Methods for determining foam during the treatment of laundry items and laundry care machines

Abstract

This invention relates to a method for determining foam during liquid treatment of laundry items in a laundry detergent machine. The laundry detergent machine includes: a housing and an lye container disposed therein for receiving liquid; a drum disposed in the lye container; a sensor continuously sensing operating data of the laundry detergent machine and supplying it to a control device connected to the sensor for controlling the laundry detergent machine based on the operating data; the control device determining foam based on the operating data. In this method, the control device includes a virtual sensor, to which operating data is continuously supplied, and the virtual sensor continuously calculates a measure of the foam present in the lye container based on the supplied operating data. The invention also relates to a laundry detergent machine configured to perform such a method.

Description

Technical Field

[0001] This invention relates to a method for determining foam during liquid treatment of laundry items in a laundry detergent machine, the laundry detergent machine comprising: a housing and an alkali container disposed therein for receiving the liquid; a motor-driven drum disposed in the alkali container, rotating about a rotation axis, and having an inner chamber for receiving the laundry items, wherein a sensor continuously receives / senses operating data of the laundry detergent machine and supplies it to a control device connected to the sensor for controlling the laundry detergent machine based on the operating data, wherein the control device determines the foam based on the operating data.

[0002] The present invention also relates to a laundry care machine for treating laundry items by means of a liquid, comprising: a housing and an alkali container disposed therein for receiving the liquid; a motor-driven drum disposed in the alkali container, rotatable about a rotation axis, and having an inner chamber for receiving the laundry items; a sensor for continuously receiving / sensing operating data of the laundry care machine; and a control device connected to the sensor for controlling the laundry care machine based on the operating data, wherein the control device is configured to determine the presence of foam in the alkali container based on the operating data. Background Technology

[0003] Document EP 0 997 570 A2 discloses a general method and a general laundry care machine as defined above.

[0004] Document DE 43 42 272 A1 discloses a method and measuring mechanism for determining the liquid level, turbidity, and foam content in a liquid used to treat laundry items in a laundry detergent machine. For this purpose, at least one optical beam is aimed at the surface of the liquid at an oblique angle of incidence, and the reflection of the beam is captured by a screen equipped with sensors. To determine the liquid level, turbidity, and foam, these reflections are then analyzed in terms of spatial distribution and amplitude. This method requires a calm and undisturbed surface of the liquid, and is therefore less suitable for use in laundry detergent machines due to the required movement of the liquid and the laundry items being treated.

[0005] Document EP 0 997 570 A2 discloses a computer-supported method and mechanism for determining clusters of foam formation in a laundry detergent machine. During the washing process in the laundry detergent machine, the pressure, temperature, and current liquid volume present therein are measured, and these measured parameters form an application vector. For these application vectors, fuzzy membership values ​​are determined for two pre-given clusters, one cluster indicating "foam is expected" and the other indicating "foam is not expected." To define the method, a control device is constructed using experimentally obtained model data sets, each set including a data vector from the aforementioned parameters and a membership value for each cluster. In practical application, the separately measured application vectors are supplied to the control device, which then provides membership values ​​for the two clusters based on these application vectors. Therefore, this method only allows for qualitative conclusions regarding the presence of foam in the laundry detergent machine. Conversely, quantitative conclusions are not possible.

[0006] Documents DE 102 34 472 A1 and DE 10 2010 028 614 A1 disclose methods and mechanisms for determining foam in laundry detergent machines, which can only be used in specific operating states of the laundry detergent machine. According to the first document, foam determination occurs after the lye solution is pumped out of the lye container of the laundry detergent machine (i.e., liquid is removed) by means of a lye pump. Here, the pressure in the lye container is measured using a pressure sensor, and the trend of these measurements is observed. If a large amount of foam is present in the lye container, this can be detected by a specific component in the trend of the measurements, and the presence of this component leads to the conclusion that foam is present in the lye container. In the second document, the power consumption of the lye pump is observed after the sub-process for treating the laundry items is completed, either when pumping out the liquid or (if the laundry items are being treated while the liquid is being circulated in the laundry detergent machine) while circulating the liquid. If the lye pump delivers foam instead of liquid, the power consumption of the lye pump decreases, and the power consumption of the lye pump is monitored accordingly. The conclusion that foam is present in the alkali container is then drawn from the potential decrease in alkali pump power consumption. However, these methods, based on the analysis of relatively small effects from operational data of laundry detergent machines, have limited reliability and are not very precise. Therefore, these methods do not allow for quantitative conclusions regarding the presence of foam.

[0007] Document DE 10 2015 205 949 A1 discloses a laundry detergent machine in which the presence of foam is verified by analyzing the drive current flowing through the motor that drives the drum during the washing of laundry items. Here, the drive current value during the drum rotation period is measured and compared with a reference value previously determined under the condition that foam is definitely not present. The relationship between the drive current value and the reference value leads to the conclusion that foam is present. This method also only provides a qualitative result, and for the same reasons as the methods in the prior art discussed above.

[0008] Documents EP 3670 729 A1 and EP 3 674 466 A1 disclose various applications of neural networks, particularly recurrent neural networks, in laundry detergent machines. EP 3670 729 A1 describes the prediction of microbial presence and subsequent growth, and derives information to process laundry in a laundry detergent machine. EP 3 674 466 A1 defines a method for processing laundry data obtained previously using a trained neural network.

[0009] The article titled "AMachine Learning-based Soft Sensor for Laundry Load Fabric Typology Estimation in Household Washer-Dryers," published by GA Susto et al. in IFAC PapersOnLine 52, 11 (2019) 116, describes a "soft sensor" for distinguishing different types of fabrics in laundry items to be processed in a washer-dryer. The "soft sensor" consistently utilizes existing sensors and signals from the washer-dryer. A regularized linear model is used here to analyze the data from the processed sensors and signals.

[0010] Directly and quantitatively measuring the foam level in the lye container of a laundry detergent machine during the programmed process of handling laundry items is practically impractical in operation because the handling of laundry items requires more or less vigorous movement. Apart from occasional static phases, the liquid being handled is in constant motion due to the rotation of the drum carrying the laundry items, and its periodic lifting and falling is facilitated by a drive mechanism connected to the drum, with the drum rotating about a substantially horizontal axis. This motion of the laundry items and the liquid is also transmitted to the foam, making it impossible to measure the level / degree of foam in the lye container during the drum's movement. Therefore, measuring the level of foam in the lye container is impractical.

[0011] Nevertheless, the presence of varying amounts of foam in the lye container determines the progress and success of the laundry care process. During direct washing, the abundant foam affects the mechanical handling of the garments by damping their movement and hindering the redistribution of the liquid used for washing (especially the lye). After actual washing, the present foam hinders the pumping of lye from the container, both because it drips very slowly from the garments due to its viscosity and because the lye pump's delivery capacity is poor. The subsequent rinsing process is impeded by the abundant foam, which retains components of the lye in the container, thus affecting its success. Finally, the foam also affects the spin-drying process, which typically concludes laundry care, particularly by hindering the removal of liquid from the lye container and by increasing friction on the rapidly rotating drum, potentially preventing it from reaching the predetermined speed.

[0012] Therefore, there is a need for methods and laundry care machines that allow quantitative measurement of foam in a lye container and provide a quantitative measure (Maß) of the current foam, and that allow such measurement at most arbitrary points in time during the ongoing laundry care process. Summary of the Invention

[0013] Accordingly, the object of the present invention is to provide a method and a laundry care machine of the categories defined in the preamble, which respectively allow quantitative measurement of foam in the lye container and provide a quantitative measure of the current foam at most arbitrary points in time during the ongoing laundry care process.

[0014] To solve this task, a method according to the corresponding invention is proposed. A laundry care machine according to the corresponding invention is also proposed to solve this task. This laundry care machine is specifically configured to perform the method according to the invention.

[0015] Preferred improvements to the present invention are set forth in the technical solutions of the present invention and in the following description, and may be used in combination with each other, provided that technical circumstances permit, even if not precisely listed herein. Preferred improvements to the laundry care machine according to the present invention correspond to preferred improvements to the method according to the present invention, and vice versa, even if not precisely listed herein.

[0016] To address this task, a method for determining foam during liquid treatment of laundry items in a laundry detergent machine is correspondingly proposed according to the present invention. The laundry detergent machine includes: a housing and an alkali container disposed therein for receiving the liquid; a motor-driven drum disposed in the alkali container, rotating about a rotation axis, and having an inner chamber for receiving the laundry items. A sensor continuously senses operating data of the laundry detergent machine and supplies it to a control device connected to the sensor for controlling the laundry detergent machine based on the operating data, wherein the control device determines the foam based on the operating data. In this method, the control device includes a virtual sensor to which the operating data is continuously supplied, and the virtual sensor continuously calculates a measure of the foam present in the alkali container based on the supplied operating data.

[0017] To address this task, a laundry care machine for treating laundry items using a liquid is also proposed according to the present invention, comprising: a housing and an alkali container disposed therein for receiving the liquid; a drum disposed in the alkali container, rotatable about a rotation axis and driven by a motor, having an inner chamber for receiving the laundry items; a sensor for continuously sensing operating data of the laundry care machine; and a control device connected to the sensor for controlling the laundry care machine based on the operating data, wherein the control device is configured to determine the presence of foam in the alkali container based on the operating data. Here, the control device includes a virtual sensor to which the operating data can be continuously supplied, and the virtual sensor is configured to continuously measure the presence of foam in the alkali container based on the supplied operating data during laundry treatment.

[0018] Therefore, according to the present invention, the presence of foam in the lye container is indirectly measured by providing a virtual sensor that determines a quantitative measure of the foam present in the lye container based on specific and meaningfully measurable operational data of laundry care. This measure (or metric) is particularly a measure used in model tests on a laundry machine where operation can be interrupted to measure foam in the lye container, and simultaneously determined with corresponding operational data in such model tests. This measure may be the height of the foam exceeding the height of the laundry lye that briefly settles in the lye container during the interruption. The correlation between the measure measured in the model tests and the simultaneously measured operational data is then implemented in the virtual sensor using algorithms, tables, or the like, so that the virtual sensor can derive the desired measure from the corresponding operational data during actual laundry care and provide it for controlling the laundry care process. In such actual operation, the measure is a theoretical value that cannot be directly replicated based on the movement process in the laundry machine, but still accurately describes the amount of foam present in the laundry machine.

[0019] Within the framework of this invention, the measurement is preferably expressed as the height or volume of foam in the drum. Here, the height of the foam can be determined, in particular, as the height at which the foam appears within the window of the door that seals the lye container of the laundry detergent.

[0020] One advantage of this invention is that the measurement provides a quantitative conclusion about the presence of foam at each point in time during the actual laundry care process. This quantitative conclusion is particularly permissible when controlling the laundry care machine, whereby adjustments can be made, in particular, using the measurement as an initial parameter, by different adjusting parameters (such as the amount and temperature of the liquid to be achieved in the lye container and the metering of detergent in the lye container). Here, there is no need to pursue the near or complete avoidance of foam through specially arranged countermeasures; the presence of a certain amount of foam, for example as a sign of adequate detergent metering, may be worthwhile depending on the desired washing process. Furthermore, this invention allows for the observation of the effects of measures, enabling the interruption, extension, repetition of a measure already taken, or the subsequent following of another measure. It is also feasible to analyze the temporal trend of the measurement and thereby derive predictions of foam occurrence in subsequent stages of laundry care.

[0021] According to a preferred improvement of the invention, the virtual sensor also determines the time variation curve of the metric, in particular the time derivative of the metric, and the control device uses the same time variation curve to control the laundry care machine.

[0022] According to another preferred improvement of the invention, the virtual sensor further determines a prediction of the time-varying curve of the metric, and the control device uses the prediction to control the laundry care machine. It is also preferred that the prediction is an estimate of the future metric, wherein the estimate corresponds to the metric after a specific period of time, between 10 seconds and 10 minutes (particularly between 2 minutes and 8 minutes, preferably about 5 minutes), has elapsed following the prediction. Thus, the virtual sensor not only provides the metric itself and its time derivative, but also extrapolates the metric to a specific future time point, preferably about five minutes for a typical 1 to 3-hour laundry care process, and provides the expected metric at that time point as a prediction. This significantly improves the determination of measures to prevent excessive foam generation. It may also be meaningful to direct the prediction to a future time point of 10 to 30 seconds, particularly about 20 seconds, especially during the laundry care process phase in which the drum rotates in opposite directions of rotation during the mentioned duration, thereby obtaining and using a prediction from one reverse cycle to the next. It is also conceivable that such predictions are set over a long period of 1 to 2 hours in order to obtain and use conclusions about the expected changes in the curve during the washing process as early as possible.

[0023] According to another preferred embodiment of the invention, the virtual sensor comprises a trained neural network that determines the metric. More preferably, the trained neural network is a recurrent neural network, particularly a trained neural network comprising a trained Elman network.

[0024] Recurrent neural networks (RNNs) differ from conventional neural networks, which transmit information in only one direction and are therefore often called "feed-forward networks," in the presence of feedback. This feedback involves one or more data points and can handle the temporal correlation of initial parameters (i.e., the correlation between initial parameters that are directly or indirectly related in time), particularly using high-dimensional time series of initial data for modeling and predicting target parameters. Similar to the approach used in high-dimensional differential equations, trained RNNs can well describe the temporal behavior of dynamic systems. Therefore, the use of corresponding trained RNNs in virtual sensors can significantly improve the modeling and prediction of foam dynamics and the typically stable and mostly monotonous foam formation during laundry care compared to previous methods. Different recurrent network structures can be used to simulate and predict foam formation during laundry care, particularly nonlinear autoregressive networks (NARX), Elman networks, gated recurrent units (GRUs), and long short-term memory networks (LSTM).

[0025] According to a preferred improvement of the invention, the operating data is selected from: the power consumption of the motor, the power consumption of the lye pump, the force consumption on the bearings of the drum, the rotational state of the drum, the liquid level in the lye container, the composition of the liquid, the temperature of the liquid, and the aforementioned time variation curves (especially the time derivative). This is primarily based on the fact that adding detergent alters the viscosity and surface tension of the liquid. This creates the preconditions for foam formation and, consequently, alters the movement of the liquid in the lye container as the drum rotates or as the liquid is pumped. Furthermore, the foam binds to a portion of the liquid, thereby lowering the liquid level in the lye container. In addition, the foam dampens the movement of the items being washed in the drum. These changes are reflected in changes in the signal of the pressure sensor, as is commonly used to measure the liquid level in the lye container, and also in changes in the drive power (or drive torque) of the motor used to rotate the drum, and in changes in the drive power (or drive torque) of the lye pump used to circulate or pump out the liquid. The presence of foam between the laundry items, which may alter their condition within the drum, is also indicated by force sensors used to measure the forces, torques, and accelerations on the drum's bearings. Finally, the drum's own rotation can be monitored and analyzed using position sensors, particularly gyroscopes. The movement of the laundry items caused by the drum's rotation manifests as rotational non-uniformity, which can be measured and analyzed. This non-uniformity is reduced in the presence of foam due to its damping properties. When the drum reverses direction, if the actuator is switched off at the end of the reverse cycle, a larger or lesser movement of the drum occurs in the opposite direction to the actuator: the driven rotation of the drum causes the laundry items on the upward-moving side of the drum to move upward, thus causing an imbalance, which is balanced by the brief reverse movement of the drum after the actuator is switched off. Therefore, measuring this reverse movement can also provide conclusions about the presence of foam.

[0026] All the aforementioned effects of foam on the operating data of the laundry detergent machine act in parallel with other effects, such as: the fabric type, structure, and quality of the laundry; the amount, temperature, and composition of the liquid used for treatment; the movement of the drum in the reverse cycle; the type of motor adjustment; imbalances in the distribution of laundry in the lye container; and the response of the vibrating suspension of the lye container to such imbalances. Therefore, attempting to record these effects in a concrete physical model may be of little help. According to the invention, an indirect measurement of the amount of foam present in the lye container is obtained by analyzing data from model tests.

[0027] According to another preferred embodiment of the invention, the laundry detergent machine has at least one actuator, which is operated by the control device when controlling the laundry detergent machine and can reduce foam in the lye container. The control device performs at least one measure to reduce the foam when using the at least one actuator, based on the measurement. More preferably, the at least one measure is selected from: adding water to the liquid; stabilizing the temperature of the liquid; reducing the rotational speed of the drum; reducing the circulation pumping of the liquid in the lye container; pumping the liquid out of the lye container; and subsequently adding water to the lye container. Additionally preferably, the control device provides information about the at least one specific measure for the user of the laundry detergent machine to know.

[0028] According to another preferred embodiment of the invention, the treatment of the laundry items is carried out in a programmed laundry care process, wherein the control device determines the duration of the laundry care process based on a measurement of foam and at least one measure taken to reduce foam, and provides information about the duration to the user of the laundry care machine.

[0029] According to another preferred embodiment of the invention, the sensor includes at least one sensor selected from: a pressure sensor in the alkali container; a power sensor on the motor; a power sensor on the alkali pump; a force sensor on the bearing of the roller; and a sensor (especially a chemical sensor) in the alkali container for determining the chemical composition of the liquid.

[0030] The power sensor on the motor (or the pump) can be a sensor that detects an electrical parameter, which is a measure of the power currently consumed (or the torque currently generated). For example, such an electrical parameter could also be the current supplied to the motor (or the pump) for operation.

[0031] According to another additional preferred improvement of the invention, the laundry detergent machine has at least one actuator operable by the control device when controlling the laundry detergent machine and capable of reducing foam in the lye container, and the control device is configured to: determine and execute at least one measure for reducing the foam based on the metric when using the at least one actuator. More preferably, the at least one actuator is selected from: the motor; a valve for allowing water to enter the lye container; an lye pump for pumping the liquid out of the lye container; and a metering device for metering foam inhibitors into the lye container. The determination of the measure is particularly based on a comparison of a currently determined metric with one or more pre-given thresholds, thereby determining the measure when the threshold or a threshold is exceeded. For example, in the spin-drying process, when a small threshold is exceeded, the spin speed is reduced and the spin-drying process continues, while when a higher threshold is exceeded, the spin-drying process is interrupted, followed by another rinsing process and then a renewed spin-drying process. Foam inhibitors can primarily be traditional fabric softeners containing cationic surfactants, which effectively neutralize the anionic surfactants present in the detergent. These softeners can be dispensed on demand, for example, via the automatic metering system of the laundry machine.

[0032] According to another preferred embodiment of the invention, the laundry detergent machine has a display device, and the control device of the laundry detergent machine is configured to: provide information about the at least one specific measure via the display device for the user of the laundry detergent machine. Additionally preferably, the laundry detergent machine is configured to: process laundry items in a programmed laundry care process, and the control device of the laundry detergent machine is configured to: determine the duration of the laundry care process based on a measurement of foam and based on at least one measure determined to reduce foam, and provide information about the duration via the display device for the user. The display device can be directly integrated into the laundry detergent machine, or it can be spatially separated from the laundry detergent machine and connected to it via a data network (e.g., the Internet, WLAN, or Bluetooth). In the latter case, the display device can be implemented on a mobile communication device (e.g., a smartphone) via a corresponding application. These two possibilities can be combined.

[0033] According to yet another preferred embodiment of the invention, the virtual sensor is generated from multiple model data sets via machine learning, wherein each model data set has a model value associated with the operational data and a metric associated with these model values. More preferably, the virtual sensor is configured to determine the metric by means of regression from these model data sets.

[0034] In principle, this invention can be used in any type of laundry and garment care machine, and in particular, the cost required to apply this invention is extremely low. Therefore, in addition to conventional washing machines, washer-dryers are also under consideration. Here, the correct spatial orientation of the rotation axis of the drum is not important. The rotation axis can be described in detail below as being substantially horizontally oriented, but it can also be vertical or oriented at any angle to the vertical. Attached Figure Description

[0035] Embodiments of the invention will now be further explained with reference to the accompanying drawings. The drawings essentially only show components of the laundry detergent machine or virtual sensor that are important for the following description. The drawings show:

[0036] Figure 1 : A schematic front view of a vertical cross-section of a laundry detergent machine, in which the lye container of the laundry detergent machine is also cut;

[0037] Figure 2 A schematic front view of a vertical section of a laundry detergent machine;

[0038] Figure 3 A schematic side view of a vertical section of a laundry detergent machine;

[0039] Figure 4 : Used in applications based on Figures 1 to 3 A diagram illustrating the function of a first embodiment of a virtual sensor in a laundry detergent machine;

[0040] Figure 5 : Used in applications based on Figures 1 to 3 A diagram illustrating the function of a second embodiment of a virtual sensor in a laundry detergent machine;

[0041] Figure 6 A schematic view of an Elman network; and

[0042] Figure 7 Different temporal trends of foam formation in laundry detergent machines when using Elman networks in virtual sensors to determine foam. Detailed Implementation

[0043] Figures 1 to 3A schematic embodiment is shown of treating laundry items 1 with a liquid 2, which tends to form foam 3 in such treatment. These laundry items 1 are located in a laundry care machine 4, which includes: a housing 5 and an lye container 6 disposed therein for receiving the liquid 2; a drum 8 disposed in the lye container 6 having an inner chamber 11 for receiving the laundry items 1; a carrying member 9 for lifting and lowering the laundry items 1; and a bottom shell 7. The drum 8 is rotatable about a rotation axis 10 and has a journal 12 supported in a bearing 13 of the lye container 6. The drum 8 can be driven by a motor 14 via the journal 12, particularly by a corresponding... Figure 3 The schematic view shows two drive wheels and a belt.

[0044] Figure 1 and Figure 2 It has essentially the same content, only Figure 1 The cross-section through the alkali container 6 is shown, while Figure 2 The alkali container 6 is shown in the front view. According to... Figure 2 A ring 26 is provided on the front side of the alkali container 6. The ring 26 is made of easily deformable rubber elastic material and is used to make a liquid-sealed connection between the alkali container 6 and the shell 5. Figure 3 A schematic example of implementing such a ring 26 can be seen here. It is also shown here how the laundry care machine 4 closes by means of a door 27 provided on the ring 26 after the laundry item 1 is placed inside.

[0045] Two lye pumps 25 and 29 are connected to the bottom shell 7 of the lye container 6. Lime pump 25 is used to pump out the liquid 2 at the end of the laundry care process, or at the end of a certain step in such a process, and after completion. Lime pump 29 is used to circulate the liquid 2 in the lye container, that is, to circulate the liquid 2, with the aim of ensuring that the laundry item 1 is better exposed to the liquid 2. Figure 3 The alkaline pump 29 is not shown for clarity.

[0046] The control device 15 is implemented, in particular, as an electronic device and is used to control the programmed laundry care process, the implementation of which is determined by the laundry care machine 4 based on individual user selections and pre-defined parameters. This control device is connected to all components of the laundry care machine 4 that must be controlled, for detecting operational data of the ongoing laundry care process, or for communicating with the individual user. The type of each corresponding connection depends on the specific circumstances; the connection can be, in particular, mechanical or electrical, and can be configured as wired or wireless.

[0047] The control device 15 also includes a virtual sensor 16, to which specific operational data of the ongoing laundry care process is continuously supplied. The virtual sensor 16 is configured to continuously measure the amount of foam 3 present in the lye container 6 based on the supplied operational data while processing the laundry item 1. Based on this measurement, the laundry care process is monitored and controlled by the control device 15, either to prevent excessive foam formation or to confirm the presence of a pre-defined amount of foam as an indicator of the presence of a pre-defined effective amount of detergent (particularly surfactant) in the liquid 2.

[0048] The measurement is preferably expressed as the height or volume of foam in the roller. Within the framework of these embodiments, the height of the foam is defined as the height at which the foam appears in the window of the door 27 that closes the alkali container 6.

[0049] In addition to measuring the foam 3 itself, the control device 15 can also determine the time variation curve of the measurement (in particular the time derivative) and use it to control the laundry care machine 4.

[0050] In addition to the time-varying curve of the metric, which can be specifically presented by the time derivative of the metric, the virtual sensor 16 also determines a prediction of the time-varying curve of the metric, and the control device 15 uses the prediction to control the laundry detergent machine 4. Here, the prediction is an estimate of the metric for a future "macro" period, wherein the estimate specifically corresponds to the metric after the prediction, at the end of a specific period between 10 seconds and 10 minutes (particularly between 2 minutes and 8 minutes, preferably about 5 minutes). Such an estimate also does not take into account the short-term fluctuations that the metric may experience due to the large movements within the laundry detergent machine 4.

[0051] It may also be meaningful to direct the prediction to a future time point 10 to 30 seconds later (especially after about 20 seconds), particularly during the laundry care process phase in which the drum 7 rotates in opposite directions of rotation during the mentioned duration, thereby obtaining and using the prediction from one reverse cycle to another.

[0052] It is also conceivable to set such forecasts for a long period of time, between 1 and 2 hours, in order to obtain and use information about the expected changes in the laundry during the washing process as early as possible.

[0053] As described more precisely based on the accompanying drawings, the virtual sensor 16 includes a trained neural network that determines the metric. This neural network is, in particular, a recurrent neural network, which not only applies to the metric but also models its temporal trend and makes predictions.

[0054] Operating data used to determine the measurement of the foam 3 includes, in particular: the power consumption of the motor 14, the power consumption of the lye pump 25 or 29, the force consumption on the bearing 11 of the drum 8, the rotational state of the drum 8, the liquid level of the liquid 2 in the lye container 6, the composition of the liquid 2, the temperature of the liquid 2, and their time derivatives. Sensors 17, 18, 19, 20, 25, and 29 are provided to continuously receive the operating data of the laundry detergent 4, and the control device 15 is connected to these sensors in an appropriate manner. These sensors 17, 18, 19, 20, 25, and 29 include: a pressure sensor 17 in the lye container 6, a power sensor 18 on the motor 14, power sensors on the lye pumps 25 and 29, a force and position sensor 19 on the bearing 13 of the drum 8, and a sensor 20 in the lye container 6 for determining the chemical composition or temperature of the liquid 2. These lye pumps 25 and 29 also function as sensors and are each equipped with a power sensor, which is not shown for clarity. According to conventional practice, the pressure sensor 17 is used to determine the liquid level 2 in the lye container 6, and is particularly used to set and / or adjust the liquid level according to pre-given parameters of each laundry care process to be performed. The power sensor 18 may be a sensor for measuring current supplied to the motor 14 for operational purposes and thus serving as a measure of the currently consumed power (or currently generated torque) when the voltage across the motor 14 is constant. The force sensor 19, according to conventional practice, is particularly used to measure the mass of the drum 8 with the laundry items 1 placed inside, and the acceleration generated by the imbalance due to the not completely uniformly distributed laundry items 1 when the drum 8 rotates rapidly (especially during spin-drying). It is generally not assumed that the laundry items 1 are so uniformly distributed in the drum 8 during spin-drying that no static and dynamic imbalance (i.e., oscillating forces and torques) occurs during rotation. Therefore, the alkali container 6 with the roller 8 is typically suspended vibratingly within the housing 5. A corresponding vibration system with springs and damping elements is not shown in the figures, but can be imagined based on relevant practice.

[0055] To determine the measurement of the foam 3, all of these sensors 17, 18, 19, 20, 25, 29, or only a subgroup of them, can currently be used. Here, the operational data obtained in the control device 15 via the corresponding sensors 17, 18, 19, 20, 25, 29 is also supplied to the virtual sensor 16 to derive quantitative conclusions about the presence of foam 3 in the alkali container 6. Figure 1A static lye container is shown, allowing the liquid 2 to form a calm level on the laundry item 1, with a layer of foam 3 settled on this level. This static arrangement is rarely observed during the laundry process. Therefore, continuously and directly measuring the amount of foam 3 (e.g., by measuring the height of the foam layer on the liquid 2) is impractical and inaccurate. Therefore, in this invention, the virtual sensor 16 derives a measure of the foam 3 present in the lye container 6 from operational data that can be meaningfully and accurately measured, instead of direct measurement. Here, the correlation between the operational data and the measure is derived from empirical data obtained in model experiments (particularly through machine learning methods). The virtual sensor 16 is generated by machine learning from multiple sets of model data, each set having a model value associated with the operational data and a measure associated with those model values. Furthermore, the virtual sensor 16 is configured to determine this measure from these sets of model data using regression.

[0056] An example of the effect of foam 3 in the components of the laundry detergent 4 is simply presented in the effect observed on the lye pump 29 during the circulation pumping of the liquid 2: drawing the liquid 2 from the lye container 6 and returning it to the lye container 6 requires a certain amount of electrical power, which can also be verified by the corresponding power sensor on the lye pump 29. However, if a sufficiently high proportion of the liquid 2 is converted into foam 3 during the laundry process, the lye pump 29 can pump the remaining liquid 2 more or less completely out of the bottom shell 7, allowing the bottom shell 7 to receive and transport the foam 3 again. Here, due to its low density, the electrical power required by the lye pump 29 decreases, and its rotational speed may also increase. Thus, the electrical power consumed by the lye pump 29 and its rotational speed significantly indicate the presence of foam 3 in the lye container 6. It may be meaningful to observe the time trend of the power consumed from the corresponding start-up of the lye pump 29. The power consumption corresponding to the delivery of liquid 2 is shown first here. Once foam 3 is delivered instead of liquid 2, the power consumption will decrease more or less suddenly.

[0057] according to Figures 1 to 3The laundry detergent machine 4 also has actuators 14, 22, 24, 25, 29, which can be manipulated by the control device 15 when controlling the laundry detergent machine 4 and, through their application, can affect (in particular reduce) the foam 3 in the lye container 6. The control device 15 is configured to: according to the measured values, use at least one actuator 14, 22, 24, 25, 29 to perform at least one measure to affect (in particular reduce) the foam 3. These actuators 14, 22, 24, 25, and 29 are: the motor 14; a valve 22 for allowing water to enter the lye container 6 via the rinsing chamber 23; an lye pump 25 for pumping the liquid 2 out of the lye container 6; an lye pump 29 for circulating the liquid 2; and a metering device 24 for metering foam inhibitor into the lye container 6 via the rinsing chamber 23.

[0058] The operation of the roller 8 can be controlled by the motor 14 in response to the presence of foam 3 in the lye container, by rotating the roller 8 faster, slower, and / or in shorter or longer time intervals. By introducing water into the lye container 6, the liquid 2 therein can be diluted and / or cooled to affect the foam. Furthermore, the temperature of the liquid 2 can be altered (particularly intentionally increased) by a heating mechanism not shown in the figures to affect the foam 3. The lye pump 25 can be used to pump a portion of the liquid 2 out of the lye container 6 and thereby advantageously influence the quantity and composition of the liquid 2 (in conjunction with the valve 22 if necessary). The operation of the lye pump 29 can be reduced or completely stopped in the presence of foam 3 to prevent the formation of more foam 3. The metering device 24 can be configured, within the framework of conventional practice, to introduce a fabric softener into the lye container 6 for the rinsing process. Such fabric softeners typically contain cationic surfactants, which can be used to neutralize the anionic surfactants present in the liquid 2 that cause foam 3 and thereby affect the formation of foam 3.

[0059] Therefore, when controlling the laundry detergent machine 4, the control device 15 manipulates at least one actuator 14, 22, 24, 25, 29 to perform at least one measure to influence (in particular reduce) the foam 3 when using the actuators 14, 22, 24, 25, 29, according to the measurement. This measure specifically includes one of the following: adding water to the liquid 2; stabilizing the temperature of the liquid 2; reducing the rotational speed of the drum 8; reducing the circulation pumping of the liquid 2; pumping the liquid 2 out of the lye container 6; and subsequently adding water to the lye container 6.

[0060] The laundry detergent 4 also has a display device 21, wherein the control device 15 is configured to: provide information via the display device 21 regarding at least one specific measure for influencing the foam 3, for the user of the laundry detergent 4 to know. Additionally, the control device 15 is configured to: determine the duration of a programmed laundry care process based on measurements of the foam 3 and based on specific measures determined to influence (particularly reduce) the foam 3, and provide information regarding this duration via the display device 21 for the user to know.

[0061] Figure 4 It shows the application in accordance with Figures 1 to 3 A diagram illustrating the function of the virtual sensor 6 in the laundry detergent 4. The virtual sensor 16 is supplied with operating data of the laundry detergent 4, determined by the sensors 17, 18, 19, and 20 and the corresponding sensors on the lye pumps 25 and 29, for executing the programmed laundry detergent process, after preprocessing (e.g., digitization and / or time differentiation) where necessary. The virtual sensor 16 then determines a measurement for the foam 3 based on this operating data as described above. This measurement can be displayed on the display device 2 for user viewing, with additional information added where necessary, such as the measures selected based on the measurement to influence the foam 3. This measurement is also supplied to the drive unit 28, which, like the virtual sensor 16, belongs to the control unit 15, and is further processed therein by generating signals for controlling the actuators 14, 22, 24, 25, and 29 and outputting them to these actuators.

[0062] Figure 5 It shows the application in accordance with Figures 1 to 3 A diagram illustrating the function of virtual sensor 16 in a laundry detergent machine. This diagram is related to... Figure 4 The function diagram is consistent with that of the virtual sensor 16, except for feedback loop 30, which displays the feedback from the virtual sensor 16. Through this type of feedback, the trained neural network in the virtual sensor 16 is a trained recurrent neural network and processes not only its initial parameters 17 to 20, 25, and 29 but also their time-varying curves. This allows the virtual sensor 16 to be used not only for the metric itself but also for determining predictions of the metric.

[0063] Figure 6An example of an Elman network as a recurrent neural network is shown. Neurons, as the transformation units of this network, are symbolically represented by circular units. An initial neuron 31 is shown on the left side of the figure, to which the operating data of the laundry machine 4 is supplied. Three hidden layer neurons 32, running in parallel with each other, are connected to the initial neuron 31, with feedback 30 established between the outputs and inputs of these hidden layer neurons 32. An output layer neuron 33 is used to output the metric and its prediction. The number of neurons 31, 32, and 33 shown does not represent a real Elman network.

[0064] exist Figure 6 In the Elman network shown, the value h of the hidden layer neuron 32 at time t is transformed by the nonlinear activation function act, and is determined by W. 1 / h Weighted input value X t Based on W, the hidden layer neurons 32 in the previous time interval h-1 are... (h-1) / h The weighted sum of the output values ​​is calculated. Next, the value W(h) of the hidden layer neuron 32 is summed with the weighted sum of the values ​​of the next layer represented by the output layer neuron 33. h / o Multiplying these by another activation function and transforming them, we obtain the network's result f(X). t This refers to the metric and its prediction. In other words, the difference between this and an artificial neural network without a recurrent structure is that the previous time interval h... t-1 The time feedback of hidden layer neurons 32 and its response to time segment h t The impact of the results. Figure 6 For clarity, only three hidden layer neurons 32 are shown. Figure 7 The predicted time trend is obtained through a trained Elman network consisting of 32 neurons in 20 hidden layers and one layer.

[0065] Figure 7 The graph illustrates different temporal trends of foam formation in the laundry detergent machine, with predicted metrics obtained in the virtual sensor 16 to determine foam using a trained Elman network, and confidence interval boundaries at 95% confidence levels for each time point. The horizontal axis of the graph shows the progress over 45 minutes, and the vertical axis shows the height of the foam in relative units: when the vertical axis is zero, there is no foam in the window of the door 27 (see [reference]). Figure 3(However, the window is not shown here). When the ordinate is 1, the foam reaches the upper edge of the window. Solid lines represent actual measurements, and dashed lines represent predicted values. Dotted lines mark confidence intervals. Here, the predictions shown in the curves are obtained as predictions for future time points 5 minutes later. The test values ​​shown in the curves are correspondingly shown with a 5-minute time delay to provide a comparison between prediction and reality. The confidence intervals vary depending on the length of the time interval for which the prediction is determined; therefore, the longer the time interval, the larger the confidence interval.

[0066] The trend in the left chart corresponds to a scenario where the laundry care process proceeds without foaming. This is also correctly reflected by the prediction with small fluctuations. The trend in the middle chart corresponds to a scenario where the laundry care process proceeds with moderate foaming. This scenario is also correctly predicted by the Elman network corresponding to the dashed line, which follows the solid line well with very small fluctuations. The trend in the right chart corresponds to a scenario where the laundry care process proceeds with a large amount of foaming. This scenario is also correctly predicted by the Elman network corresponding to the dashed line, which follows the solid line well with very small fluctuations.

[0067] Therefore, according to the present invention, the indirect measurement and prediction of the presence of foam 3 in the lye container 6 is performed by providing a virtual sensor 16, which determines a quantitative measure of the presence of foam 3 in the lye container 6 based on specific and measurable laundry care operation data. This measure may be the height of the foam 3 above the liquid 2 (particularly the laundry lye) that has briefly settled in the lye container 6 during an interruption. Thus, the correlation between the measure measured in model experiments and the simultaneously measured operation data is then implemented in the virtual sensor 16 in an algorithmic, tabular, or similar manner, so that the virtual sensor 16 can derive the desired measure from the corresponding operation data during actual laundry care and provide it for controlling the laundry care process.

[0068] The advantage of this invention is that the measurement makes it possible to draw quantitative conclusions about the presence of foam at virtually every point in time during the laundry care process. This quantitative conclusion is particularly permissible when controlling the laundry care machine 4, where adjustments can be made, in particular, using the measurement as an initial parameter, by different control parameters (such as the amount and temperature of the liquid to be achieved in the lye container and the metering of detergent in the lye container 6). The invention also allows for observation of the effects of measures influencing the foam 3, enabling the interruption, extension, repetition of a measure already taken, or its subsequent follow-up. It is also feasible to analyze the temporal trend of the measurement and thereby derive predictions of the appearance of foam 3 in subsequent stages of laundry care.

[0069] List of reference numerals

[0070] 1. Washing items

[0071] 2 Liquid

[0072] 3. Foam

[0073] 4. Laundry care machine

[0074] 5. Housing

[0075] 6. Alkali solution container

[0076] 7. Sumpf (bottom shell)

[0077] 8 rollers

[0078] 9. Carrying components

[0079] 10 Rotation axis

[0080] 11. Interior

[0081] 12 journals

[0082] 13 bearings

[0083] 14 motors

[0084] 15. Control device

[0085] 16 Virtual Sensors

[0086] 17. Pressure sensor in alkali solution container

[0087] 18 Power sensor on the motor

[0088] 19. Force or position sensor on bearing

[0089] 20. Chemical or temperature sensors in alkaline solution containers.

[0090] 21 Display devices

[0091] 22. Valve used to allow water to enter the alkali solution container.

[0092] 23 Rinse Chamber

[0093] 24 Metering device for measuring foam inhibitors

[0094] 25. Alkali pump for pumping out alkali solution

[0095] 26. Manschette (rings / loops)

[0096] 27 doors

[0097] 28 Drive control device

[0098] 29. Alkali pumps for circulating pumping

[0099] 30 Feedback Loop in Virtual Sensors

[0100] 31 Input layer neurons

[0101] 32 hidden layer neurons

[0102] 33 Output layer neurons

Claims

1. A method for determining foam (3) when laundry items (1) are treated with liquid (2) in a laundry care machine (4), The laundry care machine (4) has: Casing (5), and An alkaline solution container (6) arranged in the shell is used to receive the liquid (2). A drum (8) arranged in the alkali container (6) rotates about a rotation axis (10) and is driven by a motor (14), and the drum has an inner chamber (11) for receiving the washing items (1). in, The sensor continuously senses the operating data of the laundry care machine (4) and supplies it to the control device (15) connected to the sensor, so as to control the laundry care machine (4) according to the operating data. The control device (15) determines the foam (3) based on the operating data. Its features are, The control device (15) includes a virtual sensor (16), to which operating data is continuously supplied, and which continuously calculates a quantitative measurement of the foam (3) present in the alkali container (6) based on the supplied operating data. The measurement is the height or volume of the foam (3) in the roller (8). The virtual sensor (16) additionally determines the time-varying curve of the measurement, and the control device (15) uses the time-varying curve of the measurement to control the laundry care machine (4), wherein the operating data is selected from: the power consumption of the motor (14), the power consumption of the lye pump, the force consumption on the bearing of the drum (8), the rotation state of the drum (8), the liquid level of the liquid (2) in the lye container (6), the composition of the liquid (2), the temperature of the liquid (2), and the aforementioned time-varying curve. The virtual sensor (16) additionally determines a prediction of the time-varying curve of the metric, and the control device (15) uses the prediction to control the laundry care machine (4).

2. The method according to claim 1, in, The prediction is an estimate of a future metric. The estimated value corresponds to a measurement taken after a defined period of time, between 10 seconds and 10 minutes, following the prediction.

3. The method according to claim 1 or 2, in, The virtual sensor (16) includes a trained neural network that determines the metric.

4. The method according to claim 2, in, The estimated value corresponds to a measurement taken after a defined period of time, between 2 and 8 minutes, following the prediction.

5. The method according to claim 2, in, The estimated value corresponds to a measurement taken 5 minutes after the prediction.

6. The method according to claim 3, in, The trained neural network is a recurrent neural network.

7. The method according to any one of claims 1-2 and 4-6, in, The laundry detergent machine (4) has at least one actuator, which is operated by the control device (15) when controlling the laundry detergent machine (4), and the at least one actuator can reduce the foam (3) in the lye container, and The control device (15) performs at least one measure to reduce the foam (3) when using the at least one actuator, based on the metric.

8. The method according to claim 7, in, The control device (15) provides information about at least one defined measure for the user of the laundry care machine (4) to know.

9. The method according to claim 7, in, The treatment of the laundry items (1) is carried out in a programmed laundry care process, and The control device (15) determines the duration of the laundry care process based on measurements of the foam (3) and at least one measure taken to reduce the foam (3), and provides information about the duration to the user of the laundry care machine (4).

10. The method according to claim 8 or 9, in, The at least one measure is selected from: Add water to the liquid (2), Stabilize the temperature of the liquid (2), Reduce the rotational speed of the roller (8), and The liquid (2) is pumped out from the alkaline container (6), and Water is then poured into the alkali container (6).

11. A laundry care machine (4) for treating laundry items (1) by means of a liquid (2), comprising: Casing (5), and An alkaline solution container (6) arranged in the shell is used to receive the liquid (2). A drum (8) arranged in the alkali container (6) is rotatable about a rotation axis (10) and is driven by a motor (14). The drum has an inner chamber (11) for receiving the washing items (1). Sensors are used to continuously sense the operating data of the laundry care machine (4), and A control device (15) connected to the sensor is used to control the laundry care machine (4) based on the operating data. The control device (15) is configured to: determine the foam (3) present in the alkali container (6) based on the operating data. Its features are, The control device (15) includes a virtual sensor (16) to which operating data can be continuously supplied, and the virtual sensor is configured to: continuously determine a quantitative measure of foam (3) present in the lye container (6) based on the supplied operating data when processing the laundry item (1), the measure being the height or volume of the foam (3) in the drum (8), wherein the virtual sensor (16) is configured to: additionally determine a time-varying curve of the measure, and wherein the control device (15) is configured to: use the time-varying curve to control the laundry care machine (4), wherein the sensor includes at least one sensor selected from: a pressure sensor (17) in the lye container (6), a power sensor (18) on the motor (14), a power sensor on the lye pump, a force or position sensor (19) on the bearing of the drum (8), and a sensor (20) in the lye container (6) for determining the temperature or chemical composition of the liquid (2); and The virtual sensor (16) is configured to: additionally determine a prediction of the time-varying curve of the metric, and the control device (15) is configured to: use the prediction of the time-varying curve to control the laundry care machine (4).

12. The laundry care machine (4) according to claim 11. in, The prediction is an estimate of a future metric. The estimated value corresponds to a measurement taken after a defined period of time, between 10 seconds and 10 minutes, following the prediction.

13. The laundry care machine (4) according to claim 12. in, The estimated value corresponds to a measurement taken after a defined period of time, between 2 and 8 minutes, following the prediction.

14. The laundry care machine (4) according to claim 12. in, The estimated value corresponds to a measurement taken after a defined period of 5 minutes following the prediction.

15. The laundry care machine (4) according to any one of claims 11 to 14. in, The virtual sensor (16) includes a trained neural network that determines the metric.

16. The laundry care machine (4) according to claim 15. in, The trained neural network is a trained recurrent neural network.

17. The laundry care machine (4) according to claim 16. in, The trained neural network includes a trained Elman network.

18. The laundry care machine (4) according to any one of claims 11 to 14 and 16 to 17. The laundry detergent machine has at least one actuator, which can be operated by the control device (15) when controlling the laundry detergent machine (4), and the at least one actuator can reduce the foam (3) in the lye container (6), and in, The control device (15) is configured to: perform at least one measure to reduce the foam (3) when using the at least one actuator, based on the metric.

19. The laundry care machine (4) according to claim 18. The laundry care machine has a display device (21), and The control device (15) of the laundry care machine is configured to provide information about at least one determined measure by means of the display device (21) for the user of the laundry care machine (4) to know.

20. The laundry care machine (4) according to claim 19. The laundry care machine is configured to: process the laundry items (1) during a programmed laundry care process, and in, The control device (15) is configured to: determine the duration of the laundry care process based on a measurement of the foam (3) and based on at least one measure determined to reduce the foam (3), and to provide information about the duration by means of the display device (21) for the user to know.

21. The laundry care machine (4) according to claim 18. in, The at least one actuator is selected from: The motor (14). Valve (22) for allowing water to enter the alkaline solution container (6). An alkali pump for pumping the liquid (2) out of the alkali container (6), and Metering device for metering foam inhibitors into the alkali container (6).

22. The laundry care machine (4) according to any one of claims 11 to 14, 16 to 17 and 19 to 21. in, The virtual sensor (16) is generated from multiple model data sets through machine learning. Each model data group has a model value associated with the running data and a metric associated with that model value.

23. The laundry care machine (4) according to claim 22. in, The virtual sensor (16) is configured to determine the metric from the model data set by means of a regression approach.

Images