EP4587714A1

EP4587714A1 - Pump control system and method for controlling a pump with automatic pump model calibration

Info

Publication number: EP4587714A1
Application number: EP23764866.2A
Authority: EP
Inventors: Torben Nielsen
Original assignee: Grundfos Holdings AS
Current assignee: Grundfos Holdings AS
Priority date: 2022-09-13
Filing date: 2023-08-31
Publication date: 2025-07-23
Also published as: CN119866408A; US20260085688A1; WO2024056395A1

Abstract

The present disclosure relates to a method for controlling a pump, wherein the method comprises the following steps: - running (501) a pump at a set pump speed (ω) based on an operating value model function (P_ω(q)) corresponding to the set pump speed (ω); - determining (503) an actual operating value (P) at the set pump speed (ω); - comparing (505a, 505b) the actual operating value (P) with a maximum (P_max) of the operating value model function (P_ω(q)) and/or a minimum (P_min) of the operating value model function (P_ω(q)); and - automatically updating (507b) the operating value model function (P_ω(q)) to a higher new operating value model function (P_ω,up(q)) if the actual operating value (P) is higher than the maximum (P_max) of the operating value model function (P_ω(q)), and/or automatically updating (507a) the operating value model function (P_ω(q)) to a lower new operating value model function (P_ω,down(q)) if the actual operating value (P) is lower than the minimum (P_min) of the operating value model function (P_ω(q)).

Description

Title: Pump control system and method for controlling a pump with automatic pump model calibration Description TECHNICAL FIELD [01] The present disclosure is directed to a pump control system and a method for controlling a pump with automatic pump model calibra- tion. Furthermore, the present disclosure is directed to a pump assembly with such a pump control system or being controlled by such a meth- od. BACKGROUND [02] Pumps may be used as circulator pumps for circulating water or another liquid through a pipe system, e.g. a heating or cooling system. Typically, circulator pumps may be operated in different operating modes. For example, a circulator pump may be operated to provide a constant output pressure, i.e. in constant pressure mode. Another oper- ating mode is typically “proportional pressure mode”, wherein the op- erating points of the circulator pump follow a linear line in a pressure- flow-diagram, a so-called pq-diagram. It is a general goal of pump controlling to control the pump in such a way that the pump runs pre- cisely at a desired operating point, e.g. in the pq-diagram. Usually, the output pressure P and the flow q are measured and the measured val- ues are fed back in a closed-loop control, so that the pump runs at the desired operating point. [03] However, flow sensors are expensive and often not available for pump control. If no measured flow value is available for pump control, it is known that pump control systems and methods use the affinity laws to estimate the current operating point based on other operating val- ues, e.g. the actual power consumption P, the motor current I and/or motor speed of the pump ω. Using such operating values for an indirect estimation of the operating point, however, requires a precise model of the pump characteristic, e.g. a model function in the Pq-diagram. [04] Pump manufacturers typically determine the pump characteris- tics in a laboratory and/or field tests for certain pump types and pro- vide data sets with a plurality of model parameter sets indicative of operating value model functions that describe the pump characteris- tics of a certain model type. [05] Typically, a certain pump type comes in a plurality of variants and a large amount of combinations of customised options. The true pump characteristics may differ slightly among the plurality of variants, whereas the same model function is used by the pump control system to describe the pump characteristic. So, the true pump characteristic may differ from the assumed modelled pump characteristic. Manufac- turing tolerances add further differences between the actual pump characteristic and the assumed model pump characteristic. The true pump characteristic may even depend on the pipe system that the pump is connected to, so that there is an inherent uncertainty about the modelled pump characteristic that a pump manufacturer is able to provide with a pump control system. Last but not least, the pump char- acteristic may change dynamically over time due to wear and tear of mechanical and/or electrical components, so that an older pump be- haves differently than a brand new pump. [06] US 7,945,411 B1, DE 10 2009 050 083 B4, US 9,897,084 B1 and US 9,470,217 B1 describe control systems, wherein either the pump is forced to run at a known fix operation point for calibration, or wherein an actual flow is measured or estimated to determine the actual oper- ating point. [07] Therefore, it is an object of the present invention to provide an improved pump control system and an improved method for control- ling a pump, wherein a pump model is used to control the pump with- out a measured flow or estimated flow and without forcing the pump into known operational points. SUMMARY [08] According to a first aspect of the present disclosure, a pump control system is provided comprising: - a storage element having stored a plurality of model parameter sets for a plurality of pump speeds, wherein each model param- eter set is indicative of an operating value model function corre- sponding to a pump speed; - control electronics configured to command the pump to run at a set pump speed based on the operating the value model function corresponding to the set pump speed, wherein a corre- sponding model parameter set of the stored model parameter sets is retrieved from the storage element; and - model updating electronics, wherein the model updating elec- tronics are configured to: - determine an actual operating value at the set speed; - compare the actual operating value with a maximum of the operating value model function and/or a minimum of the oper- ating value model function; and - automatically determine an updated model parameter set if the actual operating value is higher than the maximum of the oper- ating value model function or if the operating value is lower than the minimum of the operating value model function, wherein the updated model parameter set is indicative of a higher new op- erating value model function if the actual operating value is higher than the maximum of the operating value model func- tion, and wherein the updated model parameter set is indicative of a lower new operating value model function if the actual op- erating value is lower than the minimum of the operating value model function, wherein the storage element is further configured to store the updated model parameter set. [09] The operating value may be the power consumption of the pump, a motor current, or another electrical performance value that is readily available to be determined. It should be noted that the updat- ing or calibrating of the model parameter set does not require a meas- ured flow value or an estimated flow value. Furthermore, the pump is not forced to run at a fixed known operating point. The updating or calibrating of the model parameter sets may be performed dynamical- ly and automatically during normal pump operation. The updating or calibrating can be performed continuously, regularly and/or sporadi- cally over the lifetime of the pump. [10] The inventive idea uses the fact that normal operation of a circu- lator pump is often at or around a minimum or maximum operating val- ue. If the determined actual operating value is between a minimum and a maximum of the operating value model function, it may be im- possible to evaluate the model function without having a measured or estimated flow to determine the exact operation point. Outside of the operating value range, i.e. above a maximum of the operating value model function and below a minimum of the operating value model function, a deviation between the actual operating value and the op- erating value model function cannot be explained by the uncertainty in flow. In this case, it can be assumed that the operating value model function does not reflect the true pump characteristic. Therefore, the operating value model function is updated or calibrated if the actual operating value is higher than the maximum of the operating value model function and if the actual operating value is lower than the min- imum of the operating value model function. [11] Optionally, the model updating electronics may be configured to shift or scale the operating value model function upward to the higher new operating value model function and/or to shift or scale the operating value model function downward to a lower new operating value model function. In this case, the shape of the operating value model function may remain unamended. Alternatively, or in addition, the model updating electronics may be configured to apply a correc- tion function to tweak shape of the operating value model function. For example, in case of an upward updating, the model function may be lifted only in a predefined range about the maximum of the model function. Analogously, in case of a downward updating, the model function may be lowered in a predefined range about the minimum of the model function. [12] Optionally, a shift amount and/or a scale factor may be applied, wherein the shift amount and/or the scale factor is a predetermined constant or depends in a predetermined way on set pump speed. In case of a constant shift amount and/or scale factor, the updating logic is very simple and all pump speeds are given the same weight. If, how- ever, the accuracy of the model function is known to be better for cer- tain pump speeds, it may be advantageous to apply a shift amount and/or a scale factor that depends in a predetermined way on the set pump speed. [13] Optionally, the shift amount and/or the scale factor may de- pend on a difference and/or ratio between the actual operating value and the maximum or the minimum, respectively, of the model function. Thereby, the model function is fitted to the determined actual operat- ing value. [14] Optionally, the operating value model function may be a model curve of the operating value in dependence of a pump flow within a predetermined flow range. Preferably, the model curve comprises a single well-defined minimum and maximum within the predetermined flow range. [15] Optionally, the higher new operating value model function is at least in a subrange of the predetermined flow range higher than the previous operating value model function, and the lower new operating value model function may be at least in a subrange of the predeter- mined range lower than the previous operating model function. [16] As explained above, in case that the actual operating value is between the minimum and the maximum of the operating value model function, the model function may not be automatically updated or cal- ibrated. However, it may be advantageous to run the pump in a set of calibration batch runs in this case. For example, such calibration batch runs may be started when a new pump is installed and initially operat- ed for the first time. Alternatively, or in addition, such a calibration batch runs may be part of a regular maintenance action. It is preferred that the calibration batch runs are not automatically started, but start- ed manually by skilled personnel or user. [17] Optionally, the model updating electronics may be further con- figured to: - determine a maximum actual operating value and/or a mini- mum actual operating value among the calibration batch runs, - compare the maximum actual operating value with the maxi- mum of the operating value model function and/or the mini- mum actual operating value with the minimum of the operating value model function; and - automatically update the operating value model function to a higher new operating value model function if the maximum ac- tual operating value is higher than the maximum of the operat- ing value model function, and/or automatically update the op- erating value model function to a lower new operating value model function if the minimum actual operating value is lower than the minimum of the operating value model function. [18] In other words, the calibration batch runs at different speeds are used to check if a maximum actual operating value is above a maxi- mum of the model function and whether a minimum actual operating value is below a minimum of the model function. Thereby, a broader range or area of operating points can be scanned to update or cali- brate the model parameter sets that are indicative of the model func- tions corresponding to the respective pump speeds. [19] Optionally, the model updating electronics may be integrated into the control electronics that command the pump to run at a set pump speed based on the current operating value model function. Alternatively, or in addition, the model updating electronics may be arranged separately and the model updating electronics is merely in signal connection with the storage element to store the updated mod- el parameter set received from the model updating electronics. [20] According to a second aspect of the present disclosure, a pump assembly is provided comprising a pump, an electric motor for driving the pump, and an electronics housing for housing motor control elec- tronics, wherein the above-mentioned pump control system is arranged within the electronics housing. This has the advantage that the pump assembly is able to perform an automatic pump model calibration without having any access or connection to a remote pump control system or a remote model updating electronics. [21] According to a third aspect of the present disclosure, a method for controlling a pump is provided, wherein the method comprises the following steps: - running a pump at a set pump speed based on an operating value model function corresponding to the set pump speed; - determining an actual operating value at the set pump speed; - comparing the actual operating value with a maximum of the operating value model function and/or a minimum of the oper- ating value model function; and - automatically updating the operating value model function to a higher new operating value model function if the actual operat- ing value is higher than the maximum of the operating value model function, and/or automatically updating the operating value model function to a lower new operating value model function if the actual operating value is lower than the minimum of the operating value model function. [22] Optionally, updating the operating value model function to a higher new operating value model function may include shifting or scaling the operating value model function upward and/or updating the operating value model function to a lower new operating value model function may include shifting or scaling the operating value model function downward. [23] Optionally, a shift amount and/or a scale factor may be applied, wherein the shift amount and/or the scale factor may be a predeter- mined constant or may depend in a predetermined way on the set pump speed. [24] Optionally, a shift amount and/or a scale factor may be applied, wherein the shift amount and/or the scale factor depends on a differ- ence and/or ratio between an axial operating value and the maximum of the operating value model function or the minimum of the operating value model function. [25] Optionally, the method may further comprise running the pump in a set of calibration batch runs at different pump speeds if the actual operation value is at or higher than the minimum of the operating value model function and if the actual operating value is at or lower than the maximum of the operating value model function. [26] Optionally, the method may further comprise the following steps: - determining a maximum actual operating value and/or a mini- mum actual operating value among the calibration batch runs; - comparing the maximum actual operating value with the maxi- mum of the operating value model function and/or the mini- mum actual operating value with the minimum of the operating value model function; and - automatically updating the operating model function to a high- er new operating value model function if the maximum actual operating value is higher than the maximum of the operating value model function, and/or automatically updating the oper- ating value model function to a lower new operating value model function if the minimum actual operating value is lower than the minimum of the operating value model function. [27] The method disclosed herein may be implemented in form of compiled or uncompiled software code that is stored on a computer readable medium with instructions for executing the method. Alterna- tively, or in addition, the method may be executed by software in- stalled in electronics residing within an electronics housing of a pump assembly, or software running in a cloud-based system and/or a build- ing management system (BMS), e.g. in the pump control system dis- closed herein. [28] The pump control system and method described herein may be implemented and integrated in electronics residing within an electron- ics housing of a pump assembly, or in at least one central controller controlling a plurality of pumps, e.g. as part of a building management system (BMS). It is also possible to implement the control system and method described herein at least partially in a remote cloud compu- ting environment. A cloud computing environment may be particularly useful in case of geographically widely spread fluid distribution systems, such as a municipal water supply system or a district heating or cooling system. SUMMARY OF THE DRAWINGS [29] Embodiments of the present disclosure will now be described by way of example with reference to the following figures of which: Fig.1 shows a perspective view of a pump assembly according to the present disclosure; Fig.2 shows a Pq-diagram showing an example of pump char- acteristics of a pump assembly running at twelve different speeds; Fig.3 shows the pump characteristics as shown in figure 2 for four different pumps of the same pump type; Fig.4 shows a pump characteristic in a Pq-diagram for a single pump speed; Fig.5 shows a schematic diagram of a method for controlling a pump according to the present disclosure; and Fig.6 shows a schematic diagram of further preferred steps of a method for controlling a pump according to the present disclosure. [30] Figure 1 shows a pump assembly 1 according to the present dis- closure. The hardware of the pump assembly 1 is here essentially identi- cal to the known pump type “Grundfos Magna350-60”. The hardware comprises a pump housing 3 having an inlet flange 5 and an outlet flange 7. The inlet flange 5 and the outlet flange 7 may be connected to a pipe system (not shown) for pumping water or another liquid through the pipe system. The inlet flange 5 and the outlet flange 7 are arranged coaxially on an axis L that defines the predominant pump direction. The pump assembly 1 further comprises an electric motor drive 9 for driving an impeller (not visible) residing within the pump hous- ing 3. The motor rotor and the impeller rotate about a rotor axis R ex- tending perpendicular to the axis L. The pump assembly 1 further com- prises an electronics housing 11 for housing motor control electronics. The motor control electronics are configured to command the pump motor to run at a set pump speed. A front face of the electronics hous- ing 11 comprises a human machine interface 13 comprising a display 15 and buttons 17 for outputting and inputting information, commands, settings, parameters and/or programs. The pump assembly 1 may comprise automatic pump control algorithms to run automatically and/or may be set manually to run at a certain operating point by us- ing the human machine interface 13. [31] The pump control electronics within the electronics housing 11 is programmed in such a way that the pump assembly 1 comprises a new pump control system according to the present disclosure. The pump control system of the pump assembly 1 comprises a storage element having stored a plurality of model parameter sets for a plurality of pump speeds, wherein each model parameter set is indicative of an operat- ing value model function corresponding to a pump speed. If the oper- ating value model function is very precise, the pump assembly 1 is able to run at a desired operating point without having a measured or esti- mated flow value. For example, the operating value could be the ac- tual power consumption of the pump assembly or a current drawn by the motor drive of the pump assembly 1. Another example of an oper- ating value may be an output pressure measured by a pressure sensor. The advantage of considering an actual power consumption or a mo- tor current as the modelled operating value is that these are readily available without a need for a pressure sensor. [32] The pump control system of the pump assembly 1 further com- prises control electronics configured to command the pump to run at a set pump speed based on the operating value model function corre- sponding to the set pump speed, wherein a corresponding model pa- rameter set of the stored model parameter sets is retrieved from the storage element. The initial model parameter sets stored in the storage element when a new pump assembly 1 is commissioned may be de- termined for a pump type in laboratory tests or field tests using certain fitting and modelling procedures. [33] Figure 2 shows a result of such a fitting or modelling procedure to determine model parameter sets being indicative of operating value model functions. In the example shown in figure 2, the operating value is the power consumption P of the pump assembly 1, so that figure 2 shows a Pq-diagram, wherein the power consumption P of the pump assembly 1 is shown in dependence of the pumped flow q. Figure 2 shows so-called “power curves” for twelve different pump speeds ω1 to ω12. There are numerous ways to fit the “power curve” to the data points as shown in figure 2. Figure 2 actually shows two different model functions, wherein model function #2 is a fine-tuning of the model func- tion #1. The model function may, for example, be a polynomial func- tion. The model function may be described by a model parameter set that is initially determined by a simulation or a fitting procedure under laboratory or field test conditions. [34] Figure 3 shows how the power consumption model function P_ω(q) of figure 2 fits to four different pumps of the same pump type. As can be seen, manufacturing tolerances may lead to significant devia- tions between the data points and the model function P_ω(q). Such de- viations may also be due to pump variants of the same pump type or wear and tear after a certain amount of operating hours. [35] Figure 4 illustrates how a model function P_ω(q) is automatically updated or calibrated for each individual pump assembly 1 during normal operation of the pump assembly 1 in order to reduce the devia- tions between the true current pump characteristic and the assumed modelled pump characteristic. Figure 4 shows a power curve for a sin- gle pump speed over a predetermined range of flows between a min- imum flow q_min and a maximum flow q_max. The power curve in figure 4 has a well-defined minimum P_min and a well-defined maximum P_max. Figure 4 shows three different operating areas A, B and C. Operating area A includes all operating points with a power consumption below the minimum P_min of the power consumption model function P_ω(q). The operating area B includes all operating points with a power consump- tion between P_min and P_max. Operating area C includes all operating points with a power consumption above the maximum P_max of the power consumption model function P_ω(q). The idea to improve the model function is now to lower the model function if the actual current- ly determined power consumption P is in the operating area A, i.e. be- low the minimum P_min of the model function for the currently set pump speed ω. The rationale is here that the deviation between the actual determined power consumption P and the model function P_ω(q) can- not be due to the uncertainty in flow. Analogously, the model function is raised if the actual currently determined power consumption P is in the operating area C, i.e. above the maximum P_max of the model func- tion P_ω(q). The same rationale applies here that such a deviation can- not be explained by an uncertainty in flow. In case the actual currently determined power consumption is in the operating area B, i.e. between the minimum P_min and maximum P_max of the model function, the model function P_ω(q) may be kept unamended for the time being. The reason for this is that a deviation could be due to an uncertainty in the current unknown flow q. [36] Figure 5 shows the applied method for controlling the pump as- sembly 1 in a flow diagram. In a first step 501, the pump is run at a set pump speed ω based on an operating value model function P_ω(q) cor- responding to the currently stored model parameter sets. In a second step 503, an actual power consumption P of the pump assembly 1 is determined. The actual power consumption P is then compared in fol- lowing steps 505a, b with the minimum P_min of the power consumption model function P_ω(q) and/or with the maximum P_max of the power con- sumption model function P_ω(q). It should be noted that the compari- sons 505a,b could be performed in parallel or subsequently, wherein one comparison is obsolete if the other one is affirmative. However, both comparisons 505a,b are performed if the first one of the compari- sons is negative. If comparison 505a is affirmative, i.e. P < P_min, the mod- el function P_ω(q) is lowered (step 507a) and new model parameter sets for the new lowered power consumption model function P_ω,down(q) is stored at the storage element in order to subsequently run the pump based on the new lower power consumption model function P_ω,down(q). Analogously, if comparison 505b is affirmative, i.e. P > P_max, the power consumption model function P_ω(q) is raised (step 507b) to a new higher power consumption model function P_ω,up(q). A new parameter set model parameter set indicative for the new higher power consumption model function P_ω,down(q) is stored in the storage element in order to subsequently run the pump based on the new higher power consump- tion model function P_ω,up(q). Only in case that both comparisons 505a and 505b are negative, i.e. the currently determined actual power consumption P is in the operating area B, i.e. between P_min and P_max, the power consumption model function P_ω(q) may remain unamended (step 509) for the time being. [37] Figure 6 shows further steps that could be manually triggered in case the pump assembly 1 is running in the operating area B of figure 4, i.e. that the currently determined power consumption P is at or higher than the minimum P_min of the power consumption model function P_ω(q) and if the actual power consumption P is at or lower than the maximum P_max of the power consumption model function P_ω(q). For example, in step 601, the pump assembly 1 could be run in a set of calibration batch runs at different speeds. Thereby, a larger part of the operating range can be scanned for deviations between the actual power con- sumption P and the power consumption model function P_ω(q). In step 603, a minimum actual power consumption P_batch,min and/or a maximum actual power consumption P_batch,max among the calibration batch runs is determined. Similar to figure 5, two comparison steps 605a and 605b follow, wherein the minimum actual power consumption P_batch,min is compared with the minimum P_min of the power consumption model function P_ω(q) and the maximum actual power consumption P_batch,max is compared with the maximum P_max of the power consumption model function P_ω(q). The power consumption model function P_ω(q) only re- mains unamended if both comparisons are negative. If the minimum actual operating value among the calibration batch runs is lower than the minimum P_min of the power consumption model function P_ω(q), the power consumption model function P_ω(q) is lowered (step 607a) to a lower new power consumption model function P_ω,down(q). Analogously, if the maximum actual operating value P_batch,max among the calibration batch runs is higher than the maximum P_max of the power consumption model function P_ω(q), the power consumption model function P_ω(q) is updated or calibrated (step 607b) to a new higher power consumption model function P_ω,up(q). Only in case that both comparisons 605a and 605b are negative the power consumption model function P_ω(q) may remain unamended for the time being (step 609). [38] Where, in the foregoing description, integers or elements are mentioned which have known, obvious or foreseeable equivalents, then such equivalents are herein incorporated as if individually set forth. Reference should be made to the claims for determining the true scope of the present disclosure, which should be construed so as to encompass any such equivalents. It will also be appreciated by the reader that integers or features of the disclosure that are described as optional, preferable, advantageous, convenient or the like are optional and do not limit the scope of the independent claims. [39] The above embodiments are to be understood as illustrative ex- amples of the disclosure. It is to be understood that any feature de- scribed in relation to any one embodiment may be used alone, or in combination with other features described, and may also be used in combination with one or more features of any other of the embodi- ments, or any combination of any other of the embodiments. While at least one exemplary embodiment has been shown and described, it should be understood that other modifications, substitutions and alter- natives are apparent to one of ordinary skill in the art and may be changed without departing from the scope of the subject matter de- scribed herein, and this application is intended to cover any adapta- tions or variations of the specific embodiments discussed herein. [40] In addition, "comprising" does not exclude other elements or steps, and "a" or "one" does not exclude a plural number. Furthermore, characteristics or steps which have been described with reference to one of the above exemplary embodiments may also be used in com- bination with other characteristics or steps of other exemplary embod- iments described above. Method steps may be applied in any order or in parallel or may constitute a part or a more detailed version of anoth- er method step. It should be understood that there should be embod- ied within the scope of the patent warranted hereon all such modifica- tions as reasonably and properly come within the scope of the contri- bution to the art. Such modifications, substitutions and alternatives can be made without departing from the spirit and scope of the disclosure, which should be determined from the appended claims and their legal equivalents. [41] List of reference numerals: 1 pump assembly 3 pump housing 5 inlet flange 7 outlet flange 9 motor drive 11 electronics housing 13 human machine interface (HMI) 15 display 17 buttons L axis of predominant flow direction R rotor axis ω pump speed P power consumption q flow P_ω(q) operating value model function P_min minimum of model function P_max maximum of model function P_batch,min minimum actual operating value P_batch,max maximum actual operating value

Claims

Claims 1. A pump control system comprising - a storage element having stored a plurality of model parame- ter sets for a plurality of pump speeds, wherein each model pa- rameter set is indicative of an operating value model function ( P_ω(q)) corresponding to a pump speed (ω); - control electronics configured to command the pump to run at a set pump speed (ω) based on the operating value model function (P_ω(q)) corresponding to the set pump speed (ω), wherein a corresponding model parameter set of the stored model parameter sets is retrieved from the storage element; and - model updating electronics, wherein the model updating electronics are configured to: - determine an actual operating value (P) at the set speed (ω); - compare the actual operating value (P) with a maximum (P_max) of the operating value model function (P_ω(q)) and/or a mini- mum (P_min) of the operating value model function (P_ω(q)); and - automatically determine an updated model parameter set if the actual operating value (P) is higher than the maximum (P_max) of the operating value model function (P_ω(q)) or if the ac- tual operating value (P) is lower than the minimum (P_min) of the operating value model function (P_ω(q)), wherein the updated model parameter set is indicative of a higher new operating value model function ( P_ω,up(q)) if the actual operating value (P) is higher than the maximum (P_max) of the operating value model function (P_ω(q)), and wherein the updated model parameter set is indicative of a lower new operating value model function (P_ω,down(q)) if the actual operating value (P) is lower than the minimum (P_min) of the operating value model function (P_ω(q)), wherein the storage element is further configured to store the up- dated model parameter set.

2. The pump control system of claim 1, wherein the model updating electronics are configured to shift or scale the operating value model function (P_ω(q)) upward to the higher new operating value model function (P_ω,up(q)) and/or to shift or scale the operating value model function (P_ω(q)) downward to a lower new operating value model function (P_ω,down(q)).

3. The pump control system of claim 2, wherein a shift amount and/or a scale factor is applied, wherein the shift amount and/or the scale factor is a pre-determined constant or depends in a pre- determined way on the set pump speed (ω).

4. The pump control system of claim 2, wherein a shift amount and/or a scale factor is applied, wherein the shift amount and/or the scale factor depends on a difference and/or ratio between the actual operating value (P) and the maximum (P_max) of the op- erating value model function (P_ω(q)) or the minimum (P_min) of the operating value model function (P_ω(q)).

5. The pump control system of any of the preceding claims, wherein the model updating electronics are configured to apply a correc- tion function to tweak a shape of the operating value model func- tion (P_ω(q)).

6. The pump control system of any of the preceding claims, wherein the operating value model function (P_ω(q)) is a model curve of the operating value (P) in dependence of a pumped flow (q) within a pre-determined flow range ([ q_min, q_max]).

7. The pump control system of claim 6, wherein the higher new oper- ating value model function (P_ω,up(q)) is at least in a sub-range of the pre-determined flow range ([q_min, q_max]) higher than the previ- ous operating value model function (P_ω(q)), and wherein the low- er new operating value model function (P_ω,down(q)) is at least in a sub-range of the pre-determined range [q^min, q^max] lower than the previous operating value model function (P_ω(q)).

8. The pump control system of any of the preceding claims, wherein the control electronics are further configured to run the pump in a set of calibration batch runs at different pump speeds if the actual operating value (P) is at or higher than the minimum (P_min) of the operating value model function (P_ω(q)) and if the actual operat- ing value (P) is at or lower than the maximum (P_max) of the operat- ing value model function (P_ω(q)).

9. The pump control system of claim 8, wherein the model updating electronics are further configured to: - determine a maximum actual operating value (P_batch,max) and/or a minimum actual operating value (P_batch,min) among the calibration batch runs, - compare the maximum actual operating value (P_batch,max) with the maximum (P_max) of the operating value model function (P_ω(q)) and/or the minimum actual operating value (P_batch,min) with the minimum (P_min) of the operating value model function (P_ω(q)); and - automatically update the operating value model function (P_ω(q)) to a higher new operating value model function (P_ω,up(q)) if the maximum actual operating value (P_batch,max) is higher than the maximum (P_max) of the operating value model function (P_ω(q)), and/or automatically update the operating value model function (P_ω(q)) to a lower new operating value model function (P_ω,down(q)) if the minimum actual operating value (P_batch,min) is lower than the minimum (P_min) of the operat- ing value model function (P_ω(q)).

10. The pump control system of any of the preceding claims, wherein the model updating electronics are integrated into the control electronics.

11. A pump assembly (1) comprising a pump, an electric motor (9) for driving the pump, and an electronics housing (11) for housing mo- tor control electronics, wherein the pump control system of any of the preceding claims is arranged within the electronics housing (11).

12. A method for controlling a pump, wherein the method comprises the following steps: - running (501) a pump at a set pump speed (ω) based on an operating value model function (P_ω(q)) corresponding to the set pump speed (ω); - determining (503) an actual operating value (P) at the set pump speed (ω); - comparing (505a, 505b) the actual operating value (P) with a maximum (P_max) of the operating value model function (P_ω(q)) and/or a minimum (P_min) of the operating value model function (P_ω(q)); and - automatically updating (507b) the operating value model function (P_ω(q)) to a higher new operating value model func- tion (P_ω,up(q)) if the actual operating value (P) is higher than the maximum (P_max) of the operating value model function (P_ω(q)), and/or automatically updating (507a) the operating value model function (P_ω(q)) to a lower new operating value model function (P_ω,down(q)) if the actual operating value (P) is lower than the minimum (P_min) of the operating value model function (P_ω(q)).

13. The method of claim 12, wherein updating (507b) the operating value model function (P_ω(q)) to a higher new operating value model function (P_ω,up(q)) includes shifting or scaling the operating value model function (P_ω(q)) upward and/or updating (507a) the operating value model function (P_ω(q)) to a lower new operating value model function (P_ω,down(q)) includes shifting or scaling the operating value model function (P_ω(q)) downward.

14. The method of claim 13, wherein a shift amount and/or a scale factor is applied, wherein the shift amount and/or the scale factor is a pre-determined constant or depends in a pre-determined way on the set pump speed (ω).

15. The pump control system of claim 13, wherein a shift amount and/or a scale factor is applied, wherein the shift amount and/or the scale factor depends on a difference and/or ratio between an actual operating value (P) and the maximum (P_max) of the op- erating value model function (P_ω(q)) or the minimum (P_min) of the operating value model function (P_ω(q)).

16. The method of any of the claims 12 to 15, further comprising run- ning (601) the pump in a set of calibration batch runs at different pump speeds if the actual operating value (P) is at or higher than the minimum (P_min) of the operating value model function (P_ω(q)) and if the actual operating value (P) is at or lower than the maxi- mum (P_max) of the operating value model function (P_ω(q)).

17. The method of claim 16, further comprising - determining (603) a maximum actual operating value ( P_batch,max) and/or a minimum actual operating value ( P_batch,min) among the calibration batch runs; - comparing (605a, 605b) the maximum actual operating value (P_batch,max) with the maximum (P_max) of the operating value model function (P_ω(q)) and/or the minimum actual operating value (P_batch,min) with the minimum (P_min) of the operating value model function (P_ω(q)); and - automatically updating (607b) the operating value model function (P_ω(q)) to a higher new operating value model func- tion (P_ω,up(q)) if the maximum actual operating value (P_batch,max) is higher than the maximum (P_max) of the operating value mod- el function (P_ω(q)), and/or automatically updating (607a) the operating value model function (P_ω(q)) to a lower new operat- ing value model function (P_ω,down(q)) if the minimum actual operating value (P_batch,min) is lower than the minimum (P_min) of the operating value model function (P_ω(q)).