rate is established; By algebraic elimination, a global objective function is constructed, which aims to minimize the difference between a theoretically predicted physical quantity and an observable physical quantity; A numerical optimization algorithm is used to solve the objective function, with the union of all collected data clusters as input.

[0008] Further, the stable resistance state of the pipe is achieved by artificially intervening in the pipe network regulating device, which includes but is not limited to adjusting the valve opening, switching the parallel or series pipe branches.

[0009] Further, the differentiated pipe stable resistance state means that at the same pump operating frequency, the pump outlet pressure corresponding to different resistance states should be significantly different.

[0010] Further, the method is executed under the condition that no flow sensor is deployed, and no clustering algorithm is performed on the collected data.

[0011] Further, it also includes the following steps: instantiating the identified system physical parameter set into a general digital twin model; the digital twin model is deployed in a computing device to realize at least one of the following functions: virtual sensing, energy efficiency evaluation, fault warning, performance degradation evaluation, optimal set point optimization, energy saving simulation, global optimal scheduling of pump group, equipment asset value evaluation, supplier equipment performance benchmarking.

[0012] According to the second aspect of the embodiment of the present application, a high-fidelity identification system for a pump and pipe network system based on active resistance modulation is provided, which includes: a resistance modulation module for actively setting at least two differentiated pipe stable resistance states; in each resistance state, the pump is controlled to run at different frequencies, and the outlet pressure and the inlet pressure of the pump are recorded synchronously, so as to obtain at least two structured data clusters corresponding to the at least two resistance states respectively; a construction module for constructing a system identification model, a definition of the system identification model being based on a unique, shared, to-be-identified system physical parameter set P_params, the system physical parameter set P_params including: a pump performance parameter set P_pump for describing common performance characteristics of the pump in all resistance states, and a pipe network model parameter set P_sys for describing pipe network characteristics; an identification module for constructing a global objective function with the system physical parameter set P_params as optimization variables, solving the objective function by using a numerical optimization algorithm, and identifying all parameters in the system physical parameter set P_params at one time, a constraint condition of the objective function being that the same set of pump performance parameter set P_pump and the same reference static head parameter H_static_ref must be used to fit all structured data clusters.

[0013] According to a third aspect of the embodiments of the present application, a computer readable medium having non-volatile program code executable by a processor is provided, the program code causing the processor to execute the first aspect of the pump and pipe network system high-fidelity identification method based on active resistance modulation.

[0014] The pump and pipe network system high-fidelity identification method and system based on active resistance modulation according to the embodiments of the present application have the following beneficial effects: The robustness is fundamentally improved: a set of geometrically separated pipe curves (at least two) is actively constructed, which provides strong geometric constraints for optimization solving, effectively avoids the solver from falling into local optimum or producing ambiguous solutions in a single fuzzy "data cloud", and makes the identification result no longer sensitive to the initial guess value.

[0015] Perfect parameter decoupling is achieved: it is almost impossible for any set of incorrect pump parameters to pass through multiple pipe curves located at different positions defined by multiple sets of structured data at the same time while keeping H_static_ref unchanged. This cross-validation mechanism forces the algorithm to find the only set of true physical parameters.

[0016] The data demand is dramatically reduced: since the information content of the data is actively enhanced, the present method no longer needs the massive historical data required by passive analysis. In each resistance state, only a small amount (for example, 5 to 9) of variable frequency points are collected, which is enough to achieve high-precision parameter identification, greatly shortening the test time and data processing cost.

[0017] The method is simple to operate and low in cost. The method only needs to perform several simple on-off operations on the existing valve on site, does not need to add any hardware or complex clustering algorithm, and is a new paradigm of "in-situ accurate calibration" for replacing expensive laboratory tests.

[0018] It is to be understood that both the foregoing general description and the following detailed description are exemplary and intended to provide further explanation of the subject technology claimed. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 A flow chart of a pump and pipe network system high-fidelity identification method based on active resistance modulation according to an embodiment of the present application.

[0020] Figure 2 A top structure diagram of a pump and pipe network system high-fidelity identification system based on active resistance modulation according to an embodiment of the present application. DETAILED DESCRIPTION

[0021] The preferred embodiments of the present application will be described in detail with reference to the accompanying drawings, and the present application will be further described.

[0022] First, a pump and pipe network system high-fidelity identification method based on active resistance modulation according to an embodiment of the present application will be described. Figure 1 The pump and pipe network system high-fidelity identification method based on active resistance modulation according to an embodiment of the present application is used to accurately identify the performance of the pump and the resistance characteristics of the pipe network system under the condition that no flow sensor is deployed, and has a wide range of application scenarios.

[0023] As shown in Figure 1 , the pump and pipe network system high-fidelity identification method based on active resistance modulation according to an embodiment of the present application has the following steps: In S1, at least two differentiated pipe stable resistance states are actively set; in each resistance state, the pump is controlled to operate at different frequencies, and the outlet pressure Hout and the inlet pressure Hin of the pump are recorded synchronously, so as to obtain at least two structured data clusters corresponding to at least two resistance states respectively.

[0024] Further, in the embodiment, the stable resistance state of the pipe is achieved by artificially intervening in the pipe network adjusting device, which includes but is not limited to adjusting the valve opening, switching the parallel or series pipe branch.

[0025] Further, in the embodiment, the differentiated pipe stable resistance state refers to that the pump outlet pressure corresponding to different resistance states should be significantly different under the same pump operating frequency.

[0026] In S2, a system identification model is constructed, the definition of which is based on a unique, shared, set of system physical parameters P_params, including a set of pump performance parameters P_pump describing the common performance characteristics of the pump under all resistance conditions, and a set of system model parameters P_sys describing the characteristics of the pipe network.

[0027] Further, in the present embodiment, the mathematical structure of the set of pump performance parameters P_pump is parameterized based on a physical information function (PiiF) that nonlinearly corrects the ideal affine law; the set of pump performance parameters P_pump specifically includes: a set of reference pump parameters used to define the performance curve of the pump under a certain reference operating condition; and a set of correction parameters used to parameterize one or more nonlinear functions, thereby describing how the performance curve evolves with changes in operating frequency.

[0028] Further, in the present embodiment, the set of system model parameters P_sys includes a unique, shared reference static head parameter H_static_ref and independent pipe dynamic resistance coefficients {k1, k2,..., kP} corresponding to each resistance condition.

[0029] In S3, a global objective function is constructed with the set of system physical parameters P_params as optimization variables, and the objective function is solved by a numerical optimization algorithm, thereby identifying all parameters in the set of system physical parameters P_params simultaneously, with the constraint condition being that the same set of pump performance parameters P_pump and the same reference static head parameter H_static_ref must be used to fit all structured data clusters simultaneously. The global objective function is constructed with the set of system physical parameters P_params as optimization variables, and the objective function is solved by a numerical optimization algorithm, specifically including: an algebraic expression about unknown flow rate is established based on physical equilibrium relations; a global objective function is constructed by algebraic elimination, which aims to minimize the difference between a theoretically predicted physical quantity and an observable physical quantity; a numerical optimization algorithm is used to solve the objective function with the union of all collected data clusters as input, thereby identifying the unique set of pump performance parameters P_pump, the unique reference static head H_static_ref, and the independent pipe resistance coefficients corresponding to each resistance condition simultaneously.

[0030] Further, in the present embodiment, the method is executed under the condition that no flow sensor is deployed, and no clustering algorithm is performed on the collected data.

[0031] Further, in the present embodiment, the pump and pipe network system high-fidelity identification method based on active resistance modulation of the present embodiment further comprises the following steps: instantiating the identified system physical parameter set as a general digital twin model; the digital twin model is deployed in a computing device to realize at least one of the following functions: Virtual sensing: according to the real-time input of the operating frequency (Fi), the pump outlet pressure (Hout_i) and the pump inlet pressure (Hin_i), the virtual flow value is output; Energy efficiency evaluation: based on the virtual flow value and real-time power data, the real-time efficiency of the system is calculated and output; Fault early warning: by monitoring the abnormal change of the virtual flow value or the system efficiency, an early warning signal of leakage or blockage is generated; Performance degradation evaluation: compare the pump performance parameter sets identified at different time points by the method to quantitatively evaluate the performance degradation of the pump; Optimal set point optimization: receive a target flow or target pressure instruction, based on the identified pump performance parameter set and pipe network model parameter set, reverse simulation calculation and output the optimal operating frequency setting value required to achieve the target; Energy-saving reconstruction simulation: virtually match the identified pipe network model parameter set (P_sys) with the known performance model of one or more candidate replacement pumps, and predict and output the expected operating point and energy-saving benefit of each candidate pump in the real pipe network system.

[0032] Pump group global optimal scheduling: based on the identified individualized performance parameter set (P_pump_n) of each pump and the shared pipe network model parameter set (P_sys), according to the real-time total demand, solve a global optimization problem, calculate and output the current most energy-saving pump start-stop combination and operating frequency setting value of each pump; Device asset value evaluation: compare the parameter set (P_pump) representing the current real performance state of the pump identified by the method with the baseline parameter set of a new pump of the same model, generate a quantitative performance state report for evaluating the remaining asset value of the device or guiding its remanufacturing process; Supplier device performance benchmarking: perform the method on multiple devices of the same model from different suppliers, and compare the identified performance parameter sets (P_pump) of each device with the nominal performance parameter set to evaluate the consistency, accuracy and reliability of the performance of each supplier's device.

[0033] Based on the above-described pump and pipe network system high-fidelity identification method based on active resistance modulation of the present embodiment, specific practical application examples are as follows:

Specific implementation with P=2 as an example

[0034] Step 1: Active data collection and preparation a) Collect the first data cluster (valve fully open): Open the pump outlet main line regulating valve completely. Control the frequency converter so that the pump runs stably at a series of different frequency points (for example, scan from 50 Hz to 20 Hz with a step of 5 Hz, a total of 7 data points) that can fully represent its running characteristics in this resistance state, and record the {Fi, Hout_i, Hin_i} data respectively.

[0035] b) Collect the second data cluster (high resistance / small flow condition): The operator turns the valve to the "closed" direction by hand or experience and fixes it at an appropriate angle to establish a high resistance pipeline state that is significantly different from the previous one. This method does not rely on whether there is an accurate opening indication or scale on the valve.

[0036] c) Verification and confirmation: The operator can confirm the effectiveness of the two experiments through a simple online check. At any one same test frequency (for example, 50 Hz), the pump outlet pressure Hout_2 measured by the second data cluster (high resistance) should be higher than the pressure Hout_1 measured by the first data cluster (low resistance). This real-time pressure feedback is the only objective standard for judging whether the resistance state has been "significantly changed". By "significantly" higher, it means that the pressure difference ΔH = Hout_2 - Hout_1 should reach a magnitude that is sufficient to make the two sets of pipeline dynamic resistance coefficients k1 and k2 calculated through the subsequent steps of the invention have numerical distinguishability, and can ensure the numerical stability when solving the global optimization problem. In typical industrial applications, it is recommended that Hout_2 be at least 10% higher than Hout_1.

[0037] d) Known geometric parameters: Z: The fixed vertical height difference between the installation position of the pump outlet pressure sensor and the installation position of the inlet pressure sensor, defined as Z = z_out - z_in. For example, the outlet height is 0.5 meters, then Z = 0.5 meters.

[0038] D1, D2: The nominal pipe diameters of the pump's inlet and outlet, for example, the inlet D1 is DN200 and the outlet D2 is DN150.

[0039] Pre-calculate the kinetic energy correction coefficient k0: According to the standard fluid mechanics formula k0 = (1 / A2² - 1 / A1²) / (2g), where A1 and A2 are the inlet and outlet cross-sectional areas, and g is the acceleration of gravity. This constant k0 can be accurately calculated in advance.

[0040] Where A1 and A2 are the inlet and outlet cross-sectional areas calculated from pipe diameters D1 and D2, and g is the acceleration due to gravity.

[0041] Step two: Integrated mathematical model construction Define the set of physical parameters P_params to be solved, which includes: Pump performance parameters P_pump: This parameter set includes two major components, in the present invention, all parameters of these two parts are solved as unknowns in the same optimization problem: Reference pump parameters {A, B, C}: used to define the quadratic polynomial performance curve H_base = A*Q_base² + B*Q_base + C of the pump at the reference frequency F_base.

[0042] PiiF core correction parameters {a, b, c, d, e, f}: used to parameterize a set of nonlinear scaling factor functions kh(r) and kq(r). In the present invention, these coefficients are identified synchronously with the reference pump parameters.

[0043] System model parameters P_sys: {H_static_ref, k1, k2} (total of 3).

[0044] Total of 12 unknown parameters.

[0045] Step three: Optimization problem transformation and solution Considering that in actual pipe network systems, due to factors such as upstream water source level fluctuations, the actual static lift of the system may not be a fixed constant. In order to more accurately model such systems, the present invention introduces the concept of a "reference static lift" H_static_ref. The instantaneous effective static lift of the system is modeled as a function of this reference value and the real-time measured pump inlet pressure Hin_i.

[0046] For any point i in all the collected data points, its measured value is (Fi, Hout_i, Hin_i).

[0047] Establish three core physical equations: A. Pump equation (PiiF model): describes how the true total head H_true of the pump is determined by the flow rate Q and the frequency F. In a preferred embodiment, the pump equation can be represented as: H_true = [kh(r)]² * [A*(Q / kq(r))² + B*(Q / kq(r)) + C]; where r = Fi / F_base is the ratio of the current frequency to the base frequency. kh(r) and kq(r) are non-linear scaling functions defined by the PiiF correction parameters {a, b, c, d, e, f}. In a preferred embodiment, these two functions can be parameterized as low-order polynomials: kh(r) = a*r² + b*r + c; kq(r) = d*r² + e*r + f; The parameter set {A, B, C, a, b, c, d, e, f} is the set of pump performance parameters P_pump to be solved for.

[0048] B. Measurement equation: describes how to back-calculate H_true from field measurements. According to standard fluid mechanics definitions: H_true = (Hout_i - Hin_i) + Z + k0*Q_i²; C. Pipe network equation: describes the demand of the pipe network on the head.

[0049] H_true = (H_static_ref - Hin_i) + kj*Q_i²; Back-calculate flow expression (based on B = C): Let the measurement equation equal the pipe network equation, and solve for the algebraic expression of the unknown flow Qi: Q_i_expr = sqrt((Hout_i + Z - H_static_ref) / (k_j(i) - k0)); Here k_j(i) is clearly labeled as either k1 or k2 depending on which cluster of data points i belongs to.

[0050] Construct and solve the objective function: through algebraic substitution and elimination, construct a global objective function that aims to minimize the difference between the theoretically predicted physical quantity and the observable physical quantity. In a preferred embodiment, this objective function can be expressed as: Minimize L(P_params) = Σ_i[(PiiF_Model(Q_i_expr, Fi) - k0*Q_i_expr²) - ((Hout_i - Hin_i) + Z)]²; Using non-linear least squares and applying reasonable physical boundary constraints, solve for all 12 unknown parameters in the system physical parameter set P_params.

[0051] Solving: Solve the objective function in a programming environment such as Python using a nonlinear least squares optimization library (e.g. scipy.optimize.least_squares). To ensure stable convergence of the optimization process and obtain physically meaningful solutions, reasonable initial guess values and boundary constraints must be set for the unknown parameters to be solved. In a specific embodiment, these constraints can at least include: Constraints on pump reference parameters: A must be negative (A < 0) to ensure that the pump's H-Q performance curve is convex upwards.

[0052] C (zero-flow head) must be positive (C > 0).

[0053] Constraints on pipe network parameters: The reference static head H_static_ref must be positive.

[0054] All pipe dynamic resistance coefficients k1, k2,..., kM must be positive (kj > 0) to ensure that the pipe resistance curve is monotonically increasing.

[0055] Constraints on PiiF correction parameters: Based on physical experience, soft constraints can be imposed on the behavior of kh(r) and kq(r) near r = 1, for example, kh(1) = 1, kq(1) = 1, to ensure smooth evolution near the reference frequency.

[0056] By imposing these physical prior-based constraints, the optimization algorithm is effectively guided to search within the space of physically feasible solutions until the algorithm converges, ultimately obtaining the optimal estimated values of all parameters.

[0057] Testing with two sets (P = 2) of significantly different resistance states is the minimum but sufficiently effective implementation of the invention. In each resistance state, a number of (e.g. 5 to 9) variable frequency data points within the pump's effective operating frequency range are collected, which can usually construct an overdetermined equation system to ensure the theoretical uniqueness of the solution.

[0058] In some complex working conditions with extremely high identification accuracy requirements or large system noise, the skilled person can set and measure a third (or more) independent resistance state to further enhance the constraint strength of the optimization problem and the robustness of the solution. This "on-demand increase in testing" capability is a major advantage of the invention, which provides a deterministic engineering means to overcome the potential "ill-conditioned data" (i.e. insufficient resistance state difference) problem, and transforms the entire calibration process from a static process to a dynamic, verifiable closed loop, thereby transforming the passive approach to data quality in the passive approach into a deterministic and guaranteed result in the present approach.

[0059] As described above, in the pump and pipe network system high-fidelity identification method based on active resistance modulation according to the embodiment of the application, the following beneficial effects are achieved: The robustness is fundamentally improved: a set of at least two geometrically separated pipe curves is actively constructed, which provides strong geometric constraints for optimization solution, effectively avoids the solver from falling into local optimum or producing ambiguous solutions in a single fuzzy "data cloud", and makes the identification result no longer sensitive to the initial guess value.

[0060] The parameters are perfectly decoupled: it is almost impossible for any set of incorrect pump parameters to simultaneously and perfectly pass through multiple pipe curves at different positions defined by multiple sets of structured data while keeping H_static_ref unchanged. This cross-validation mechanism forces the algorithm to find the only set of real physical parameters.

[0061] The data requirement is dramatically reduced: since the information content of the data is actively enhanced, the method no longer needs the massive historical data required by passive analysis. A small number (for example, 5 to 9) of frequency points are collected at each resistance state, which is sufficient to achieve high-precision parameter identification, greatly shortening the test time and data processing cost.

[0062] The operation is simple and the cost is low: the method only needs to perform several simple on-off operations on the existing valves on site, without the need to add any hardware or complex clustering algorithm, and is a new paradigm for "in-situ accurate calibration" that replaces expensive laboratory tests with high cost performance.

[0063] The above is described in combination with the accompanying Figure 1 A pump and pipe network system high-fidelity identification method based on active resistance modulation according to the embodiment of the application is described. Further, the application can also be applied to a pump and pipe network system high-fidelity identification system based on active resistance modulation.

[0064] As Figure 2 shown, according to the second aspect of the embodiment of the application, a pump and pipe network system high-fidelity identification system based on active resistance modulation is provided, which comprises: A resistance modulation module 100 is configured to actively set at least two differentiated pipe stable resistance states; in each resistance state, the pump is controlled to operate at different frequencies, and the outlet pressure and the inlet pressure of the pump are recorded synchronously, so as to obtain at least two structured data clusters corresponding to at least two resistance states respectively; The constructing module 200 is configured to construct a system identification model, and a definition of the system identification model is based on a unique, shared, and to-be-identified system physical parameter set P_params, which includes a pump performance parameter set P_pump for describing common performance characteristics of the pump under all resistance states, and a pipe network model parameter set P_sys for describing pipe network characteristics. The identifying module 300 is configured to construct a global objective function with the system physical parameter set P_params as optimization variables, and solve the objective function by using a numerical optimization algorithm, so as to identify all parameters in the system physical parameter set P_params at one time, and a constraint condition of the objective function is that the same set of pump performance parameter set P_pump and the same reference static head parameter H_static_ref must be used to fit all structured data clusters.

[0065] According to a third aspect of the embodiment of the present application, a computer readable medium with non-volatile program code executable by a processor is provided, and the program code causes the processor to execute the first aspect of the active resistance modulation based pump and pipe network system high-fidelity identification method.

[0066] Where the readable storage medium can be a computer-readable storage medium, also can be a communication medium. The communication medium includes any medium that facilitates transfer of a computer program from one place to another. The computer storage medium can be any available medium that can be accessed by a general purpose or special purpose computer. For example, the readable storage medium is coupled to the processor, so that the processor can read information from the readable storage medium, and can write information to the readable storage medium. Of course, the readable storage medium can also be an integral part of the processor. The processor and the readable storage medium can be located in the application-specific integrated circuit. In addition, the application-specific integrated circuit can be located in the device. Of course, the processor and the readable storage medium can also exist as discrete components in the communication device. The readable storage medium can be read-only memory (ROM), random access memory (RAM), CD-ROM, magnetic tape, floppy disk and optical data storage device, etc. The present application also provides a program product, which includes execution instructions stored in a readable storage medium. The at least one processor of the device can read the execution instructions from the readable storage medium, and the at least one processor executes the execution instructions so that the device implements the above-mentioned various embodiments of the active resistance modulation-based pump and pipe network system high-fidelity identification method. In the above-mentioned embodiments of the device, it should be understood that the processor can be a central processing unit (English: Central Processing Unit, for short: CPU), and can also be other general-purpose processors, digital signal processors (English: Digital Signal Processor, for short: DSP), application-specific integrated circuits (English: Application Specific Integrated Circuit, for short: ASIC), etc. The general-purpose processor can be a microprocessor, or the processor can be any conventional processor, etc. The steps of the method disclosed in combination with the present application can be directly embodied as execution completed by a hardware processor, or executed by a combination of hardware and software modules in the processor.

[0067] It should be noted that in the present specification, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or includes elements inherent to such a process, method, article or device. Without more limitations, the elements defined by the statement "comprising" do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0068] While the application has been described in detail by reference to preferred embodiments thereof, it should be recognized that the description set forth herein is by way of example and that modifications of the procedures described can be employed without departing from the scope of the application. Accordingly, the scope of the application should be determined by the appended claims and equivalents thereof.

Claims

1. A high-fidelity identification method for pump and pipe network systems based on active resistance modulation, characterized by, The method comprises the following steps: actively setting at least two differentiated steady resistance states of the pipeline; in each of the resistance states, controlling the pump to operate at different frequencies, and synchronously recording the outlet pressure and the inlet pressure of the pump, so as to obtain at least two structured data clusters corresponding to the at least two resistance states respectively; constructing a system identification model, the definition of the system identification model being based on a unique, shared, set of system physical parameters P_params, the set of system physical parameters P_params comprising: a set of pump performance parameters P_pump for describing the common performance characteristics of the pump in all resistance states, and a set of pipeline model parameters P_sys for describing the characteristics of the pipeline network; constructing a global objective function taking the set of system physical parameters P_params as optimization variables, and solving the objective function by using a numerical optimization algorithm, so as to identify all the parameters in the set of system physical parameters P_params simultaneously, the constraint condition of the objective function being that the same set of pump performance parameters P_pump and the same reference static head parameter H_static_ref must be used to fit all the structured data clusters.

2. The method of claim 1, wherein, The mathematical structure of the set of pump performance parameters P_pump is parameterized based on a physical information model that nonlinearly corrects the ideal affine law; the set of pump performance parameters P_pump specifically comprises: a set of reference pump parameters for defining the performance curve of the pump in a certain reference operating state; and a set of correction parameters for parameterizing one or more nonlinear functions, so as to describe how the performance curve evolves with the change of the operating frequency.

3. The method of claim 1 or 2, wherein the method further comprises: The set of pipeline model parameters P_sys comprises a unique, shared reference static head parameter H_static_ref and an independent pipeline dynamic resistance coefficient corresponding to each of the resistance states.

4. The method of claim 1, wherein the method is characterized by, constructing a global objective function by using an algebraic elimination method, the global objective function aiming to minimize the difference between a theoretically predicted physical quantity and an observable physical quantity; using a numerical optimization algorithm to solve the objective function, with the union set of all the collected data clusters being taken as the input. The steady resistance states of the pipeline are achieved by artificially intervening in the pipeline regulating device, which includes but is not limited to adjusting the opening of the valve, switching the parallel or series pipeline branches. The differentiated steady resistance states of the pipeline refer to the fact that the pump outlet pressures corresponding to different resistance states should be significantly different under the same pump operating frequency.

5. The method of claim 1, wherein the method further comprises: determining a flow rate of the pump based on the determined pressure difference and the determined flow rate of the pump. The method is performed under the condition that no flow sensor is deployed, and no clustering algorithm is needed to be performed on the collected data.

6. The method of claim 1, wherein the method further comprises: ​ 7. The method of claim 1, wherein the method further comprises: determining a flow rate of the pump based on the determined pressure difference and the determined flow rate of the pump. ​ 8. The method of claim 1, wherein the method further comprises: determining a flow rate of the pump based on the determined pressure difference and the determined flow rate of the pump. Further comprising the steps of: instantiating the identified system physical parameter set as a common digital twin model; deploying the digital twin model in a computing device for implementing at least one of the following functions: virtual sensing, energy efficiency assessment, fault early warning, performance degradation assessment, optimal set point optimization, energy saving retrofit simulation, global optimal scheduling of pump groups, equipment asset value assessment, supplier equipment performance benchmarking.

9. A high-fidelity identification system for pump and piping systems based on active resistance modulation, characterized by, Comprising: a resistance modulation module for actively setting at least two differentiated pipe steady resistance states; in each of the resistance states, controlling the pump to run at different frequencies and synchronously recording the outlet pressure and inlet pressure of the pump, thereby obtaining at least two structured data clusters corresponding to the at least two resistance states respectively; a construction module for constructing a system identification model, the definition of the system identification model being based on a unique, shared, system physical parameter set P_params to be identified, the system physical parameter set P_params comprising: a pump performance parameter set P_pump for describing the common performance characteristics of the pump in all resistance states, and a pipe network model parameter set P_sys for describing the characteristics of the pipe network; an identification module for constructing a global objective function with the system physical parameter set P_params as the optimization variable, solving the objective function by a numerical optimization algorithm, and synchronously identifying all parameters in the system physical parameter set P_params at one time, the constraint condition of the objective function being: all the structured data clusters must be fitted simultaneously using the same set of pump performance parameter set P_pump and the same reference static head parameter H_static_ref.

10. A computer readable medium having non-transitory program code executable by a processor, the program code comprising instructions for: The program code causes the processor to run the high-fidelity identification method of a pump and pipe network system based on active resistance modulation according to any one of claims 1-8.

Images