Fluid flow

Abstract

A method for estimating a volume of liquid output from an electric pump during an operating period of the electric pump, the method comprising: determining a water flow delay of the pump, where the water flow delay is a period between a first time corresponding to a first water flow drawn by the pump and a second time corresponding to a second water flow drawn by the pump; and generating an estimate of the volume of liquid output from the electric pump during the operating period based on the flow delay.

Description

[0001] Technical Field and Background Art

[0002] The present technology relates to flow estimation in the field of electric pumps. More particularly, but not exclusively, the present technology relates to estimating the volume or flow of liquid output from an electric pump without the use of a flow meter.

[0003] Beverage preparation devices or liquid dispensing machines typically include one or more electric pumps for moving liquid through a liquid path within the device or machine. In such cases, the flow rate of liquid output from the electric pumps (typically expressed as volume per unit time) is measured and summed to determine the total volume of liquid output from the beverage preparation device, for example to indicate when to stop outputting liquid from the device.

[0004] Conventionally, flow meters are used to measure the flow rate of liquid output from an electric pump. However, such flow meters can increase manufacturing complexity of the entire system, increase manufacturing costs, introduce additional failure points, and introduce irregularities into the liquid path. Therefore, it may be appropriate to estimate the flow rate of liquid output from an electric pump (which may be used interchangeably with the conceptual description of volume per unit time herein) without using a flow meter.

[0005] Known methods are proposed in EP2548483 and US2013094840 which attempt to estimate the water flow through the heater based on measured water temperature and a physical formula relating the energy required to heat a specific amount of water. Other known methods are proposed in WO12062628 and EP3097827.

[0006] However, as recognized by the inventors of the present method, existing methods for estimating the flow rate of a liquid without a dedicated flow meter may lack accuracy and in some cases require the introduction of other complex hardware equipment or rely on additional values ​​of estimated parameters, thereby introducing additional uncertainty into the flow estimate. Summary of the invention

[0007] Particular aspects and embodiments are listed in the accompanying claims.

[0008] Viewed from a first aspect, there is provided a method for estimating the volume of liquid output from an electric pump during an operating period of the electric pump, the method comprising: determining a water flow delay of the pump, wherein the water flow delay is a period between a first time corresponding to a first water flow drawn by the pump and a second time corresponding to a second water flow drawn by the pump; and generating an estimate of the volume of liquid output from the electric pump during the operating period based on the water flow delay. Thus, there is provided a method for efficiently and accurately estimating the flow output from the electric pump by using parameters directly related to the operation of the electric pump. Therefore, a flow meter can be omitted.

[0009] Viewed from another aspect, there is provided a liquid dispensing machine comprising an electric pump and a processing circuit configured to perform the method described herein. Thus, there is provided a liquid dispensing machine configured to efficiently and accurately estimate the flow rate output from the electric pump by using parameters directly related to the operation of the electric pump. Thus, a flow meter may be omitted from the liquid dispensing machine.

[0010] Viewed from another aspect, a computer readable medium is provided, the computer readable medium comprising instructions, which when executed by a programmable liquid dispensing machine causes the programmable liquid dispensing machine to perform the method described herein. Thus, a computer readable medium is provided, the computer readable medium comprising instructions, which when executed by a programmable liquid dispensing machine causes the programmable liquid dispensing machine to efficiently and accurately estimate the flow rate output from the electric pump by using parameters directly related to the operation of the electric pump.

[0011] Viewed from another aspect, a method for training a regression model to estimate the volume of liquid output from an electric pump during an operation period of the electric pump is provided, the model comprising: obtaining training data, the training data comprising values ​​of a water flow delay of a test device and the volume of liquid output by the test device during operation of the test device, wherein: the test device corresponds to the electric pump; and the water flow delay is a period between a first time corresponding to a first flow of water drawn by the test device and a second time corresponding to a second flow of water drawn by the test device; and performing a regression analysis on the training data to provide one or more parameters that can be used to estimate the volume of liquid output from the electric pump based on measurements of the water flow delay of the electric pump. Thus, a method for training a regression model to efficiently and accurately estimate the flow output from the electric pump by using parameters directly related to the operation of the electric pump is provided.

[0012] Viewed from another aspect, a beverage preparation device is provided, the beverage preparation device comprising: a fluid input end connected to receive a fluid for beverage preparation; an electric pump connected to the fluid input end and having a control input end; a water flow sensor connected to a power connection of the electric pump; a beverage preparation output end connected to receive a fluid dispensed by the electric pump; and a controller connected to the water flow sensor and the control input end, wherein the controller is configured to calculate the volume of fluid dispensed by the electric pump based on a water flow draw delay measured by the water flow sensor, and to send a control signal to the controller input end to deactivate the electric pump in response to the fluid volume reaching a threshold volume. Thus, a beverage preparation device is provided, the beverage preparation device being configured to efficiently and accurately estimate a flow rate output from the electric pump by using parameters directly related to the operation of the electric pump. Thus, a flow meter can be omitted from the beverage preparation device.

[0013] Other aspects will become apparent after review of the disclosure, particularly after review of the Brief Description of the Drawings, Detailed Description, Drawings, and Claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Examples of the present disclosure will now be described, by way of example only, with reference to the accompanying drawings, in which:

[0015] Figure 1 : Schematically illustrates a beverage preparation device configured to operate according to the teachings of the present disclosure.

[0016] Figure 2 : illustrates the relationship between flow rate and water flow delay according to the teachings of the present disclosure.

[0017] Figure 3 : Schematically illustrates a method for estimating the volume of liquid output from an electric pump during an operation period of the electric pump according to the teachings of the present disclosure.

[0018] Figure 4 : illustrates the relationship between the water flow drawn by an electric pump according to the teachings of the present disclosure and time.

[0019] Figure 5 : illustrates the relationship between flow rate and water flow delay for three power supply voltages according to the technology of the present disclosure.

[0020] Figure 6 :include Figure 6 a and Figure 6 b, These graphs illustrate the relationship between time and pump power and time and cumulative volume according to the disclosed techniques.

[0021] Figure 7: illustrates the relationship between time and cumulative volume of liquids at different temperatures according to the technology of the present disclosure.

[0022] Figure 8 : Schematically illustrates a method for calculating the cumulative volume of liquid output from an electric pump according to the technology of the present disclosure.

[0023] Fig. 9 : Schematically illustrates a method for estimating the volume of liquid output from an electric pump when the power supply voltage varies according to the technology of the present disclosure.

[0024] Fig.10 : A method for training a regression model to estimate the volume of liquid output from an electric pump during an operating period of the electric pump is schematically illustrated.

[0025] Fig.11 : illustrates the estimation performance of the technology of the present disclosure for estimating the volume of liquid output from an electric pump during an operation period of the electric pump.

[0026] Although the present disclosure is susceptible to various modifications and alternative forms, specific exemplary methods are shown by way of example in the drawings and described in detail herein. However, it should be understood that the drawings and the accompanying specific embodiments are not intended to limit the present disclosure to the specific forms disclosed, but the present disclosure is intended to cover all modifications, equivalents and alternatives that fall within the spirit and scope of the claimed invention.

[0027] It will be appreciated that features of the above-described examples of the present disclosure may be used conveniently and interchangeably in any appropriate combination. DETAILED DESCRIPTION

[0028] The present method as described below relates to an estimate of the flow rate output from an electric pump. Although these methods are described in the context of a beverage preparation device, it should be understood that the techniques for flow rate estimation are applicable to a wide variety of arrangements in which an electric pump is used to output a flow rate of liquid. Such an estimate can be used to control the amount of liquid output by the pump, for example, to output a known volume of liquid in an activation window that permits control of the pump. Such an estimate can also or alternatively be used to keep track of the total liquid output by the pump (e.g., within a specific time period and / or within the operating life of the pump).

[0029] Figure 1 The beverage preparation device is schematically illustrated. In some examples, the beverage preparation device 100 or the liquid dispensing machine can be a coffee machine or a capsule-based coffee machine. As mentioned above, the technology described herein can be applied to any device or machine including an electric pump configured to output a liquid or fluid. In fact, in some examples, such an electric pump can be an oscillating pump or a positive displacement pump, such as an oscillating piston pump or an oscillating plunger pump.

[0030] In the present example, the beverage preparation device 100 comprises a fluid input 110 connected to receive a fluid for beverage preparation. The beverage preparation device 100 also comprises an electric pump 130 connected to the fluid input 110 and having a control input 140. The fluid input 110 and the electric pump 130 are connected by a fluid path 115 to allow fluid to flow from the fluid input 110 to the electric pump 130. The beverage preparation device 100 also comprises a water flow sensor 160, which is connected to the power connection of the electric pump 130 (indicated by connection 155) and is configured to measure the water flow drawn by the electric pump 130.

[0031] The beverage preparation apparatus 100 further comprises a beverage preparation output 150 connected to receive fluid dispensed by the electric pump 130. The electric pump 130 and the beverage preparation output 150 are connected by a fluid path 145 to allow fluid to flow from the electric pump 130 to the beverage preparation output 150.

[0032] The water flow sensor 160 is further connected to the controller 170 (indicated by connection 165) so that a measurement of the water flow drawn by the electric pump 130 is provided to the controller 170. The controller 170 is also connected to the control input 140 of the electric pump 130 (indicated by connection 175). The controller 170 is configured to calculate the volume of fluid dispensed by the electric pump 130 based on the water flow draw delay measured by the water flow sensor 160, and to send a control signal to the control input 140 of the electric pump 130 to deactivate the electric pump 130 in response to the fluid volume reaching a threshold volume.

[0033] It should be understood that the beverage preparation device 100 may include Figure 1 For example, additional components may be present in the fluid path between the fluid input end 110 and the electric pump 130 and in the fluid path between the electric pump 130 and the beverage preparation output end 150. For example, some form of beverage preparation chamber and / or heating component may be provided between the electric pump 130 and the beverage preparation output end 150, and / or some form of fluid input filter may be present between the fluid input end 110 and the electric pump 130.

[0034] Figure 2 The relationship between the liquid flow rate or volume of liquid per unit time of liquid output from the electric pump during operation of the electric pump and the water flow delay parameter of the electric pump is shown. In particular, the inventors of the present method have determined that the volume of liquid output from the electric pump is related to the water flow delay of the electric pump. The water flow delay of the electric pump as used herein is a period of time between a first time corresponding to a first water flow drawn by the electric pump and a second time corresponding to a second water flow drawn by the pump. In the following and with particular reference to Figure 4 The first water flow and the second water flow are discussed in more detail.

[0035] like Figure 2 As shown, generally, the greater the water flow delay of the electric pump, the smaller the volume of liquid output from the electric pump per unit time, in other words, the smaller the flow rate of liquid output from the pump. Conversely, a smaller water flow delay generally corresponds to a larger flow rate of liquid output from the pump. After determining this correlation, the inventors of the present invention realized that this relationship allows an estimate of the volume of liquid output from the electric pump during the operating period of the electric pump to be generated based on the determination of the water flow delay of the electric pump.

[0036] Figure 3 A method 300 for estimating a volume of liquid output from an electric pump during an operation period of the electric pump according to the teachings of the present disclosure is schematically illustrated. The method 300 comprises the following steps.

[0037] In step S301, a water flow delay of the electric pump is determined. As described above, the water flow delay is the period between a first time corresponding to a first water flow drawn by the pump and a second time corresponding to a second water flow drawn by the pump.

[0038] In step S302 , an estimate of the volume of liquid output from the electric pump during the operation period is generated based on the water flow delay.

[0039] Thus, the present method enables the volume of liquid output from the pump during the operating period of the electric pump to be estimated without the use of a dedicated flow meter. In doing so, the manufacturing complexity of the system can be reduced, the manufacturing cost of the system can be reduced, a point of failure of the system (the flow meter) can be removed, and irregularities in the liquid path can be reduced. Any one or more of these results can be provided in any specific implementation, thereby providing a more efficient manufacturing process and / or more reliable operation.

[0040] Furthermore, since the estimate is generated based on the water flow delay of the electric pump, the need to introduce other complex hardware devices is avoided, and reliance on parameters that have already been estimated is also avoided. Therefore, the method can avoid introducing additional uncertainty into the flow estimate when the flow meter is omitted. Therefore, the method can provide a more accurate flow estimate in the absence of a flow meter. Furthermore, since the flow estimate is based on the water flow delay of the electric pump (i.e., based on a single hardware device), the computational efficiency is improved.

[0041] The above discussion and Figure 3 refers to the water flow delay of the pump, the first time corresponding to the first water flow drawn by the pump and the second time corresponding to the second water flow drawn by the pump. Figure 4 The water flow delay and the determination of the first time and the second time and the water flow are described in more detail.

[0042] Figure 4 The relationship between the pump flow (i.e. the flow of water drawn by the electric pump) and time is shown. More specifically, Figure 4 Five pump cycles are shown over a period of approximately 100 ms. Each pump cycle is characterized by an increase from approximately zero flow drawn by the pump, after a short period of noise-like behavior, the flow rapidly increasing in a substantially linear manner (i.e., with a substantially constant rate of change), then increasing less rapidly and reaching a peak flow, then returning again to noise-like behavior and finally to a time when no flow is drawn. Figure 4 In the depicted example, the peak water flow is about 1 A, and the water flow at the transition from having a substantially linear increase to having a less linear increase is about 0.3 A. However, it should be understood that the present disclosure is not limited to such specific values, and that in fact the specific values ​​for any given implementation will vary depending on the specific pump.

[0043] Figure 4 The time 410 on the figure may indicate a first time, and the time 420 may indicate a second time. As shown, the first time 410 corresponds to a first water flow 415 drawn by the pump, and the second time 420 corresponds to a second water flow 425 drawn by the pump. The first water flow may substantially correspond to a zero crossing of the AC power supply voltage supplied to the pump. For example, the first water flow may be a water flow drawn by the pump when the AC power supply voltage supplied to the pump crosses a zero crossing voltage point, or when the voltage sign of the AC voltage changes from negative to positive, although in other examples it may be vice versa. Alternatively, the first water flow may be a predefined water flow, in which case the first time will be the time when the water flow drawn by the pump reaches the predefined water flow. In addition, the first water flow may be a water flow drawn by the pump determined by analyzing the water flow drawn by the pump within a rolling time window, wherein it is determined that the water flow has increased beyond noise-like behavior. Alternatively, the first water flow may be a water flow determined by empirical or experimental analysis of the water flow characteristics of the electric pump or a pump corresponding to the electric pump.

[0044] The second time 420 may correspond to a time near the end of, or within, a period of time during which the flow of water drawn by the pump has a substantially constant rate of change. In other words, the second time 420 may be a time at the end of, immediately after, or within a period of time when the flow of water drawn by the pump increases at a constant rate. Figure 4 The second time 420 corresponds to the end of the period during which the flow of water drawn by the pump has a substantially constant rate of change, which point is indicated by 425. As described above, Figure 4In the example shown, this second time 420 and point 425 correspond to a flow of water drawn by the pump having a value of approximately 0.3 A, as shown by point 430 on the graph. The second time 420 may be a time determined by performing an empirical or numerical analysis of the flow of water drawn by the electric pump or a pump corresponding to the electric pump over time. With respect to the first time, the second time 420 may be determined based on an analysis of the flow of water drawn by the pump over a rolling time window.

[0045] Furthermore, in some examples, the second time 420 may be a time determined based on the water flow drawn by the pump at that time. Specifically, it may be appropriate to ensure that the second time 420 does not correspond to a water flow drawn by the pump beyond a predetermined position of the water flow curve. This may be appropriate, for example, to allow for changes in liquid pressure when the pressure is lower. This may be relevant in the context of a beverage preparation device or in fact feeding liquid through a connection tube that provides a lower or inconsistent liquid pressure rather than feeding liquid from a built-in reservoir.

[0046] In some examples, the second time 420 may be determined based on the second time 420 being within a predetermined duration from the first time 410. As an example, the predetermined duration may be X seconds, and the first time may correspond to Y seconds. Following this example, the second time 420 may be determined as a time within X+Y seconds, or equal to X+Y seconds.

[0047] It should be understood that the first time 410 and the second time 420 can be determined by measuring the water flow drawn by the pump. For example, the water flow meter functionality can be used to measure the water flow during the entire operation of the pump. In doing so, the water flow delay can be determined from the measurement of the water flow drawn only by the pump. This can generate flow estimates efficiently and quickly.

[0048] Alternatively, in some examples, the first time 410 may be determined based on a measurement of the AC power voltage supplied to the pump during the operating period of the pump. The second time 420 may be determined based on a measurement of the water flow drawn by the pump during the operating period. Thus, the calculated difference between the first time and the second time provides the water flow delay of the pump. Following this example and referring to Figure 4 , the water flow delay may be calculated as the difference between the first time 410 and the second time 420 , for example by subtracting the first time 410 from the second time 420 .

[0049] In this way, the first time 410 can be determined by measuring the AC power supply voltage of the pump, for example by determining the time corresponding to the zero crossing of the power supply voltage as the first time. Figure 4As shown, the water flow drawn by the pump may be unstable, noisy, close to zero flow, and thus using a predefined flow to determine the first flow, and then determining the first time as the time when the pump draws the predefined flow, may be prone to error. Instead, and as discussed, a measurement of the AC power supply voltage and the zero crossing of the AC voltage supply may be used. In doing so, the accuracy of the determination of the first time 410 may be improved, resulting in a more accurate flow estimate.

[0050] like Figure 4 As shown, the water flow delay can be determined by subtracting the first time 410 from the second time 420. So far, the zero crossings of the AC supply voltage have all referred to increasing zero crossings, i.e., the zero crossings where the voltage turns from negative to positive. In other examples, the falling zero crossings of the AC supply voltage can be used, i.e., corresponding to the water flow drawn by the pump in the second half of the pump cycle. This approach can be able to ignore any noise-like behavior around the zero water flow point, because the first time when the water flow reaches zero (or substantially zero) can be used as the zero crossing. Referring again to Figure 4 This is described. The times marked 435 and 440 may correspond to the falling zero crossings of the AC power voltage supplied to the pump. In this example, the water flow delay is determined as follows: water flow delay = ((440-435) / 2)-(440-420). In some hardware configurations, measuring the water flow delay in this manner may be more accurate.

[0051] In some examples, in step S301 of method 300, the water flow delay determination may be repeated before method 300 proceeds to the generation of the volume estimate in step S302. Thus, in this example, the volume estimate may be generated based on an average water flow delay of the pump. In some examples, the water flow delay is determined for each pump cycle, which may correspond to determining the water flow delay approximately every 20 ms, and then after five water flow delay determinations, an estimate of the volume may be generated based on an average of the five determined water flow delays. Thus, in some examples, an estimate of the volume may be generated every 100 ms, while the water flow delay may be determined every 20 ms. It should be understood that the number of water flow delays used for averaging may vary.

[0052] Reference again Figure 3In some examples, the method 300 may further include steps S301a and S302a. At step S301a, the average value of the AC power supply voltage supplied to the electric pump is determined. The average value of the AC power supply voltage may be a power supply voltage provided from a mains power grid or a power grid. Therefore, in some examples, the average value refers to the root mean square average value of the AC power supply voltage over a period of time. Examples of average AC power supply voltages that may be supplied to the electric pump include 110V, 120V, 220V, 230V, and 240V, etc. It should be understood that the average AC power supply voltage supplied to the electric pump may refer to the nominal power supply voltage, i.e., the voltage at the interconnection point between the power facility and the device or the voltage actually used by the electric pump. The AC power supply voltage may be determined in a variety of ways, such as by direct measurement, by measuring the power supply voltage from the power grid, or based on a tangent applied to the power supply voltage from the power grid.

[0053] Although step S302a Figure 3 , but it should be understood that step S302a optionally modifies step S302 rather than representing an additional separate step. At step S302a, an estimate of the volume of liquid output from the electric pump during the operating period is generated based on the average value of the water flow delay and the AC power supply voltage.

[0054] Therefore, in this example, the average value of the water flow delay of the pump and the AC power supply voltage supplied to the pump is determined, and the flow rate estimate is generated based on the average value of the water flow delay and the AC power supply voltage. The inventors of the present method have determined (e.g. Figure 2 The flow rate and water flow delay dependency depends on the supply voltage, e.g. whether 230V, 195V or 265V is supplied to the pump from the grid. This can be Figure 5 As seen in Figure 1, the graph shows the relationship between liquid flow rate or volume of liquid per unit time and water flow delay for different levels of average AC supply voltage supplied to the pump. Generally, a higher supply voltage (265V in this case) results in a lower average determined water flow.

[0055] Therefore, in some example methods including steps S301a and S302a, the flow estimation value generation is further based on the power supply voltage. In some specific implementations, introducing this additional parameter into the estimation method can lead to further improvement in the accuracy of the flow estimation value.

[0056] In some examples, generating the estimate of the volume of liquid output from the pump is further based on a technique selected from a plurality of techniques, the selection based on an indication of whether a power of the pump during an operating period of the pump is above a predefined power threshold.

[0057] Figure 6a shows the power of the electric pump over time during operation. As shown, the power of the pump does not start at 100%, but there are ramp-up periods of varying duration. In some examples, such as those involving beverage preparation devices, the power of the pump is increased over a period of time (e.g., Figure 6 a) is intentionally increased to full power within approximately 7 seconds. This may be due to a heating operation to be applied to the liquid flowing through the pump, for example to provide a hot beverage. In such cases, the temperature regulation may be such that the pump cannot be started at full power for hot beverages. This may occur, for example, if the heating of the liquid after it has been discharged from the pump is performed by a heater that requires some time to reach full operating temperature (for example due to heat losses in the preheating phase) and / or a heater with a low thermal mass. By such a pump ramp-up, as Figure 6 As shown in a, the cumulative volume of liquid discharged from the pump behaves differently at different times.

[0058] The behavior is Figure 6 b. For a time period corresponding to the time period during which the pump's power is ramped up (approximately 7 seconds in this example), the cumulative volume increases at a varying rate and at a different rate than when the pump is operating at 100%. Therefore, by taking the pump's power into account when selecting a technique for generating an estimate, a more accurate estimate of the flow rate and volume of liquid output from the pump over time can be achieved.

[0059] For example, one of the multiple techniques may be optimized for generating an estimate when the pump power is below a predetermined amount (e.g., 100% of full power or 70% of full power) of the nominal or maximum operating power of the pump. A different one of the multiple techniques may be optimized for generating an estimate when the pump power is operating around a predetermined power level (e.g., substantially at 100% or above a predetermined threshold of 70%). Thus, selecting which of the multiple techniques to use in estimate generation based on the power of the pump and specifically whether the power is above a predefined power threshold may result in a more accurate flow estimate for certain specific implementations.

[0060] As will be appreciated, the indication as to whether the power of the pump is above or below a predefined threshold may be measured or calculated directly or indirectly.Alternatively, an indication may be received.

[0061] In some examples, generating the estimate of the volume of liquid output from the pump is further based on a technique selected from a plurality of techniques based on an indication of whether a temperature of the liquid is greater than a predefined temperature threshold or whether a heating operation is applied to the liquid.

[0062] As discussed, if a heating operation is to be applied to the liquid being pumped, the pump may need to undergo a power ramp-up. Therefore, the selection may be based on an indication that a heating operation is to be applied to the liquid, rather than relying on an indication of whether the pump power is above a threshold or actually measuring or calculating the pump power. In some cases, this may be caused by a user of the beverage preparation device selecting a hot beverage to be prepared. In other cases, the selection may be based on whether the temperature of the liquid is greater than a predefined temperature, thereby determining that a heating operation has been applied. In some examples, the temperature may be sensed by a temperature sensor. In other examples, the processing circuit is configured to receive an indication that a heating operation is to be applied to the liquid, and select a technique based on the indication.

[0063] Figure 7 The cumulative volume of hot and cold beverages versus time in the context of a beverage preparation device is shown. As shown, the rate at which the cumulative volume increases (i.e., the extraction of liquid) depends on whether a hot or cold beverage is extracted. In some examples, this may be due to flow restrictions caused by the heater path, and / or due to different physical properties of the liquid at different temperatures, through which the liquid output from the pump passes in order to heat the pumped liquid. Therefore, depending on the temperature of the liquid or whether a heating operation is to be applied to the liquid (i.e., the final temperature of the liquid), it may be appropriate to use a separate or different technique.

[0064] In some examples, generating an estimate of the volume of liquid output from the pump is further based on a technique selected from a plurality of techniques, the selection being based on an indication of whether a duration of an operation period of the pump is greater than a predefined duration threshold. Rather than receiving an indication of a pump power, or a heating operation to be applied, or a temperature of the liquid as described above, the selection may be based on the duration of operation of the pump, i.e., the length of time the pump has been operating for extraction or output of liquid.

[0065] from Figure 6 As can be seen in Figure 1, in some examples, the pump ramp-up is reliably completed after a duration of 7 seconds of pump operation, so that after 7 seconds the pump can be expected to operate at 100% power. Therefore, the duration of time the pump is operating can be used as a proxy to indicate that the pump is operating at full power. Different pumps, and indeed different heating methods, may have different characteristics, and therefore different durations will indicate full pump power, and the duration of the predefined operation (7 seconds in this example) will vary accordingly in different specific implementations.

[0066] In some examples, the selection is based on a combination of the above selection criteria. In some examples, the selection is a two-stage selection, the first stage selection is based on whether to apply a heating operation, and the second stage selection is based on the duration of the pump operation. For example, it can be determined first whether to apply a heating operation to the liquid. This can be determined in response to an indication of whether to apply a heating operation. If it is determined that a heating operation is not applied to the liquid, a technology in the technology can be selected, which can be suitable for cold liquids and / or optimized for cold liquids. Alternatively, if it is determined that a heating operation is to be applied to the liquid, different technologies (suitable for hot liquids and / or optimized for hot liquids) can be applied. For example, a technology can be optimized for the first period of operation of a hot beverage, and a second technology can be optimized for the second period after the first period of operation of the hot beverage. In other words, once it is determined that a heating operation is to be applied to the liquid, the selection is based on the duration (i.e., time) of the operation. In at least some specific implementations, a single technology can be applied to cold drinks, considering that the temperature adjustment for cold drinks and the pump ramp-up are expected to be less relevant (or not relevant at all) accordingly.

[0067] Figure 8 Schematically illustrated is a method 800 for calculating the cumulative volume of liquid output from an electric pump in accordance with the techniques of this disclosure. It should be understood that method 800 may be operated sequentially after method 300. From one perspective, method 300 results in an estimate of the volume output by the pump during a first operating period of the pump (e.g., 100 ms), and method 800 then defines a repetition of method 300 during a second operating period of the pump, and summing the estimates of the volumes from the first operating period and the second operating period to calculate the total volume of liquid output by the pump during a time period equal to the total of the first operating period and the second operating period.

[0068] More specifically, at step S801, in response to determining that the second operation of the electric pump has elapsed, a second water flow delay of the electric pump is determined and a second estimate of the volume of liquid output from the pump during the second operation period is generated. The method 300 and associated examples may be repeated at step S801. In addition, the elapse of the second operation period may be determined based on a timer (e.g., a timer indicating that a predefined time period corresponding to the operation period has elapsed). In some examples, the length of the operation period is 100 ms, and the first operation period and the second operation period are the same.

[0069] At step S802, the total volume of liquid output from the pump during the first operating period and the second operating period is calculated by summing the first estimated value of the volume and the second estimated value of the volume. In this way, the total volume of the liquid output can be determined. Then, the calculated total volume can be used to control the pump, for example, to deactivate the pump once a predetermined total volume is reached. In an example in which the described method is performed in the context of a beverage preparation device, the beverage preparation device can stop outputting liquid based on determining that the calculated total volume has reached a predetermined threshold. In this example, the predetermined threshold can be determined based on user input to the beverage preparation device (e.g., by selecting a beverage with a predefined volume).

[0070] In some examples, the generation of the second estimate of step S801 is based on a second technique, and the estimate from the first operating period is a first technique. The second technique may be selected from a plurality of techniques (e.g., a plurality of techniques for selecting a technique for the estimate in the first operating period). In practice, the second technique may be selected from a plurality of techniques based on at least one of: an indication of whether the power of the pump during the second operating period of the pump is above a predefined power threshold, an indication of whether the temperature of the liquid is greater than a predefined temperature threshold or whether a heating operation is to be applied to the liquid, and an indication of whether the duration of the first operating period and the second operating period of the pump is greater than a predefined duration threshold. It should be understood that the predefined power threshold, the predefined temperature, and the predefined duration threshold may be the same for the first estimate and the second estimate, or they may be different.

[0071] Thus, for an estimate of the volume from the first operating period, the second estimate may use techniques that are appropriate for and / or optimized for the conditions experienced during the second operating period. Thus, in at least some implementations, a more accurate estimate of the total volume output from the pump across the first operating period and the second operating period may be achieved.

[0072] In some examples, the first technique (i.e., the technique associated with generating a first estimate of the volume output during the first operating period) and the second technique are the same. For example, if the power of the pump has not exceeded a predefined power threshold between the first operating period and the second operating period, or the total duration of the pump operation has not exceeded a predefined duration threshold, the first technique and the second technique for the first operating period and the second operating period may be the same. In addition, as discussed above, based on an indication that no heating operation is applied to the liquid, the technique may be the same across the first operating period and the second operating period. In some examples, based on a determination that no heating operation is applied to the liquid, the second technique is automatically set to the first technique. For cold liquids (e.g., cold beverages), no pump power ramp-up is required, so a single technique can be used for the duration of the cold liquid being drawn.

[0073] Alternatively, the first technique and the second technique may be different. For example, the first operating period and the second operating period may span a point where the pump power crosses a predefined power threshold or the pump's operating duration crosses a predefined duration threshold. Thus, the techniques of the present disclosure allow for fine-grained application of different techniques to best suit the prevailing conditions, thereby permitting a technician to customize any particular implementation to its specific circumstances, thereby achieving a higher accuracy estimate of the volume output from the pump.

[0074] Fig. 9 Schematically illustrated is a method for estimating an estimated value of the volume of liquid output during a second operating period following a first operating period. It should be understood that method 900 may be sequentially followed by method 300 or method 800. In fact, these methods may be combined in different ways. Reference is made to the second operating period in a manner similar to method 800, but this does not require that these methods are mutually exclusive alternatives, but rather that the two methods may be combined and the two "second" periods may be the same period, or the two "second" periods may be different from each other, so that one such "second" period may be conceptually renamed as a "third" period.

[0075] The inventors of the present disclosure have determined that, in some cases, the AC supply voltage may vary across the operating period of the pump. Figure 5 As shown, for some electric pumps, the water flow delay and flow rate estimate may be dependent on the supply voltage. Therefore, for AC supply voltages that vary significantly, a further series of techniques may be employed to improve the accuracy of the flow / volume estimate. Thus, from one perspective, the method 900 considers whether the supply voltage (e.g., the nominal supply voltage (e.g., 230V) or the supply voltage at the pump) varies outside of a predetermined threshold between subsequent time periods, and if so, handles the change accordingly.

[0076] More specifically, in this example method 900, steps S902, S903, S904, and S905 may be performed after the determination at step S901, specifically, in response to determining that the second operation period of the pump has passed, and determining that the difference between the average value of the AC power supply voltage in the first operation period and the average value of the AC power supply voltage in the second operation period is greater than a predefined difference. As mentioned, the AC power supply voltage may be the AC power supply voltage supplied to the electric pump, as it may be a voltage supplied for use by the pump or it may be a nominal power supply voltage from a mains power grid. The AC power supply voltage may be determined in a variety of ways, such as by direct measurement, by measuring the power supply voltage from the power grid, or based on a phase tangent of a voltage supplied from the power grid. The predefined difference may be determined based on a percentage of the AC power supply voltage in the first operation period.

[0077] In step S902, a second water flow delay of the electric pump is determined. This can be determined as described herein.

[0078] At step S903, based on the average value of the AC power supply voltage in the second operation period and the second technique, an estimated value of the second water flow delay of the electric pump is generated. In this example, the second technique is used to eliminate the effect of the varying voltage on the water flow delay and generate an estimated value of the water flow delay.

[0079] At step S904, the estimated value of the second water flow delay is subtracted from the second water flow delay to calculate the adjusted water flow delay. In other words, the normalized water flow delay is calculated by subtracting the predicted water flow delay predicted using the second technique from the determined water flow delay.

[0080] At step S905, an estimate of the volume of liquid output from the pump during the second operating period is generated based on the adjusted water flow delay and the third technique. In this example, the third technique may be different from the second technique. The second technique in this example may be different from the second technique described in other examples herein.

[0081] Thus, in other words, method 900 provides a method by which a more accurate estimate of the volume output from a pump can be determined when the power supply voltage varies significantly (i.e., when the power supply voltage is treated as a continuous variable). In some cases, and in some locations, a largely constant power supply voltage (e.g., grid voltage) cannot be guaranteed, and thus the present technique can be used in related implementations to mitigate any negative impact on the accuracy of flow estimates in such cases without requiring the use of a dedicated flow meter in such cases and locations.

[0082] In some examples, the techniques described herein, the first technique, the second technique, and / or the third technique are one or more trained regression models trained based on performing regression analysis on training data, the training data including values ​​of water flow delays of a test device and volumes of liquid output by the test device during operation of the test device, wherein the test device includes a pump that corresponds in a related manner to an electric pump. In some examples, the trained regression model can be a trained linear regression model. In addition, as discussed above, the technique corresponding to the first 7 seconds of the draw (or any predetermined time period) can be a model that produces a constant value. For example, in some cases, a constant value can be used to appropriately estimate the first 7 seconds of the draw, and then a second technique (e.g., a trained linear regression model) can be used after the 7 second duration has passed. It should be understood that the duration (7 seconds in this example) can vary depending on the specific implementation.

[0083] In other examples, the technique corresponding to the first 7 seconds of extraction (or any predetermined time period) may be a linear regression model. In this example, the first technique may be different from the linear regression model used after the first 7 seconds.

[0084] In some examples, multiple test devices (such as test pumps or test beverage preparation devices) may be operated in conjunction with a flow meter and operated to determine the water flow delay of the pump to generate training data including measured values ​​of flow and water flow delay that can be used to generate estimates of flow.

[0085] An example trained linear regression model will now be described, which may correspond to any of the techniques discussed above. The trained linear regression model may include several determined constant model terms, which together with the water flow delay and the supply voltage can be used to calculate the predicted volume of liquid output from the pump. Specifically, in this example, the predicted volume per unit time = parameter 1 + (parameter 2 × water flow delay) + (parameter 3 × supply voltage) + (parameter 4 × water flow delay × supply voltage). In this example, parameter 1, parameter 2, parameter 3, and parameter 4 are constant terms determined using linear regression analysis. In this way, the water flow delay and the supply voltage can be input to the trained linear regression model (i.e., input to the equation, as described above) and combined with the determined model parameters to produce an output corresponding to an estimated value of the volume output during a period of time. It should be understood that other equations can characterize the linear regression model. The linear regression model discussed herein refers to one or more model parameters determined from a linear regression analysis (ie, training) and equations linking the model parameters to input and output variables (in this example, water flow delay, supply voltage, and volume estimate).

[0086] An example method for trained linear regression model selection will now be described, which can be used as part of any of the techniques described herein. In this example, it is first determined whether a heating operation is to be applied to the liquid. For example, the determination can be based on whether a user of the beverage preparation device selects a hot beverage to be prepared. If it is determined that a heating operation is not to be applied (for example, because the user has selected a cold drink), the model can be used to predict the volume of the liquid output from the pump as follows: predicted volume per unit time = parameter 1 + (parameter 2 × water flow delay) + (parameter 3 × power supply voltage). In this example, parameter 1, parameter 2, and parameter 3 are constant terms determined using linear regression analysis. If it is determined that a heating operation is to be applied to the liquid, a determination is made as to whether the pump is operating at or above a predetermined amount of the pump's nominal or maximum operating power (e.g., 100% of full power or 70% of full power). If it is determined that the pump is operating at or above the predetermined amount, the model can be used to predict the volume of the liquid output from the pump as follows: predicted volume per unit time = parameter 1 + (parameter 2 × water flow delay) + (parameter 3 × power supply voltage). In this example, Parameter 1, Parameter 2, and Parameter 3 are constant terms determined using linear regression analysis, and may be different from constant terms associated with a situation where no heating operation is applied to the liquid. If it is determined that the pump is operating at a power lower than the predetermined amount, the model may be used to predict the volume of liquid output from the pump as follows: Predicted volume per unit time = Parameter 1 + (Parameter 2 × pump power) + (Parameter 3 × time elapsed since the start of extraction) + (Parameter 4 × temperature of the liquid). In this example, Parameter 1, Parameter 2, Parameter 3, and Parameter 4 are constant terms determined using linear regression analysis, and may be different from constant terms associated with a situation where no heating operation is applied to the liquid or the pump is operating at or above the predetermined amount of power.

[0087] In the model discussed above, the water flow delay can be a normalized value. This normalization can be applied to solve the physical changes between the beverage preparation devices of the same type. For example, several beverage preparations of the same type may actually show changes in behavior between different devices in the device, for example, due to changes in the performance of pumps, power transmission circuits, heaters, etc. Therefore, when manufacturing a given beverage preparation device, normalization can be performed by performing several beverage preparation operations under the highest operating voltage of the device when no beverage preparation ingredients / capsules are used. In the case of a machine designed to work under the mains voltage of most Western European countries / regions, this voltage can be 265V. Since the water flow delay is minimum when the voltage is the highest, these beverage preparation operations will reveal the minimum possible water flow delay of the pump of a specific beverage preparation device. When the model is applied to control the pump during the operation use of the beverage preparation device, the normalization of the later value may include deducting the minimum possible water flow delay of the pump of the specific beverage preparation device.

[0088] In the model discussed above, the power supply voltage value can be a normalized value. This normalization can be applied to solve the physical changes between the beverage preparation devices of the same type. For example, several beverage preparations of the same type may actually show changes in behavior between different devices in the device, for example, due to changes in the performance of pumps, power delivery circuits, heaters, etc. Therefore, when manufacturing a given beverage preparation device, when the controlled voltage with two or more different voltage values ​​is supplied, normalization can be performed by testing the voltage measured by the specific beverage preparation device. For example, the voltage measurement of the specific beverage preparation device can be measured under each of 195V, 230V and 265V. Then determine the offset between the measured voltage at each supply value and the known supply voltage, and the offset can be used to adjust the measured voltage during the operation of the beverage preparation device. This type of measurement can be performed multiple times under each known power supply voltage and the average offset can be determined. In order to solve the possibility that the offset has different amplitudes and / or different percentages of the actual power supply voltage at different measurement points, interpolation or fitting can be applied to find the appropriate offset value applied at any given measurement power supply voltage during operation.

[0089] As discussed, each trained regression model may include one or more predetermined model parameters. In addition, generating an estimate of the volume may include: inputting the water flow delay and an average value of the AC power voltage supplied to the pump into the trained regression model; and performing one or more mathematical operations using the one or more predetermined model parameters, the water flow delay, and the average value of the AC power voltage.

[0090] like Fig.10 As shown, in step S1001 , training data is obtained, the training data including the value of the water flow delay of the test device and the volume of liquid output by the test device during operation of the test device.

[0091] At step S1002 , a regression analysis is performed on the training data to provide one or more parameters that can be used to estimate the volume of liquid output from the electric pump based on a measurement of the water flow delay of the electric pump.

[0092] In the example method 1000 , the test device corresponds to an electric pump and the water flow delay is a period between a first time corresponding to a first water flow drawn by the test device and a second time corresponding to a second water flow drawn by the test device.

[0093] In some examples, the test device includes a plurality of test beverage preparation devices. Training may include operating the machine to perform a predefined number of extractions in accordance with a test protocol that forces the pump to operate across a variety of different set points (e.g., different performances). During this training, there may be a physical flow meter that measures the actual flow rate, and other variables including water flow delays are also recorded. The data may be analyzed, and different types of regression models (linear regression models in the example) may be trained (with different parameter combinations) to fit the measured flow from the physical flow meter. For different regression models (i.e., models with different parameters), performance indicators (e.g., the proportion of extractions that exceed a predefined tolerance compared to the physical flow meter) are compared, and the regression model is optimized and a model of appropriate performance (such as the best performance model determined by the relevant operational considerations of any given specific implementation) is selected for use.

[0094] In some examples, evaluating the performance of the selected regression model may include implementing the selected regression model within firmware of the beverage preparation device. The device may operate with the electric pump controlled by the model rather than a physical flow meter, and the physical flow meter may serve as ground truth for evaluating the performance of the model.

[0095] The present technique can provide an accurate estimate of the volume of liquid output from an electric pump during operation of the electric pump. Fig.11 An example diagram showing the performance of this technology is shown. Specifically, Fig.11 The relative error for each draw is shown for different voltages (in this example at 195V, 213V, 230V, 247V and 265V). The relative error is calculated based on comparing the cumulative volume for a first device that uses a dedicated flow meter to measure flow and calculate the cumulative volume over time and for a second device that does not use a dedicated flow meter but uses the present technology to estimate flow and output cumulative volume over time. As an example, Fig.11 The relative error of the volume prediction of liquid output from a pump using the disclosed technique is shown compared to an apparatus using a dedicated flow meter to measure the volume of liquid output from the pump over 5438 iterations. Fig.11 As shown, in these iterations, 90% of the volume predictions were within ±10% relative error, 96% of the volume predictions were within ±12% relative error, and 99% of the volume predictions were within ±15% relative error.

[0096] It will be appreciated that a liquid dispensing machine comprising an electric pump and a processing circuit configured to perform the methods described herein may be provided instead of Figure 1 A beverage preparation device 100.

[0097] Therefore, from one perspective, a method for estimating the volume of liquid output from an electric pump during an operating period of the electric pump is described, the method comprising: determining a water flow delay of the pump, wherein the water flow delay is a period between a first time corresponding to a first water flow drawn by the pump and a second time corresponding to a second water flow drawn by the pump; and generating an estimate of the volume of liquid output from the electric pump during the operating period based on the water flow delay.

[0098] Additional examples of this approach are listed in the numbered clauses below.

[0099] Item 1. A method for estimating the volume of liquid output from an electric pump during an operating period, the method comprising: determining a water flow delay of the pump, wherein the water flow delay is a period between a first time corresponding to a first water flow drawn by the pump and a second time corresponding to a second water flow drawn by the pump; and generating an estimate of the volume of liquid output from the electric pump during the operating period based on the water flow delay.

[0100] Clause 2. The method of clause 1, wherein the first water flow substantially corresponds to a zero crossing of an AC power voltage supplied to the pump.

[0101] Clause 3. The method of any preceding clause, further comprising: determining an average value of an AC power supply voltage supplied to the electric pump, wherein generating the estimate of the volume is further based on the average value of the AC power supply voltage.

[0102] Clause 4. A method according to any of the preceding clauses, wherein determining the water flow delay of the pump comprises: determining the first time based on a measurement of the AC power voltage supplied to the pump during the operating period; determining the second time based on a measurement of the water flow drawn by the pump during the operating period; and calculating the difference between the first time and the second time as the water flow delay of the pump.

[0103] Clause 5. A method according to any preceding clause, wherein the second time corresponds to a time near the end of or within a time period during which the flow of water drawn by the pump has a substantially constant rate of change.

[0104] Clause 6. A method according to any preceding clause, wherein generating the estimate of the volume of liquid output from the pump is further based on a technique selected from a plurality of techniques, the selection being based on an indication of whether the power of the pump during the operating period of the pump is above a predefined power threshold.

[0105] Clause 7. A method according to clauses 1 to 5, wherein generating the estimate of the volume of liquid output from the pump is further based on a technique selected from a plurality of techniques, the selection being based on an indication of whether the temperature of the liquid is greater than a predefined temperature threshold or whether a heating operation is to be applied to the liquid.

[0106] Clause 8. A method according to clauses 1 to 5, wherein generating the estimate of the volume of liquid output from the pump is further based on a technique selected from a plurality of techniques, the selection being based on an indication of whether the duration of the operating period of the pump is greater than a predefined duration threshold.

[0107] Clause 9. A method according to clauses 6 to 8, wherein the operating period of the pump is a first operating period, the water flow delay is a first water flow delay, and the estimated value of the volume is a first estimated value of the volume, the method further comprising: in response to determining that a second operating period of the electric pump has passed, determining a second water flow delay of the electric pump, and generating a second estimated value of the volume of liquid output from the pump during the second operating period based on the second water flow delay; and calculating the total volume of liquid output from the pump during the first operating period and the second operating period by summing the first estimated value of the volume and the second estimated value of the volume.

[0108] Clause 10. A method according to clause 9, wherein the technology is a first technology, and wherein the second estimate of the volume is generated based on a second technology selected from the plurality of technologies, the selection being based on at least one of: an indication of whether the power of the pump during the second operating period of the pump is higher than the predefined power threshold, an indication of whether the temperature of the liquid is greater than the predefined temperature threshold or whether a heating operation is to be applied to the liquid, and an indication of whether the duration of the first operating period and the second operating period of the pump is greater than the predefined duration threshold.

[0109] Clause 11. The method of clause 10, wherein the first technique and the second technique are the same.

[0110] Clause 12. The method of clause 10, wherein the first technology and the second technology are different.

[0111] Clause 13. A method according to clauses 6 to 8, wherein the operating period of the pump is a first operating period, and the method further includes: in response to determining that the second operating period of the pump has passed, and determining that the difference between the average value of the AC power supply voltage in the first operating period and the average value of the AC power supply voltage in the second operating period is greater than a predefined difference: determining a second water flow delay of the electric pump; generating an estimated value of the second water flow delay of the electric pump based on the average value of the AC power supply voltage in the second operating period and a second technology; subtracting the estimated value of the second water flow delay from the second water flow delay to calculate an adjusted water flow delay; and generating an estimated value of the volume of liquid output from the pump during the second operating period based on the adjusted water flow delay and a third technology.

[0112] Clause 14. A method according to clauses 6 to 13, wherein the technique, the first technique, the second technique and / or the third technique is a trained regression model trained based on a regression analysis performed on training data, wherein the training data includes the value of the water flow delay of the test device and the volume of the liquid output by the test device during operation of the test device, wherein the test device corresponds to the electric pump.

[0113] Clause 15. A method according to clause 14, wherein the trained regression model includes one or more predetermined model parameters, and wherein generating the estimated value of the volume includes: inputting the average value of the water flow delay and the AC power voltage supplied to the pump into the trained regression model; and performing one or more mathematical operations using the one or more predetermined model parameters, the water flow delay and the average value of the AC power voltage.

[0114] Clause 16. A liquid dispensing machine comprising an electric pump and processing circuitry configured to perform the method of clauses 1 to 15.

[0115] Clause 17. The liquid dispensing machine of Clause 16, wherein the liquid dispensing machine is configured to cease dispensing liquid in response to determining that the estimate of the volume of liquid output from the pump is equal to a predetermined threshold volume.

[0116] Clause 18. A computer readable medium comprising instructions that, when executed by a programmable liquid dispensing machine, cause the programmable liquid dispensing machine to perform the method of clauses 1 to 15.

[0117] Item 19. A method for training a regression model to estimate the volume of liquid output from an electric pump during an operating period of the electric pump, the model comprising: obtaining training data, the training data comprising values ​​of a water flow delay of a test device and the volume of liquid output by the test device during operation of the test device, wherein: the test device corresponds to the electric pump; and the water flow delay is a period between a first time corresponding to a first water flow drawn by the test device and a second time corresponding to a second water flow drawn by the test device; and performing regression analysis on the training data to provide one or more parameters that can be used to estimate the volume of liquid output from the electric pump based on measurements of the water flow delay of the electric pump.

[0118] Item 20. A beverage preparation device, comprising: a fluid input end, the fluid input end being connected to receive a fluid for beverage preparation; an electric pump, the electric pump being connected to the fluid input end and having a control input end; a water flow sensor, the water flow sensor being connected to a power connection of the electric pump; a beverage preparation output end, the beverage preparation output end being connected to receive a fluid dispensed by the electric pump; and a controller, the controller being connected to the water flow sensor and the control input end, wherein the controller is configured to calculate the volume of fluid dispensed by the electric pump based on a water flow drawing delay measured by the water flow sensor, and to send a control signal to the controller input end to deactivate the electric pump in response to the fluid volume reaching a threshold volume.

[0119] The methods discussed above may be performed under the control of a computer program executed on a computing device (e.g., a programmable beverage preparation device or a programmable liquid dispensing machine). Thus, the computer program may include instructions for controlling a computing device to perform any of the methods discussed above. The program may be included in a computer-readable medium. The computer-readable medium may include non-transitory type media, such as physical storage media, such as storage disks and solid-state devices. The computer-readable medium may additionally or alternatively include transient media such as carrier signals and transmission media, which may, for example, transmit instructions between multiple separate computer systems and / or between components within a single computer system.

[0120] Thus, the present teachings provide a method for estimating the flow rate of a liquid output by an electric pump so as to provide an accurate estimate without the need for a flow meter. These techniques, while based on the use of a water flow delay of the pump, can be adapted by using complementary techniques and methods to account for complexities such as liquid temperature, variable quality of supply voltage, etc. The estimates obtained by these methods can be used to control when to shut off a pump for dispensing a measured or predetermined amount of liquid, which may find utility in devices such as beverage preparation machines, where a defined amount of liquid is prescribed for any given beverage prepared by the machine.

[0121] The various embodiments described herein are presented only to aid understanding and teach the claimed features. These embodiments are provided only as representative samples of embodiments and are not exhaustive and / or exclusive. It should be understood that the advantages, embodiments, examples, functions, features, structures and / or other aspects described herein should not be regarded as limitations on the scope of disclosure defined by the claims or limitations on the equivalents of the claims, and other embodiments may be utilized and modifications may be made without departing from the scope and / or spirit of the claims.

Claims

1. A method for estimating the volume of liquid output from an electric pump during an operation period of the electric pump, the method comprising: determining a water flow delay of the pump, wherein the water flow delay is a period of time between a first time corresponding to a first water flow drawn by the pump and a second time corresponding to a second water flow drawn by the pump; as well as Based on the water flow delay, an estimate of the volume of liquid output from the electric pump during the operating period is generated.

2. The method of claim 1, wherein the first water flow substantially corresponds to a zero crossing of an AC power supply voltage supplied to the pump.

3. The method according to any preceding claim, further comprising: An average value of an AC power supply voltage supplied to the electric pump is determined, wherein generating the estimate of the volume is further based on the average value of the AC power supply voltage.

4. The method of any preceding claim, wherein determining the water flow delay of the pump comprises: determining the first time based on a measurement of an AC power voltage supplied to the pump during the operating period; determining the second time based on a measurement of a flow of water drawn by the pump during the operating period; and The difference between the first time and the second time is calculated as the water flow delay of the pump.

5. A method according to any preceding claim, wherein the second time corresponds to a time towards the end of or within a time period during which the flow of water drawn by the pump has a substantially constant rate of change.

6. A method according to any preceding claim, wherein: Generating the estimate of the volume of liquid output from the pump is further based on a technique selected from a plurality of techniques, the selection being based on an indication of whether the power of the pump during the operating period of the pump is above a predefined power threshold; and / or Generating the estimate of the volume of liquid output from the pump is further based on a technique selected from a plurality of techniques, the selection based on an indication of whether a temperature of the liquid is greater than a predefined temperature threshold or whether a heating operation is applied to the liquid; and / or Generating the estimate of the volume of liquid output from the pump is further based on a technique selected from a plurality of techniques, the selection based on an indication of whether a duration of the operation period of the pump is greater than a predefined duration threshold.

7. The method of claim 6, wherein the operation period of the pump is a first operation period, the water flow delay is a first water flow delay, and the estimate of volume is a first estimate of volume, the method further comprising: responsive to determining that a second operating period of the electric pump has elapsed, determining a second water flow delay of the electric pump, and generating a second estimate of the volume of liquid output from the pump during the second operating period based on the second water flow delay; as well as A total volume of liquid output from the pump during the first operating period and the second operating period is calculated by summing the first estimate of the volume and the second estimate of the volume.

8. A method according to claim 7, wherein the technique is a first technique, and wherein the second estimate of the volume is generated based on a second technique selected from the plurality of techniques, the selection being based on at least one of: an indication of whether the power of the pump during the second operating period of the pump is above the predefined power threshold, an indication of whether the temperature of the liquid is greater than the predefined temperature threshold or whether a heating operation is to be applied to the liquid, and an indication of whether the duration of the first operating period and the second operating period of the pump is greater than the predefined duration threshold, and optionally wherein the first technique and the second technique are the same or different.

9. The method of claim 6, wherein the operation period of the pump is a first operation period, the method further comprising: In response to determining that a second operation period of the pump has elapsed and determining that a difference between an average value of the AC power supply voltage in the first operation period and an average value of the AC power supply voltage in the second operation period is greater than a predefined difference: determining a second water flow delay of the electric pump; generating an estimate of the second water flow delay of the electric pump based on the average value of the AC power supply voltage in the second operating period and a second technique; subtracting the estimated value of the second water flow delay from the second water flow delay to calculate an adjusted water flow delay; as well as An estimate of the volume of liquid output from the pump during the second operating period is generated based on the adjusted water flow delay and a third technique.

10. A method according to claims 6 to 9, wherein the technique, the first technique, the second technique and / or the third technique is based on a trained regression model trained by performing regression analysis on training data, the training data comprising values ​​of water flow delay of a test device and a volume of liquid output by the test device during operation of the test device, wherein the test device corresponds to the electric pump.

11. The method of claim 10, wherein the trained regression model comprises one or more predetermined model parameters, and wherein generating the estimate of the volume comprises: inputting the water flow delay and an average value of the AC power voltage supplied to the pump into the trained regression model; and One or more mathematical operations are performed using the one or more predetermined model parameters, the water flow delay, and the average value of the AC power voltage.

12. A liquid dispensing machine comprising an electric pump and a processing circuit configured to perform the method according to claims 1 to 11.

13. The liquid dispensing machine of claim 12, wherein the liquid dispensing machine is configured to cease dispensing liquid in response to determining that the estimate of the volume of liquid output from the pump is equal to a predetermined threshold volume.

14. A computer readable medium comprising instructions which, when executed by a programmable liquid dispensing machine, cause the programmable liquid dispensing machine to perform the method of claims 1 to 11.

15. A method for training a regression model to estimate the volume of liquid output from an electric pump during an operation period of the electric pump, the model comprising: obtaining training data, the training data comprising values ​​of a water flow delay of a test device and a volume of liquid output by the test device during operation of the test device, wherein: The test device corresponds to the electric pump; and The water flow delay is a period of time between a first time corresponding to a first water flow drawn by the test device and a second time corresponding to a second water flow drawn by the test device; and A regression analysis is performed on the training data to provide one or more parameters that can be used to estimate the volume of liquid output from the electric pump based on a measurement of water flow delay of the electric pump.

Images