Dishwasher and control method thereof

Abstract

The invention belongs to the technical field of dish-washing machines, and provides a dish-washing machine and a control method thereof.The method comprises the steps that a vibration signal of a spraying arm of the dish-washing machine is obtained by responding to a washing program started by the dish-washing machine; processing the vibration signal to extract a time domain characteristic parameter and / or a frequency domain characteristic parameter in the vibration signal; the time domain characteristic parameters and / or the frequency domain characteristic parameters are input into a load recognition model obtained through pre-training, the loading capacity grade and the distribution uniformity grade, output by the load recognition model, of tableware loaded in the dish washing machine are obtained, and the load recognition model is obtained by training a sample data set through a machine learning algorithm. The sample data set comprises vibration characteristic samples under different loading capacities and vibration characteristic samples under different tableware distribution states, and each vibration characteristic sample is associated with a corresponding loading capacity grade label and a distribution uniformity grade label; according to the loading capacity grade and the distribution uniformity grade of the tableware, washing parameters of the dish washing machine are adjusted.

Description

Technical Field

[0001] This application belongs to the field of dishwasher technology, and more specifically, relates to a dishwasher and its control method. Background Technology

[0002] In related technologies, dishwashers generally rely on preset washing programs to operate. Washing parameters such as washing temperature and washing time are written into the preset washing program when the dishwasher leaves the factory, and the washing parameters cannot be dynamically adjusted according to actual usage.

[0003] However, when the dishwasher is lightly loaded or empty, the high-energy-consuming preset washing parameters will cause unnecessary waste of water and electricity; while when the dishwasher is heavily loaded, the fixed cleaning force is difficult to penetrate stacked dishes, which can easily leave stains and result in unstable cleaning effect. Summary of the Invention

[0004] The purpose of this application is to provide a dishwasher and its control method, which aims to solve the technical problem that traditional dishwashers rely on preset washing programs to determine washing parameters inaccurately, resulting in resource waste and unstable cleaning effect.

[0005] To achieve the above objectives, according to the first aspect of this application, a method for controlling a dishwasher is provided, the method comprising: In response to the dishwasher starting the washing program, the vibration signal of the dishwasher's spray arm is obtained; The vibration signal is processed to extract time-domain and / or frequency-domain feature parameters. The time-domain feature parameters and / or frequency-domain feature parameters are input into the pre-trained load recognition model to obtain the load level and distribution uniformity level of the tableware loaded in the dishwasher. The load recognition model is trained on the sample dataset by a machine learning algorithm. The sample dataset contains vibration feature samples under different load levels and vibration feature samples under different tableware distribution states. Each vibration feature sample is associated with a corresponding load level label and distribution uniformity level label. Adjust the dishwasher's washing parameters according to the load level and distribution uniformity level of the tableware.

[0006] The beneficial effects of this application embodiment compared with the prior art are as follows: This method abandons pressure sensors or optical sensors that are susceptible to interference and have high costs, and instead acquires the vibration signal of the spray arm. By utilizing the strong correlation between the load change of the spray arm and the vibration signal, hardware costs are reduced and anti-interference ability is improved. Secondly, by extracting the time-domain feature parameters and frequency-domain feature parameters in the vibration signal, the vibration differences of the dishwasher under different loading and distribution states are determined. Then, by using a load recognition model trained with multiple samples, the accurate judgment of the loading level and distribution uniformity level is achieved, avoiding the blindness of fixed washing parameters. Finally, based on the loading level and distribution uniformity level of the tableware, the washing parameters of the dishwasher are dynamically adjusted to avoid resource waste and ensure stable cleaning effect.

[0007] According to a second aspect of this application, a dishwasher is provided, including a spray arm base, a vibration sensor mounted on the spray arm base, a main controller, and an actuator assembly; The main controller, connected to the vibration sensor, is configured to control the vibration sensor to collect vibration signals from the dishwasher's spray arms in response to the start of the dishwasher's washing program. The main controller, connected to the execution components, is also configured to process the vibration signal, extract time-domain feature parameters and / or frequency-domain feature parameters; input the time-domain feature parameters and / or frequency-domain feature parameters into a pre-trained load recognition model to obtain the load level and distribution uniformity level of the dishes loaded in the dishwasher output by the load recognition model; and control the execution components to perform corresponding operations based on the load level and distribution uniformity level to adjust the washing parameters of the dishwasher. The load identification model is trained on a sample dataset using a machine learning algorithm. The sample dataset contains vibration feature samples under different loads and vibration feature samples under different tableware distribution states. Each vibration feature sample is associated with a corresponding load level label and distribution uniformity level label.

[0008] According to a third aspect of this application, an electronic device is provided, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the electronic device performs the method as described in any one of the above.

[0009] According to a fourth aspect of this application, a computer-readable storage medium is provided that stores a computer program, which, when executed by a processor, implements the method as described in any one of the above.

[0010] According to a fifth aspect of this application, a computer program product is provided that, when run on an electronic device, causes the electronic device to perform the method described in any one of the first aspects above.

[0011] It is understandable that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a schematic diagram of the structure of a dishwasher provided in an embodiment of this application; Figure 2 This is a flowchart illustrating a dishwasher control method provided in an embodiment of this application; Figure 3 This is a schematic flowchart of an optional dishwasher control method provided in an embodiment of this application; Figure 4 This is a schematic flowchart of an optional dishwasher control method provided in an embodiment of this application; Figure 5 This is a schematic flowchart of an optional dishwasher control method provided in an embodiment of this application; Figure 6 This is a schematic diagram of the structure of a dishwasher control device provided in an embodiment of this application; Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0014] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0015] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0016] It should also be understood that, in the description of this application, unless otherwise stated, the " / " used in the specification and appended claims indicates that the related objects are in an "or" relationship. For example, A / B can mean A or B. The "and / or" in this application is merely a description of the relationship between the related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. Furthermore, in the description of this application, unless otherwise stated, "multiple" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0017] Furthermore, to facilitate a clear description of the technical solutions in the embodiments of this application, the terms "first" and "second" are used in the embodiments of this application to distinguish identical or similar items with essentially the same function and effect. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, but are only used for distinguishing descriptions, and the terms "first" and "second" do not necessarily imply that they are different, nor should they be construed as indicating or implying relative importance.

[0018] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0019] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0020] In related technologies, some solutions attempt to detect the load in the dishwasher using pressure sensors or optical sensors. However, pressure sensors are easily affected by fluctuations in the flow of washing water, resulting in low detection accuracy. Optical sensors are easily obscured by oil stains on the surface of tableware, leading to poor reliability. Furthermore, both types of sensors require additional hardware costs, which is not conducive to large-scale mass production and promotion.

[0021] Therefore, there is an urgent need for a low-cost, interference-resistant technology that can accurately identify the load status of tableware and dynamically adapt washing parameters to solve the dual problems of resource waste and poor cleaning effect of traditional dishwashers.

[0022] This application provides an example of a dishwasher; please refer to [link / reference]. Figure 1 As shown, Figure 1 The diagram shows a structural schematic of a dishwasher provided in this application. It is provided as an example and not as a limitation. The control method of the dishwasher provided in this application can be applied to or operated in a dishwasher. The dishwasher includes a spray arm base 101, a vibration sensor 102 mounted on the spray arm base 101, a main controller 103, and an execution component 104.

[0023] The main controller 103, connected to the vibration sensor 102, is configured to control the vibration sensor to collect vibration signals from the dishwasher's spray arms in response to the start of the dishwasher's washing program.

[0024] The main controller 103, connected to the execution component 104, is also configured to process the vibration signal, extract time-domain feature parameters and / or frequency-domain feature parameters; input the time-domain feature parameters and / or frequency-domain feature parameters into a pre-trained load recognition model to obtain the load level and distribution uniformity level of the tableware loaded in the dishwasher output by the load recognition model; and control the execution component to perform corresponding operations based on the load level and distribution uniformity level to adjust the washing parameters of the dishwasher.

[0025] The load identification model is trained on a sample dataset using a machine learning algorithm. The sample dataset contains vibration feature samples under different loads and vibration feature samples under different tableware distribution states. Each vibration feature sample is associated with a corresponding load level label and distribution uniformity level label.

[0026] Indicatively, the spray arm base is the mounting base for the spray arm of the dishwasher. It has mounting holes inside to accommodate the rotating shaft of the spray arm, and a sensor mounting slot is reserved on the side wall near the rotating shaft. The vibration sensor is fixedly installed in the sensor mounting slot.

[0027] As an illustration, a piezoelectric accelerometer can be selected as the vibration sensor. The sensing end of the sensor is closely attached to the side wall of the spray arm base, which can directly collect the vibration signal transmitted through the spray arm base when the spray arm rotates. The sampling frequency can be preset (e.g., 1kHz) to capture subtle changes in the vibration signal under different loads.

[0028] Indicatively, the main controller can be a microprocessor that establishes communication connections with vibration sensors and actuators via a CAN bus, possessing functions such as signal reception, data processing, model calculation, and command transmission. The actuators controlled by the main controller can include a washing pump, a heater, a washing liquid distributor, and a drive motor for the spray arm. The washing pump is used to regulate the washing water pressure and volume, the heater is used to control the washing water temperature, the washing liquid distributor is used to quantitatively release the washing liquid, and the drive motor for the spray arm is used to control the speed and rotation mode of the spray arm. Each actuator can respond to the pulse signal from the main controller to perform corresponding actions.

[0029] In some embodiments, when the user connects to the main controller 103 via the dishwasher's control panel 105, which can be understood as a display control board or display controller, it has functions such as signal reception, data processing, model calculation, and command transmission. Besides including a basic control chip or processing chip for processing user touch or button operation data, it also includes... Figure 1 As shown, the peripherals of the control panel 105 include a buzzer, physical buttons, and a display panel for receiving user touch operations. For example, it can start washing programs such as daily wash and intensive wash in response to the user's press of the start button on the control panel 105 or the start button on the display panel. The user can also start the washing program through a mobile terminal associated with the dishwasher (such as a smartphone, PAD, or smart control robot). Afterwards, the control panel 105 sends the start command or start signal of the dishwasher to the main controller 103 to trigger the main controller 103 to control the drive motor of the spray arm to drive the spray arm at a preset standard speed (such as 1500 r / min, which can be understood as the reference speed for stable rotation of the spray arm and can be finely adjusted through the main controller parameter configuration interface) for a set duration (such as 30s) to eliminate the mechanical inertial vibration interference of the spray arm at the initial stage of start-up and ensure that the vibration signal collected later can accurately reflect the tableware loading status.

[0030] After the spray arm has been running for a set time, the main controller sends a data acquisition command to the vibration sensor. The vibration sensor then begins to acquire the vibration signal transmitted to the base of the spray arm as it rotates. The acquisition time is set to 10 seconds. During the acquisition process, the vibration sensor converts the analog vibration signal into an analog electrical signal and transmits it to the main controller in real time via the CAN bus.

[0031] In some embodiments, after receiving the analog electrical signal, the main controller can first preprocess the analog electrical signal. For example, it can reduce the noise of the analog electrical signal by using a built-in digital filtering algorithm (using a 50Hz notch filter to remove mains interference and a 10-500Hz bandpass filter to retain the effective vibration frequency band), and then convert the noise-reduced analog signal into a digital signal through an analog-to-digital conversion module. Subsequently, the main controller extracts time-domain feature parameters and / or frequency-domain feature parameters from the digital signal. The time-domain feature parameters include the peak amplitude (unit: g, where g is the gravitational acceleration) and root mean square amplitude of the vibration signal. The frequency-domain feature parameters are calculated by fast Fourier transform and specifically include the main vibration frequency (unit: Hz) and the proportion of spectral energy in the first preset frequency band (10-50Hz, low frequency band) and the second preset frequency band (100-300Hz, high frequency band).

[0032] In some embodiments, the main controller inputs the extracted time-domain feature parameters and / or frequency-domain feature parameters into the pre-trained load recognition model to obtain the load recognition model outputting the level of tableware loading and the level of distribution uniformity of the tableware in the dishwasher. In this embodiment, the load recognition model is trained and generated using a random forest machine learning algorithm. The training process can be as follows: When constructing the sample dataset, five loading scenarios corresponding to different loading levels are set in the dishwasher: empty (approximately 0% capacity), 1 / 4 load (approximately 25% capacity, such as a total tableware weight of 1-2 kg), 1 / 2 load (approximately 50% capacity, such as a total tableware weight of 2-3 kg), 3 / 4 load (over 70% capacity, such as a total tableware weight of 3-4 kg), and full load (over 100% capacity, such as a total tableware weight of 4-5 kg). Under each loading scenario, four distribution state scenarios are set: uniform distribution (tableware is evenly placed in each area of ​​the rack), local stacking (tableware is concentrated in the center of the rack), edge offset (tableware is biased towards the edge of the rack), and mixed disorder (tableware is placed irregularly).

[0033] For each loading scenario, the spray arm is controlled to run at a standard speed and vibration signals are collected. For example, if 50 sets of valid vibration signals are collected as samples for each scenario, a total of 5×4×50=1000 sets of samples are collected. After extracting the above-mentioned time-domain feature parameters and / or frequency-domain feature parameters for each set of samples, the above-mentioned time-domain feature parameters and / or frequency-domain feature parameters are respectively labeled with the corresponding loading level labels (e.g., L0 for empty load, L1 for 1 / 4 load, L2 for 1 / 2 load, L3 for 3 / 4 load, and L4 for full load) and distribution uniformity level labels (e.g., U0 for uniform distribution, U1 for local stacking, U2 for edge offset, and U3 for mixed and disordered distribution). The labeled sample dataset is divided into a training set and a test set in a 7:3 ratio. The model parameters are iteratively optimized through the training set, and the model accuracy is verified through the test set. The final load recognition model has an accuracy of no less than 92% for the loading level and no less than 90% for the distribution uniformity level.

[0034] After the load recognition model outputs the load level and distribution uniformity level of the dishes inside the dishwasher, the main controller generates corresponding parameter adjustment instructions based on these levels and sends them to the execution components. For example, if the load level is L4 (full load), the main controller increases the washing pump speed from the base speed of 1800 r / min to 2200 r / min to increase the washing water pressure, raises the washing water temperature from the base temperature of 50℃ to 60℃ to enhance cleaning power, and extends the washing time from the base time of 30 minutes to 40 minutes. Conversely, if the load level is L0 (empty), the washing pump speed is reduced to 1200 r / min and the water temperature is reduced to 40℃. Washing time is shortened to 20 minutes to reduce resource consumption. For example, if the distribution uniformity level is U3 (mixed and messy), the main controller controls the detergent dispenser to add 5ml of detergent to the basic amount (15ml), and at the same time controls the spray arm drive motor to increase the rotation speed frequency of the spray arm from 1 time / minute (i.e., the speed changes once every 60 seconds) to 2 times / minute. By rotating at different speeds, the water flow can be more evenly distributed to cover the messy dishes, avoiding cleaning dead corners. If the distribution uniformity level is lower than the preset threshold (e.g., U2 and below), the main controller can also synchronously control the dishwasher's display panel to display a tableware placement guide diagram, prompting the user to adjust the placement of the tableware to further ensure the cleaning effect.

[0035] During the washing of dishes using the adjusted washing parameters, the main controller can also trigger secondary signal acquisition and recognition at half the total washing time (e.g., at the 20th minute when the total washing time is 40 minutes): it controls the vibration sensor to collect the vibration signal of the spray arm again, repeats the above feature extraction and model recognition steps, and if the loading level or distribution uniformity level of the secondary recognition deviates from the initial recognition result by more than 10% (e.g., the initial recognition is L3 and the secondary recognition is L2), the main controller will fine-tune the remaining washing parameters, with the fine-tuning range not exceeding 20% ​​of the initial adjustment parameters (e.g., the washing pump speed is fine-tuned from 2200 r / min to 2000 r / min), to ensure that the washing parameters during the washing process always match the actual tableware loading status, taking into account both cleaning effect and energy saving.

[0036] In one possible implementation, the washing parameters include at least one of the following: washing pump speed, washing temperature, washing liquid volume, washing duration, and washing water volume; the actuating components include at least one of the following: spray arm drive, washing liquid distributor, washing pump, and heater.

[0037] In some embodiments, the main controller controls the execution components to perform corresponding operations based on the load level and distribution uniformity level, in order to adjust the washing parameters of the dishwasher, specifically configured as follows: If the uniformity level is lower than the preset uniformity threshold, it is determined that the dishwasher has uneven distribution of tableware.

[0038] If the load capacity level exceeds the preset load capacity threshold, it is determined that the dishwasher is overloaded.

[0039] If the uniformity level is lower than the preset uniformity threshold and the load level exceeds the preset load threshold, it is determined that the dishwasher is overloaded and the tableware is unevenly distributed.

[0040] If the dishwasher is overloaded, at least one of the detergent dispenser, the washing pump, and the heater will perform corresponding operations based on the load level to adjust at least one of the washing pump speed, washing temperature, washing time, and washing water volume.

[0041] If the dishwasher has uneven distribution of dishes, at least one of the detergent dispenser and spray arm drive components will perform corresponding operations based on the distribution uniformity level to adjust the detergent usage and the washing pump speed.

[0042] In some embodiments, the loading capacity level and distribution uniformity level can be clearly defined in advance. For example, but not limited to, the loading capacity level is divided into 5 levels based on the total weight of the tableware: L0 (empty, 0kg), L1 (1 / 4 load, 1-2kg), L2 (1 / 2 load, 2-3kg), L3 (3 / 4 load, 3-4kg), and L4 (full load, 4-5kg), with higher level values ​​representing a larger loading capacity. As another example, but not limited to, the distribution uniformity level is divided into 4 levels based on the regularity of the tableware arrangement: U0 (uniform distribution, tableware is not stacked and is evenly distributed in all areas of the basket), U1 (slightly uneven, with a small amount of stacking in some areas), U2 (moderately uneven, stacking on one side or edge offset), and U3 (severely uneven, messy stacking in multiple areas), with higher level values ​​representing a more uneven distribution.

[0043] In some embodiments, basic values ​​for each washing parameter can be preset, such as a basic washing pump speed of 1800 r / min, a basic washing temperature of 50°C, a basic washing time of 30 min, a basic washing water volume of 8 L, a basic washing liquid volume of 15 ml, and a basic speed change frequency of the spray arm of 1 time / minute (i.e., the speed is switched once every 60 seconds). Each washing parameter is adjusted according to a preset gradient as the washing level changes.

[0044] In some embodiments, after the main controller identifies the loading level and distribution uniformity level, it controls the washing pump, heater, timing module, and water volume control unit (integrated into the washing pump) to perform adjustments according to the following rules: the higher the loading level, the higher the washing pump speed, the higher the washing temperature, the longer the washing time, and the more washing water, and thus the higher the frequency of the spray arm rotation affected by the washing pump speed; the lower the distribution uniformity level, the more washing liquid is used and the higher the washing pump speed, and thus the higher the frequency of the spray arm rotation affected by the washing pump speed.

[0045] For example, when the load level is identified as L0 (no load), in order to avoid waste of resources, the main controller controls the washing pump speed to be reduced to 1200 r / min (33% lower than the base value), the washing temperature to be reduced to 40℃ (20% lower), the washing time to be shortened to 20 min (33% shorter), and the washing water volume to be reduced to 5L (37.5% lower). At this time, only the minimum parameters required for basic cleaning are maintained, which greatly reduces water and electricity consumption.

[0046] For example, when the main controller identifies the load level as L1 (1 / 4 load), it adjusts the washing parameters appropriately: the washing pump speed is increased to 1500 r / min (reduced by 17%), the washing temperature is maintained at 45℃ (reduced by 10%), the washing time is shortened to 25 min (reduced by 17%), and the washing water volume is reduced to 6.5L (reduced by 18.75%), balancing cleaning needs with energy-saving goals.

[0047] For example, when the main controller identifies the load level as L2 (1 / 2 load), it uses basic washing parameters, namely, washing pump speed of 1800 r / min, washing temperature of 50℃, washing time of 30 min, and washing water volume of 8L, which is suitable for daily cleaning needs with medium load.

[0048] For example, when the main controller identifies the load level as L3 (3 / 4 load), it appropriately increases the washing parameters: the washing pump speed is increased to 2000 r / min (an increase of 11%), the washing temperature is increased to 55℃ (an increase of 10%), the washing time is extended to 35 min (an extension of 17%), and the washing water volume is increased to 9L (an increase of 12.5%). By increasing the water pressure and temperature, the cleaning ability is enhanced, which is adapted to the cleaning needs under higher load levels.

[0049] For example, when the main controller identifies the load level as L4 (full load), it maximizes the washing parameters: the washing pump speed is increased to 2200r / min (22% increase), the washing temperature is increased to 60℃ (20% increase), the washing time is extended to 40min (33% extension), and the washing water volume is increased to 10L (25% increase). At this time, the high speed ensures the water flow penetration, the high temperature accelerates the dissolution of stains, and the long time ensures thorough cleaning, solving the problem of insufficient cleaning in heavy load scenarios.

[0050] In one embodiment, for example, when the main controller identifies the distribution uniformity level as U0 (uniform distribution), it uses basic parameters, namely, 15ml of washing liquid and 1 spray arm speed change frequency per minute. At this time, the water flow can evenly cover all the tableware without the need to add additional washing liquid or adjust the speed mode.

[0051] For example, when the main controller identifies the distribution uniformity level as U1 (slightly uneven): slightly adjust the washing parameters, increase the amount of washing liquid to 18ml (an increase of 20%), and increase the spray arm speed frequency to 1.2 times / minute (shortening the switching interval to 50s). By slightly increasing the detergent concentration and slightly changing the speed, the cleaning effect of local stacked areas can be improved.

[0052] For example, when the main controller identifies the distribution uniformity level as U2 (moderate unevenness): moderately adjust the washing parameters, increase the amount of washing liquid to 21ml (40% increase), and increase the spray arm speed frequency to 1.5 times / minute (shorten the switching interval to 40s). The higher speed frequency allows for more flexible water flow direction, covering areas of tableware that are offset at the edge or stacked on one side, and additional detergent compensates for local cleaning power.

[0053] For example, when the main controller identifies the distribution uniformity level as U3 (severely uneven): the washing parameters are significantly adjusted, the amount of detergent is increased to 24ml (an increase of 60%), the spray arm speed frequency is increased to 2 times / minute (shortening the switching interval to 30s), the high-frequency speed change makes the water flow form a dynamic impact, effectively penetrating the gaps between messy stacked tableware, and the high concentration of detergent ensures that stubborn stains are dissolved, solving the problem of cleaning blind spots caused by uneven distribution.

[0054] In this embodiment, the main controller can pre-store the correspondence table between the above-mentioned "level-parameter". After identifying a specific level, it directly calls the parameter value in the table and sends pulse commands to the corresponding execution components via the CAN bus. For example, when identified as L4+U3, it synchronously sends a "2200r / min" command to the washing pump, a "60℃" command to the heater, a "40min" command to the timing module, a "24ml" command to the detergent dispenser, and a "2 times / minute" command to the spray arm drive. Each execution component responds and executes within 100ms after receiving the command, and the main controller collects the feedback signals of each execution component in real time (such as washing pump speed feedback and detergent dosage feedback). If the parameter deviation is detected to exceed ±5%, the adjustment command is resent to ensure the accuracy and stability of the washing parameter adjustment.

[0055] Through the above-mentioned graded adjustment logic, the washing parameters can be dynamically adapted under different load conditions, which not only avoids excessive energy consumption under light load, but also solves the problem of insufficient cleaning under heavy load and uneven distribution, significantly improving the intelligence and practicality of the dishwasher.

[0056] This application provides an example of a dishwasher control method; please refer to... Figure 2 As shown, Figure 2 A schematic flowchart of a dishwasher control method provided in this application is shown. This is an example and not a limitation; the method can be applied to or operated in a dishwasher. The method includes: S201, in response to the dishwasher starting the washing program, obtains the vibration signal of the dishwasher's spray arm.

[0057] S202, Process the vibration signal to extract time-domain and / or frequency-domain characteristic parameters from the vibration signal.

[0058] S203, input the time-domain feature parameters and / or frequency-domain feature parameters into the pre-trained load recognition model to obtain the load level and distribution uniformity level of the tableware loaded in the dishwasher output by the load recognition model.

[0059] S204, adjust the dishwasher's washing parameters according to the load level and distribution uniformity level of the tableware.

[0060] The load identification model is trained on a sample dataset using a machine learning algorithm. The sample dataset contains vibration feature samples under different loads and vibration feature samples under different tableware distribution states. Each vibration feature sample is associated with a corresponding load level label and distribution uniformity level label.

[0061] In some embodiments, when the user starts a washing program (such as normal wash, heavy-duty wash, etc.) through the dishwasher's control panel or associated mobile terminal, the dishwasher's main controller triggers the drive motor of the spray arm to drive the spray arm at a preset standard speed (such as 1500 r / min, which can be understood as the reference speed for stable rotation of the spray arm and can be finely adjusted through the main controller parameter configuration interface) for a set duration (30s) to eliminate mechanical inertial vibration interference in the initial stage of spray arm startup (0s to ts of starting the washing program, where t needs to be set according to different washing programs such as heavy-duty wash and normal wash), ensuring that the vibration signals collected subsequently can accurately reflect the loading status of the tableware in the dishwasher.

[0062] After the spray arm has been running for a set time, the dishwasher's main controller sends a data acquisition command to the vibration sensor installed on the spray arm base. For example, a piezoelectric accelerometer can be used as the vibration sensor. The sensing end of the sensor is tightly bonded to the metal sidewall of the spray arm base using high-temperature resistant silicone (suitable for the dishwasher's operating temperature of 50-70℃). The sampling frequency is set to 1kHz (covering the effective vibration frequency band of 10-500Hz generated by the rotation of the spray arm), and the acquisition time is set to 10s. During the acquisition process, the sensor converts the mechanical vibration transmitted to the spray arm base when the spray arm rotates into an analog electrical signal, which is then transmitted in real time to the signal input terminal of the main controller via a shielded wire, avoiding electromagnetic interference from the motor and heater inside the dishwasher during signal transmission.

[0063] In some embodiments, after receiving the analog electrical signal, the main controller of the dishwasher first performs noise reduction processing on the analog electrical signal through the built-in digital filtering algorithm of the main controller. For example, a 50Hz notch filter is used to remove mains frequency interference (the dishwasher is powered by 220V / 50Hz, which is prone to generating 50Hz electromagnetic noise). Then, a 10-500Hz bandpass filter is used to retain the effective vibration frequency band related to the spray arm load, filtering high-frequency noise (such as sensor circuit noise) and low-frequency interference (such as cabinet vibration). Subsequently, the analog-to-digital conversion module (12-bit sampling accuracy) of the main controller converts the noise-reduced analog electrical signal into a digital signal.

[0064] In some embodiments, the dishwasher's main controller extracts time-domain and / or frequency-domain feature parameters from the digital signal. Furthermore, the main controller inputs the extracted time-domain feature parameters (peak amplitude, root mean square amplitude) and / or frequency-domain feature parameters (dominant vibration frequency, energy ratio of high and low frequency bands) into a pre-trained load recognition model.

[0065] The time-domain characteristic parameters include the peak amplitude of the vibration signal (unit: g, where g is the acceleration due to gravity) and the root mean square amplitude (obtained by taking the square root of the square mean of all sampling points of the digital signal, reflecting the average energy of the vibration signal). The frequency-domain characteristic parameters are obtained by converting the time-domain digital signal into a frequency-domain signal through Fast Fourier Transform (FFT) to extract the dominant vibration frequency (unit: Hz, i.e., the frequency point with the highest energy in the frequency-domain signal) and the proportion of spectral energy.

[0066] In addition, the frequency domain characteristic parameters can be divided into a first preset frequency band (10-50Hz, low frequency band) and a second preset frequency band (100-300Hz, high frequency band), and the total spectral energy in the two frequency bands can be calculated separately. Then, the sum of the spectral energy in the entire effective frequency band (10-500Hz) can be divided to obtain the energy percentage of the corresponding frequency band (unit: %).

[0067] In some embodiments, the load identification model is trained using a random forest machine learning algorithm, and its training process and model characteristics are as follows: First, set up 5 loading scenarios in the dishwasher: L0 (empty, tableware weight 0kg), L1 (1 / 4 load, tableware weight 1-2kg, 8-12 standard plates), L2 (1 / 2 load, tableware weight 2-3kg, 13-20 standard plates + 3-5 soup bowls), L3 (3 / 4 load, tableware weight 3-4kg, 21-28 standard plates + 6-8 soup bowls + 10-15 pairs of chopsticks), and L4 (full load, tableware weight 4-5kg, 29-35 standard plates + 9-12 soup bowls + 16-20 pairs of chopsticks + 5-8 spoons).

[0068] Then, four distribution states were set for each type of loading scenario: U0 (uniform distribution, tableware is evenly distributed in each area of ​​the basket rack according to the rule of "plates are placed upright and soup bowls are placed in a scattered manner"), U1 (slightly uneven, tableware is stacked in 1 / 4 area, stacking height does not exceed 2 layers), U2 (moderately uneven, tableware is stacked in 1 / 3 area, stacking height is 2-3 layers or tableware is biased towards one side of the basket rack), and U3 (severely uneven, tableware is randomly stacked in 1 / 2 area, stacking height is more than 3 layers or tableware is concentrated at the edge of the basket rack).

[0069] For each "load-distribution status" combination scenario, the spray arm was controlled to run at 1500 r / min for 30 seconds, and then a vibration signal was collected for 10 seconds. 50 sets of valid signals were collected for each scenario (samples with abnormal signal amplitude were removed), for a total of 5×4×50=1000 sets of samples. After extracting the above-mentioned time-domain feature parameters and / or frequency-domain feature parameters from each set of samples, the time-domain feature parameters and / or frequency-domain feature parameters were labeled with the corresponding load level labels (L0-L4) and distribution uniformity level labels (U0-U3), forming a sample dataset.

[0070] The sample dataset was divided into a training set (700 groups) and a test set (300 groups) in a 7:3 ratio. The random forest algorithm (with 100 decision trees and a maximum depth of 10) was used to iteratively optimize the model parameters on the training set. The model accuracy was verified on the test set. The final load recognition model achieved an accuracy of 93% in recognizing the load level and 91% in recognizing the distribution uniformity level, which meets the actual usage requirements. The model parameters (such as decision tree node thresholds and feature weights) after training were stored in the Flash memory of the main controller for easy real-time retrieval.

[0071] After receiving the feature vector, the load identification model outputs the corresponding load level (e.g., L3) and distribution uniformity level (e.g., U2) through its built-in classification logic, and feeds the level results back to the main controller. The main controller pre-stores a "level-parameter" correspondence table. Based on the load level and distribution uniformity level output by the load identification model, it retrieves the corresponding washing parameter adjustment instructions from the table to control the dishwasher's execution components (wash pump, heater, detergent dispenser, spray arm drive) to perform operations. The washing parameters include wash pump speed, washing temperature, detergent usage, washing time, and water volume.

[0072] By following the above steps, the washing process of the dishwasher can be dynamically adapted, which avoids the waste of resources and unstable cleaning effect of the preset washing programs of traditional dishwashers, and improves the intelligence and practicality of washing.

[0073] In one possible implementation, please refer to Figure 3 As shown, Figure 3 A schematic flowchart of an optional dishwasher control method provided in this application is shown, which, in response to the dishwasher starting a washing program, acquires a vibration signal from the dishwasher's spray arm, including: S301, in response to the dishwasher starting the washing program, controls the spray arm to run at a preset speed for a set duration.

[0074] S302, after the spray arm runs at a preset speed for a set time, controls the vibration sensor installed on the spray arm base of the dishwasher to collect the vibration signal of the spray arm.

[0075] S303, acquire the vibration signal of the spray arm collected by the vibration sensor.

[0076] In some embodiments, when a user starts any washing program (such as daily wash or quick wash) through the dishwasher's control panel (such as a touch screen or physical button) or associated mobile terminal (such as a mobile APP), the dishwasher's main controller immediately responds to the start command and triggers the control spray arm to run at a preset speed for a set duration to eliminate mechanical inertial vibration interference in the initial stage of spray arm start-up (0s to ts of starting the washing program, where t needs to be set according to different washing programs such as heavy wash or normal wash), and avoids unstable vibration causing subsequent signal acquisition distortion.

[0077] For example, the main controller communicates via the CAN bus with the spray arm drive unit (which uses a 42-stepper motor with a rated torque of 0.5N). The main controller sends a speed control command (m, adapted to the rotational load requirements of the spray arm) to control the spray arm to operate at a preset speed (e.g., 1500 r / min). This preset speed is a benchmark speed that ensures the spray arm maintains stable rotation and conforms to actual washing conditions under different tableware loads. Simultaneously, the main controller uses a built-in timing module to keep track until the set duration (e.g., 30 seconds). Experimental data shows that the spray arm requires a 15-25 second transition period from startup to stable rotation, during which the amplitude fluctuation range can reach ±0.2g; after 30 seconds of operation, the amplitude fluctuation can be controlled within ±0.05g, and the vibration state is completely stable, accurately reflecting the load impact of tableware loading on the spray arm.

[0078] As an illustration, a piezoelectric accelerometer can be selected as the vibration sensor; the spray arm base is made of 304 stainless steel (which has good vibration conductivity), and a sensor mounting groove is reserved on the side wall near the rotation axis of the spray arm. The sensor is tightly attached and fixed in the mounting groove with high temperature resistant silicone, ensuring that the vibration energy generated by the rotation of the spray arm is completely transmitted to the sensor sensing end through the base, without significant vibration attenuation.

[0079] In some embodiments, after the spray arm has been running stably at a preset speed for 30 seconds, the main controller sends a data acquisition trigger signal to the vibration sensor installed on the spray arm base via the I / O port. The vibration sensor then begins to acquire the vibration signal of the spray arm. The acquisition time is set to 10 seconds, which can acquire 10,000 effective sampling points (1kHz×10s), covering 2-3 complete rotation cycles of the spray arm (1500r / min corresponds to one rotation every 0.04s), thus fully capturing the vibration patterns under different loading conditions. If the acquisition time is too short (e.g., less than 5 seconds), the vibration features may not be fully extracted due to insufficient sample size. If the acquisition time is too long, it will extend the total washing program time, affecting the user experience. During the acquisition process, the vibration sensor converts the mechanical vibration into a 0-5V analog electrical signal, which is transmitted in real time to the analog signal input terminal of the main controller through a twisted-pair shielded wire (the shield is grounded to effectively isolate electromagnetic interference generated by the motor and heater inside the dishwasher), avoiding amplitude attenuation during signal transmission.

[0080] Afterwards, the analog signal input terminal of the main controller receives the analog electrical signal transmitted by the vibration sensor. First, it uses the built-in signal conditioning circuit (a non-inverting amplifier circuit composed of operational amplifiers with a magnification factor of 10) to amplify the weak vibration electrical signal to the standard processing range of 0-5V. Then, the main controller starts the 12-bit analog-to-digital converter module (conversion rate of 1MHz) to convert the amplified analog electrical signal into a digital signal of 0-4095 and stores it in real time in the main controller's RAM buffer (the buffer capacity is 128KB, which can meet the storage requirements of 10s of data acquisition).

[0081] In one possible implementation, the washing parameters include at least one of the following: washing pump speed, washing temperature, washing liquid volume, washing duration, and washing water volume.

[0082] Adjust the dishwasher's washing parameters according to the load capacity and distribution uniformity levels, including: If the uniformity level is lower than the preset uniformity threshold, it is determined that the dishwasher has uneven distribution of tableware.

[0083] If the load capacity level exceeds the preset load capacity threshold, it is determined that the dishwasher is overloaded.

[0084] If the uniformity level is lower than the preset uniformity threshold and the load level exceeds the preset load threshold, it is determined that the dishwasher is overloaded and the tableware is unevenly distributed.

[0085] If the dishwasher is overloaded, adjust at least one of the following based on the load level: washing pump speed, washing temperature, washing time, and washing water volume. If the dishwasher has uneven distribution of dishes, adjust the amount of detergent and the speed of the washing pump based on the distribution uniformity level.

[0086] As an illustration, the main controller pre-stores two types of judgment thresholds as the basis for identifying dishwasher overload and uneven distribution of tableware. The preset uniformity threshold and preset loading amount threshold can be set based on a large amount of experimental data in the early stage to adapt to the daily tableware loading scenarios of families.

[0087] The preset loading threshold is set to loading level L4, which corresponds to a fully loaded state. For example, this could be a total tableware weight of 4-5kg, including 29-35 standard plates, 9-12 soup bowls, 16-20 pairs of chopsticks, and 5-8 spoons. The reason for setting the preset loading threshold is that when the loading exceeds L4, the dishwasher basket's load-bearing capacity approaches its limit, and the fixed washing parameters in related technologies cannot penetrate the stacked tableware, failing to meet cleaning needs. Therefore, it is necessary to increase the parameters to enhance the cleaning power.

[0088] The preset uniformity threshold is set to distribution uniformity level U1, which corresponds to a slightly uneven state, specifically when 1 / 4 of the area has stacked tableware and the stacking height does not exceed 2 layers. The preset uniformity threshold is set because experiments have shown that when the distribution uniformity level drops to U1 or below, local areas will form cleaning blind spots due to the stacking of tableware, resulting in insufficient water flow coverage. It is necessary to adjust the parameters to compensate for the cleaning power.

[0089] In addition, both the preset uniformity threshold and the preset load capacity threshold can be fine-tuned through the parameter setting interface displayed on the dishwasher's control panel. For example, for special tableware such as large soup bowls and woks, users can lower the load capacity threshold to L3 to adapt to special load scenarios.

[0090] In some embodiments, the higher the loading capacity level, the higher the corresponding washing pump speed, washing temperature, washing time, and washing water volume; the lower the distribution uniformity level, the more washing liquid is used and the higher the washing pump speed, which in turn affects the variable speed frequency of the spray arm rotation.

[0091] In some embodiments, after receiving the load level and distribution uniformity level output by the load identification model, the main controller first performs anomaly judgment. The judgment logic is divided into two scenarios. One is a single anomaly scenario: if only the load level is detected to reach or exceed L4 and the distribution uniformity level is U0 (uniform distribution), it is determined to be a pure overload scenario; if only the distribution uniformity level is detected to reach or fall below U1 and the load level is between L0 and L3 (not overloaded), it is determined to be a pure uneven distribution scenario. The other is a compound anomaly scenario: if the load level is detected to reach or exceed L4 and the distribution uniformity level is detected to reach or fall below U1, it is determined to be a compound overload + uneven distribution scenario. In this case, the two types of parameter adjustment logic corresponding to overload and uneven distribution need to be executed simultaneously.

[0092] If the dishwasher is determined to be overloaded (including pure overload and overload dimensions in compound anomalies), the main controller sends parameter adjustment commands to the wash pump, heater, timing module, and water volume control unit (integrated into the wash pump) based on the specific load level. The adjustment rule follows that "the higher the load level, the greater the parameter adjustment range." For example, but not limited to, when the load level is L3 (3 / 4 load, corresponding to a total tableware weight of 3-4kg, including 21-28 standard plates, 6-8 soup bowls, and 10-15 pairs of chopsticks), the wash pump speed is increased from the base value of 1800r / min to 2000r / min, an increase of 11%, to enhance water flow penetration; the wash temperature is increased from the base value of 50℃ to 55℃, an increase of 10%, to accelerate the dissolution of grease stains; the wash time is extended from the base value of 30min to 35min, an extension of 17%, to ensure that stains have enough time to be decomposed; and the wash water volume is increased from the base value of 8L to 9L, an increase of 12.5%, to ensure that the water flow can cover more tableware. When the load level is L4 (full load), the washing pump speed is further increased to 2200 r / min (22% increase), the washing temperature is increased to 60℃ (20% increase), the washing time is extended to 40 min (33% extension), and the washing water volume is increased to 10L (25% increase), to meet the high cleaning requirements under full load through a greater increase in parameters. If an abnormal overload exceeding L4 occurs (total weight of tableware exceeds 5kg), the parameters will be adjusted to the upper limit: washing pump speed 2400 r / min (33% increase), washing temperature 65℃ (30% increase), washing time 45 min (50% extension), and washing water volume 11L (37.5% increase), to avoid incomplete cleaning due to overload.

[0093] If the dishwasher is determined to have uneven distribution of tableware (including pure uneven distribution and uneven distribution dimensions in compound anomalies), the main controller sends parameter adjustment instructions to the detergent dispenser and spray arm drive based on the specific distribution uniformity level. The adjustment rule follows that "the lower the distribution uniformity level (the more uneven the distribution), the greater the parameter adjustment range." For example, but not limited to, when the distribution uniformity level is U1 (slight unevenness), the detergent volume is increased from the base value of 15ml to 18ml, an increase of 20%, to compensate for cleaning blind spots by increasing the local detergent concentration; by controlling the washing pump speed to increase the rotation frequency of the spray arm from the base value of 1 time / minute (spinning speed every 60s) to 1.2 times / minute (spinning speed every 50s), an increase of 20%, to cover lightly stacked areas by slightly changing the water flow direction. When the distribution uniformity level is U2 (moderately uneven, corresponding to 1 / 3 of the area with stacked tableware or biased towards one side of the basket), the amount of detergent used is increased to 21ml (an increase of 40%), and the spray arm speed change frequency is increased to 1.5 times / minute (switching speed every 40 seconds, an increase of 50%). The higher concentration of detergent and more frequent speed changes enhance the cleaning power for moderately stacked areas. When the distribution uniformity level is U3 (severely uneven, corresponding to 1 / 2 of the area with messy tableware stacked more than 3 layers high), the amount of detergent used is increased to 24ml (an increase of 60%), and the spray arm speed change frequency is increased to 2 times / minute (switching speed every 30 seconds, an increase of 100%). The high-frequency speed change allows the water flow to create a dynamic impact, penetrating the gaps between the messy stacked tableware, while the high concentration of detergent ensures that stubborn stains are fully dissolved, significantly reducing cleaning blind spots.

[0094] The main controller sends parameter adjustment commands to each actuator via the CAN bus, and all actuators initiate adjustment actions within 100ms of receiving the command. Specifically, the washing pump changes its output speed via frequency conversion, the heater adjusts its heating power via pulse width modulation (PWM) signals, the detergent dispenser controls the dosage by controlling the on-time of the solenoid valve (e.g., 4.2s for 21ml of detergent, 3s for a baseline of 15ml), and the spray arm drive switches speed modes via a stepper motor driver. During parameter execution, the main controller collects feedback signals in real time through various sensors: the washing pump speed is monitored by a motor encoder with an accuracy of ±10r / min; the washing temperature is monitored by an NTC temperature sensor with an accuracy of ±1℃; the detergent dosage is monitored by a flow sensor with an accuracy of ±0.5ml; and the spray arm speed is monitored by feedback signals from the drive components. If the actual value of a certain parameter is detected to deviate from the commanded value by more than ±5% (e.g., the commanded speed of the washing pump is 2000 r / min, but the actual speed is only 1900 r / min), the main controller will immediately resend the adjustment command and simultaneously increase the drive power of the corresponding execution component (e.g., increase the drive voltage of the washing pump by 5%) until the washing parameter deviation is less than ±5%, ensuring that the washing parameters accurately match the actual load state of the dishwasher.

[0095] In one possible implementation, please refer to Figure 4 As shown, Figure 4 A schematic flowchart of an optional dishwasher control method provided in this application is shown. The process of constructing the sample dataset includes: S401, obtain a variety of pre-set different loading capacity scenarios, and a variety of different distribution state scenarios set under each loading capacity scenario.

[0096] Each loading scenario corresponds to a preset total weight range of tableware, and each distribution scenario corresponds to a preset tableware placement rule.

[0097] S402, after the dishwasher starts the washing program and controls the spray arm to run at a preset speed for a set time, collects the vibration signal of the spray arm under each load scenario.

[0098] S403 extracts time-domain and / or frequency-domain feature parameters from the acquired vibration signals to form vibration feature samples corresponding to each loading scenario.

[0099] S404 labels each vibration feature sample with its corresponding load level and distribution uniformity level, and then summarizes them to form a sample dataset.

[0100] First, based on preliminary research into household tableware usage habits, a two-dimensional scenario was designed, encompassing both loading capacity and distribution scenarios, to ensure the sample covers various typical usage situations. The loading capacity scenario was divided into five categories according to the total weight of the tableware, each corresponding to a specific weight range and tableware combination: L0 is the empty scenario, with a total tableware weight of 0kg, meaning the dishwasher contains no tableware; L1 is the 1 / 4 load scenario, with a total tableware weight of 1-2kg, specifically including 8-12 standard 20cm diameter plates (simulating the amount of tableware after a single person's daily meal); L2 is the 1 / 2 load scenario, with a total tableware weight of 2-3kg, including 13-20 standard plates + 15cm diameter plates. L3 is a 3 / 4 load scenario, with a total tableware weight of 3-4kg, including 21-28 standard plates, 6-8 soup bowls, and 10-15 pairs of wooden chopsticks (simulating tableware for a 4-5 person family meal); L4 is a full load scenario, with a total tableware weight of 4-5kg, including 29-35 standard plates, 9-12 soup bowls, 16-20 pairs of chopsticks, and 5-8 stainless steel spoons (simulating tableware for a 6-7 person family gathering).

[0101] Subsequently, for each loading capacity scenario, four distribution scenarios were further set up, each with specific tableware placement rules: U0 is a uniform distribution scenario, requiring tableware to be placed according to the rule of "plates upright in the basket rack slots, soup bowls scattered on the lower layer of the basket rack, and chopsticks and spoons classified and placed in dedicated storage compartments", ensuring that no area is stacked; U1 is a slightly uneven scenario, allowing tableware to be stacked in 1 / 4 area (such as the upper right corner of the basket rack), with a stacking height of no more than 2 layers, and the remaining areas are placed according to the uniform rule; U2 is a moderately uneven scenario, allowing tableware to be stacked in 1 / 3 area (such as the left half of the basket rack), with a stacking height of 2-3 layers, or the tableware is generally biased towards the edge of the basket rack (the tableware density in the edge area is twice that in the center area); U3 is a severely uneven scenario, allowing tableware to be stacked randomly in 1 / 2 area (such as the front half of the basket rack), with a stacking height of more than 3 layers, and without distinguishing tableware types (plates, soup bowls, and chopsticks are mixed together), simulating the extreme case of users placing tableware randomly. Through the above design, 5×4=20 combinations of "load capacity-distribution status" scenarios are formed, which fully cover the normal and extreme load conditions in home use.

[0102] Then, for each of the 20 combined scenarios, a standardized vibration signal acquisition process was executed. The acquisition operation strictly followed the signal acquisition logic of the actual washing program of the dishwasher to ensure that the sample data was consistent with the signal characteristics of the actual usage scenario. Before acquisition, the tableware for the corresponding scenario was placed in the dishwasher basket according to preset rules. After closing the dishwasher door, the washing program was started through the main controller (e.g., a uniform daily wash mode was selected to avoid the influence of different washing programs on the spray arm control logic). After the washing program started, the main controller controlled the spray arm drive component according to preset logic, so that the spray arm ran at a preset speed of 1500r / min for 30 seconds (consistent with the pre-run process in the actual washing, eliminating the interference of startup inertia). After the spray arm had run stably for 30 seconds, the main controller triggered the piezoelectric vibration sensor installed on the spray arm base to start acquiring the vibration signal of the spray arm. The acquisition time was set to 10 seconds (which can cover 2-3 spray arm rotation cycles to ensure the capture of complete vibration characteristics).

[0103] To avoid the randomness of single data collection, 50 sets of vibration signals were repeatedly collected as samples for each scenario combination. During the collection process, if sensor contact problems occurred (signal amplitude continuously <0.05g) or abnormal vibration of the base occurred (signal amplitude continuously >2g), the data set was immediately discarded and recollected to ensure that each sample set was valid. After collection, each set of vibration signals was stored in the computer database using the naming rule of "Scene Number - Collection Number" (e.g., "L2-U1-03" represents the 3rd sample set of L2 loading + U1 distribution state) for subsequent feature extraction operations.

[0104] For all vibration signal samples stored in the database, feature extraction is performed one by one. The extraction logic is completely consistent with the feature extraction logic of the main controller in the actual washing process, ensuring that the sample features are consistent with the input feature types during model inference. Preprocessing of the vibration signal is also possible, but not limited to: using a 50Hz notch filter to remove mains interference, then using a 10-500Hz bandpass filter to retain the effective vibration frequency band, and then converting the analog signal to a digital signal through a 12-bit analog-to-digital converter; after preprocessing, two types of feature parameters are extracted from the digital signal: time-domain feature parameters include peak amplitude (by iterating through the maximum value of the digital signal and converting it according to the sensor calibration relationship of "amplitude g = 0.001 × voltage V") and root mean square amplitude (by taking the square root of the square mean of all sampling points); frequency-domain feature parameters are obtained through Fast Fourier Transform (FFT), including the main vibration frequency (the frequency point with the highest energy in the frequency domain signal), the proportion of spectral energy in 10-50Hz (the first preset frequency band, low frequency band) and the proportion in 100-300Hz (the second preset frequency band, high frequency band) (the percentage of the total energy of the two frequency bands in the total energy of 10-500Hz is calculated respectively).

[0105] The “peak amplitude + root mean square amplitude + main vibration frequency + low frequency energy ratio + high frequency energy ratio” extracted from each sample are combined into a feature vector, which serves as the core feature representation of the sample. At the same time, it is stored in association with the original vibration signal to ensure the traceability of features and signals.

[0106] For each feature vector sample, a labeling operation is performed. The labels are divided into two categories, corresponding to the loading level and distribution uniformity level of the scenario: Based on the loading scenario to which the sample belongs, a loading level label is assigned (L0 corresponds to label "0", L1 to "1", L2 to "2", L3 to "3", and L4 to "4"); based on the distribution scenario to which the sample belongs, a distribution uniformity level label is assigned (U0 corresponds to label "0", U1 to "1", U2 to "2", and U3 to "3"). For example, a sample belonging to the "L3 loading level + U2 distribution state" combined scenario is labeled as "loading level 3, distribution uniformity level 2".

[0107] After annotation, the "feature vector + dual label" of all samples are summarized in a structured format (such as CSV) to form a complete sample dataset. This sample dataset contains 20 combined scenarios × 50 sets of samples = 1000 valid samples. Then, the dataset is divided into a training set (700 samples, used for model parameter training) and a test set (300 samples, used for model accuracy verification) in a 7:3 ratio. The partitioning process uses random sampling to ensure that the scene distribution of the training set and the test set is consistent, avoiding model bias caused by uneven scene distribution.

[0108] In one possible implementation, after inputting time-domain feature parameters and / or frequency-domain feature parameters into a pre-trained load recognition model to obtain the load quantity level and distribution uniformity level of the dishes loaded in the dishwasher output by the load recognition model, the method further includes: If the uniformity level is lower than the preset uniformity threshold, it is determined that the dishwasher has uneven distribution of tableware.

[0109] If the load capacity level exceeds the preset load capacity threshold, it is determined that the dishwasher is overloaded.

[0110] If the uniformity level is lower than the preset uniformity threshold and the load level exceeds the preset load threshold, it is determined that the dishwasher is overloaded and the tableware is unevenly distributed.

[0111] If the dishwasher is overloaded and / or the dishes are unevenly distributed, a warning message will be displayed.

[0112] The prompt message is output via the dishwasher's control panel and / or via a mobile terminal associated with the dishwasher, and the prompt message includes at least instructions on how to rearrange the dishes.

[0113] After the load recognition model outputs the load level and distribution uniformity level of the tableware inside the dishwasher, the main controller first calls two pre-stored judgment thresholds to perform abnormal state judgment. Among them, the preset load threshold is set to load level L4, which corresponds to the full load state of the dishwasher, specifically the total weight of tableware is 4-5kg, including 29-35 standard plates, 9-12 soup bowls with a diameter of 15cm, 16-20 pairs of wooden chopsticks, and 5-8 stainless steel spoons. When the load level output by the model is L4 or exceeds L4 (such as the total weight of tableware exceeding 5kg, the corresponding level is marked as L5), the main controller determines that the dishwasher is "overloaded". The preset uniformity threshold is set to distribution uniformity level U1, which corresponds to a slightly uneven state, specifically when 1 / 4 of the area has tableware stacking (stack height not exceeding 2 layers) or when the tableware density in a local area is 1.5 times higher than that in other areas. Since the higher the distribution uniformity level value, the more uneven the distribution (U0 is uniform, U1 is slightly uneven, U2 is moderately uneven, and U3 is severely uneven), when the distribution uniformity level output by the model is U1, U2, or U3, the main controller determines that the dishwasher "has uneven tableware distribution".

[0114] The main controller's judgment logic covers three scenarios: if only "load level ≥ L4" and distribution uniformity level is U0, it is judged as a "single overload scenario"; if only "distribution uniformity level ≥ U1" and load level ≤ L3, it is judged as a "single uneven distribution scenario"; if both "load level ≥ L4" and "distribution uniformity level ≥ U1" are met, it is judged as a "combined overload + uneven distribution scenario". The judgment result is stored in real time in the main controller's temporary buffer area as the basis for triggering prompt information output.

[0115] When the main controller determines that the dishwasher is overloaded and / or the dishes are unevenly distributed, it will simultaneously output prompt information through the dishwasher's control panel, such as the display panel, buzzer, and / or associated mobile terminal, to ensure that the user can get timely feedback whether they are in the kitchen or away from the dishwasher.

[0116] In some embodiments, the display panel on the dishwasher's control panel presents prompts in a combination of text and icons. The display panel can be a touchscreen LCD that supports multi-color text and dynamic icons. For a "single overload scenario," the screen displays bold red text in the center: "Warning: Overloaded tableware, please reduce the load," accompanied by a yellow "Weight Exceeded Limit" icon (a tilted balance scale), and specific guidance text: "It is recommended to keep the total weight of tableware within 4-5 kg; 3-5 standard plates can be removed." For a "single uneven distribution scenario," the guidance is refined according to the distribution uniformity level: if it is U1 (slightly uneven), orange text "Tip: Tableware is partially stacked; please distribute it more evenly," accompanied by a blue "Uneven Distribution" icon (a stacked tableware icon), and the guidance content is... "It is recommended to distribute the stacked plates in the right-side basket 1 / 4 area to the empty area on the left." If it is U2 (moderate unevenness) or U3 (severe unevenness), the red text "Warning: The tableware is severely unevenly distributed and needs to be rearranged" will be displayed, and the instructions will be "Please sort the messy tableware in the front half of the basket, place the plates upright, and distribute the soup bowls to the lower basket." For "combined scenarios", two types of prompts will be integrated. First, an overload warning will be displayed, and then an uneven distribution prompt will be displayed. Each type of prompt will be displayed for 3 seconds and then cycled until the user triggers the "confirm" touch button or adjusts the distribution of tableware to restart the washing program.

[0117] In some embodiments, notification messages can be pushed to a mobile terminal associated with the dishwasher via a dedicated app. The mobile terminal must be paired with the dishwasher beforehand via Bluetooth or Wi-Fi. For example, after the main controller detects an anomaly, it sends a notification command to the associated app via the dishwasher's built-in Wi-Fi module. Upon receiving the command, the app immediately displays a pop-up notification titled "Dishwasher Loading Anomaly Reminder," with content consistent with the notification displayed on the dashboard, and includes "View Details" and "Processed" operation buttons. Clicking "View Details" redirects to the "Loading Guide" page within the app, which includes a diagram of the dishwasher's current loading status (marking overloaded or unevenly distributed areas), detailed placement steps (e.g., "Step 1: Remove 2 soup bowls; Step 2: Move the upper right plate to the lower left"), and a standard placement example diagram to help users make adjustments more intuitively. Furthermore, if the user does not click "Processed" within 10 minutes, the app will send another push notification to ensure the user receives timely feedback on overloading or uneven tableware distribution and can make adjustments according to the guide, solving the problem of traditional dishwashers lacking intelligent feedback.

[0118] In addition, the main controller will continue to monitor user operations after the prompt message is output: if the user opens the dishwasher door (detected by the door sensor), the prompt will be paused; if the washing program is restarted after the door is closed, the main controller will re-execute vibration signal acquisition and load identification. If the abnormal state has been resolved, the prompt will no longer be output; if the abnormality still exists, the dual-channel prompt will be triggered again until the abnormality is resolved, ensuring that the dishwasher always operates under a reasonable load state and avoiding washing failure or equipment damage caused by abnormal loading.

[0119] One possible implementation is, such as Figure 5 As shown, after adjusting the dishwasher's washing parameters according to the tableware loading level and distribution uniformity level, the method further includes: S501 controls the dishwasher to run a preset cycle according to the adjusted washing parameters.

[0120] S502, the vibration signal of the spray arm is collected again and the feature parameter extraction and model recognition steps are repeated to obtain the load level and distribution uniformity level identified in the second identification.

[0121] S503, if the dishwasher is still determined to be overloaded and / or the tableware is unevenly distributed based on the load level and distribution uniformity level obtained from the secondary identification, the washing parameters of the dishwasher are adjusted a second time.

[0122] S504 controls the dishwasher to operate according to the second-adjusted washing parameters.

[0123] In some embodiments, to ensure that the washing process always adapts to the dynamic changes in the tableware loading status and avoids insufficient cleaning due to initial identification errors or mid-process load changes, the main controller adjusts the washing parameters based on the initially identified loading level and distribution uniformity level and sends them to each execution component. First, it controls the dishwasher to start the washing operation according to the adjusted parameters and simultaneously triggers the built-in timing module to set a preset cycle. The duration of this cycle is determined based on the total washing time, uniformly set to half of the total washing time after the initial adjustment, with a minimum of 5 minutes and a maximum of 15 minutes. For example: if the total washing time after the initial adjustment is 30 minutes (corresponding to L2 loading level + U0 distribution status), the preset cycle is set to 15 minutes; if the total washing time after the initial adjustment is 20 minutes (corresponding to L0 empty load), the preset cycle is set to 10 minutes; if the total washing time is 40 minutes (corresponding to L4 full load + U3 severe unevenness), the preset cycle is set to 15 minutes (taking the upper limit). This is because when the washing cycle is halfway through, the dishes have already come into initial contact with the water flow and may have slightly shifted (such as stacked plates becoming loose or edge dishes shifting). At this time, secondary recognition can more accurately capture the dynamic load status and avoid the deviation between the initial static recognition and the dynamic load in the middle. At the same time, the duration of half a cycle can allow enough time to observe load changes without causing frequent adjustments and extending the total washing time due to the cycle being too short.

[0124] During the preset cycle of operation, the main controller receives feedback signals from each actuator (such as washing pump speed and washing temperature) in real time to ensure that the parameters are executed stably according to the adjusted values. If a parameter deviates by more than ±5%, it is immediately corrected in real time, providing a stable operating environment for secondary identification.

[0125] Once the dishwasher has completed the preset cycle according to the adjusted parameters, the main controller pauses the current washing action (only pausing the spraying and heating, keeping the dishes soaking in the washing water to prevent the dishes from drying or the stains from solidifying), and then starts the "secondary signal acquisition-feature extraction-model recognition" process that is completely consistent with the initial recognition, ensuring that the standards of the two recognitions are consistent and the results are comparable.

[0126] The main controller performs the same preprocessing (filtering and analog-to-digital conversion) on the secondary vibration signal as it did on the primary signal. Then, it extracts time-domain feature parameters (peak amplitude and root mean square amplitude) and frequency-domain feature parameters (main vibration frequency, energy proportion in the 10-50Hz low-frequency band, and energy proportion in the 100-300Hz high-frequency band) to form a secondary feature vector. This secondary feature vector is then input into the same pre-trained load recognition model as the primary signal to obtain the load level and distribution uniformity level of the secondary recognition output by the load recognition model. For example, if the primary recognition is L3 load level + U2 distribution, the secondary recognition may change to L3 load level + U1 distribution due to tableware displacement, or to L4 load level + U3 distribution due to increased stacking.

[0127] The main controller calls the same preset thresholds as the initial one (load level threshold L4, uniformity threshold U1) to determine the abnormality of the secondary identification results: if the secondary identification load level ≤ L3 and the distribution uniformity level = U0, it is determined that "the abnormal state has been resolved", and no parameter adjustment is required. The main controller controls the dishwasher to continue running the remaining washing time according to the parameters adjusted in the initial step; if the secondary identification still meets "load level ≥ L4 (overload)" and / or "distribution uniformity level ≥ U1 (uneven distribution)", it is determined that "the abnormal state continues", and the secondary parameter adjustment is immediately initiated.

[0128] It should be understood that the rules for secondary adjustments follow the principle of incremental increases based on the initial adjustment values. The magnitude of the increase is determined by the deviation between the secondary identification level and the initial level. For example, for secondary adjustments in overload scenarios: if the secondary identification load level is one level higher than the initial level (e.g., L3 initially, L4 later), then each washing parameter will be increased by 10% based on the initial adjustment of the washing pump speed, washing temperature, washing time, and washing water volume (e.g., the initial washing pump speed is 2000 r / min, the secondary adjustment is 2200 r / min; the initial washing temperature is 55℃, the secondary adjustment is 60.5℃). If the secondary identification level is the same as the initial level (e.g., L4 initially, L4 later), then each parameter will be increased by 5% based on the initial adjustment value (e.g., the initial washing time is 40 minutes, the secondary adjustment is 42 minutes) to avoid incomplete cleaning due to continuous overload.

[0129] For secondary adjustments to uneven distribution scenarios, if the distribution uniformity level identified in the secondary recognition is improved by 1 level compared to the initial level (e.g., initial U1, secondary U2), then each washing parameter is increased by 20% based on the initial adjustment of detergent volume and spray arm speed frequency (e.g., initial detergent volume 18ml, secondary adjustment 21.6ml; initial speed frequency 1.2 times / minute, secondary adjustment 1.44 times / minute). If the secondary recognition level is the same as the initial level (e.g., initial U3, secondary still U3), then each parameter is increased by 10% based on the initial adjustment value (e.g., initial speed frequency 2 times / minute, secondary adjustment 2.2 times / minute). By using a higher concentration of detergent and more frequent speed switching, the persistent uneven distribution can be addressed.

[0130] The main controller sends the adjusted parameters to each execution component. The execution components respond and initiate adjustments within 100ms. The main controller then controls the dishwasher to continue running the remaining wash time. During the remaining run, the main controller monitors parameter execution in real time using sensors (e.g., washing pump speed monitored by an encoder, detergent volume monitored by a flow sensor). If the washing parameters deviate by more than ±5%, the controller immediately resends the correction command. Simultaneously, a third identification and adjustment is not initiated. This avoids frequent adjustments that extend the total wash time, and the second adjustment is sufficient to cover most dynamic load variations. Experimental data shows that after the second adjustment, the cleaning pass rate can be increased from 92% after the initial adjustment to 98%, meeting actual usage needs. If extreme anomalies still occur after the second adjustment (e.g., a continuous deviation of washing parameters exceeding 10%), the dishwasher's control panel will display a "Load abnormality, please check the dishes" message, ensuring the safe operation of the dishwasher.

[0131] In one possible implementation, the time-domain characteristic parameters include the peak amplitude, root mean square amplitude, and vibration duration of the vibration signal, while the frequency-domain characteristic parameters include the main vibration frequency of the vibration signal, the proportion of spectral energy in the first preset frequency band, and the proportion of spectral energy in the second preset frequency band, wherein the frequency range of the second preset frequency band is greater than that of the first preset frequency band.

[0132] In one possible implementation, during the training of the load identification model, the lower the load level, the higher the main vibration frequency and the lower the peak amplitude of the vibration signal; the higher the load level, the lower the main vibration frequency and the higher the peak amplitude of the vibration signal, and the spectral energy shifts towards the first preset frequency band.

[0133] It should be understood that time-domain feature parameters are extracted based on the time dimension data of vibration signals, directly reflecting the intensity, average energy and effective acquisition duration of the spray arm vibration, including parameters such as peak amplitude, root mean square amplitude and vibration duration. The extraction process requires the preprocessing of the vibration signal (filtering to remove mains interference and analog-to-digital conversion) before calculation based on the preprocessed digital signal.

[0134] Peak amplitude is the maximum vibration intensity of the vibration signal within the acquisition time, reflecting the maximum vibration amplitude of the spray arm under load. During extraction, all digital signal sampling points (10,000 in total, sampling frequency 1kHz) within the 10-second acquisition time are traversed first to find the voltage value corresponding to each sampling point. Then, the amplitude value of each sampling point is calculated according to the calibration relationship of piezoelectric vibration sensor "amplitude g = 0.001 × voltage V" (each 1mV output voltage of the sensor corresponds to an amplitude of 0.001g, where g is the gravitational acceleration 9.8m / s²). Finally, the maximum value among all amplitude values ​​is taken as the peak amplitude.

[0135] In actual load scenarios, this parameter shows a significant upward trend with the increase of load: when unloaded (L0), the peak amplitude is usually 0.1-0.3g (corresponding to a voltage of 0.1-0.3V, digital signal value 82-245), and the vibration amplitude is small because there is no load obstruction on the spray arm; when fully loaded (L4), the peak amplitude can reach 0.5-0.8g (corresponding to a voltage of 0.5-0.8V, digital signal value 409-655), and the vibration amplitude is significantly increased when rotating due to the increased resistance of the tableware to the spray arm; in uneven distribution scenarios, if there is local stacking (such as U3 severe unevenness), the peak amplitude will be 0.1-0.2g higher than that when uniformly distributed (U0) with the same load, and the vibration fluctuation is aggravated due to the difference in local resistance.

[0136] The root mean square amplitude is the average energy representation of the vibration signal over the acquisition period. It can avoid the random deviation of a single peak amplitude and more comprehensively reflect the overall intensity of the vibration. The calculation steps are as follows: first, obtain the amplitude values ​​of 10,000 sampling points according to the above method, then calculate the sum of squares of all amplitude values, divide it by the total number of sampling points (10,000) to obtain the mean square value, and finally take the square root of the mean square value to obtain the root mean square amplitude.

[0137] Vibration duration refers to the effective duration of vibration signal acquisition. The duration of each acquisition is consistent to avoid feature deviations due to duration differences. In actual operation, after the spray arm runs stably at 1500 r / min for 30 seconds, the main controller sends an acquisition command to the vibration sensor. The time from receiving the command to stopping acquisition is controlled to 10 seconds, which can cover 2-3 complete rotation cycles of the spray arm (1500 r / min corresponds to 1 rotation every 0.04 seconds), but is not limited to 10 seconds. It can completely capture the periodic vibration characteristics of the spray arm during rotation, without missing the periodic vibration fluctuations caused by the load, and without increasing the data processing volume.

[0138] Frequency domain characteristic parameters are extracted after converting the time-domain digital signal into a frequency-domain signal using Fast Fourier Transform (FFT). These parameters reflect the frequency distribution of the vibration signal. Since the vibration frequency of the spray arm varies significantly under different loads (the larger the load, the lower the vibration frequency), these parameters can effectively distinguish the load levels. These parameters include the main vibration frequency, the proportion of spectral energy in the first preset frequency band, and the proportion in the second preset frequency band. The first preset frequency band is set to 10-50Hz (low frequency band), and the second preset frequency band is set to 100-300Hz (high frequency band). The frequency range of the second preset frequency band (200Hz) is greater than that of the first preset frequency band (40Hz). This division is based on previous experiments: the effective frequency of the spray arm vibration is concentrated in the 10-300Hz range, and the load change has the most significant impact on the energy distribution of the 10-50Hz and 100-300Hz frequency bands.

[0139] The dominant vibration frequency is the frequency point with the highest energy in the frequency domain signal, representing the dominant frequency of the spray arm vibration. When extracting it, first perform FFT transformation on the 10-second time domain digital signal (using 1024-point FFT, with a frequency resolution of about 0.977Hz, which can accurately distinguish the energy difference between adjacent frequencies) to obtain the energy value (in dB) corresponding to different frequency points. Then, traverse all frequency points to find the frequency point with the largest energy value. This frequency point is the dominant vibration frequency.

[0140] The percentage of spectral energy in the first preset frequency band (10-50Hz) is the percentage of the total energy of all frequency points within the 10-50Hz band to the total energy of the effective frequency band (covering all vibration frequencies of the spray arm), reflecting the proportion of low-frequency vibration energy. Since the greater the load, the more significant the low-frequency vibration, this parameter increases with the load: at no load (L0), low-frequency energy is low, accounting for only 10-15%; at L1, it rises to 15-20%; at L2, 20-25%; at L3, 25-30%; and at L4, it can reach 30-40%. In unevenly distributed scenarios, local stacking increases low-frequency vibration. For example, at full load (L4), the percentage of uniformly distributed U0 is 32%, while at severely uneven U3, it rises to 38%. This parameter can be used to distinguish differences in distribution under the same load.

[0141] The proportion of spectral energy in the second preset frequency band (100-300Hz) is the percentage of the total energy of all frequency points within the 100-300Hz band to the total energy of the effective frequency band (10-500Hz), reflecting the proportion of high-frequency vibration energy. Since higher loads result in weaker high-frequency vibrations, this parameter decreases with increasing load and is negatively correlated with the proportion in the first preset frequency band: at no load (L0), high-frequency energy is abundant, accounting for 30-40%; at L1, it drops to 25-30%; at L2, 20-25%; at L3, 15-20%; and at L4, only 10-15%. In unevenly distributed scenarios, the proportion of high-frequency energy is 5-8% lower than when the load is evenly distributed. For example, at L2 with a uniform load distribution (U0), the proportion of high-frequency energy is 23%, but at U2 with moderate unevenness, it drops to 17%, further confirming the influence of distribution on vibration frequency distribution.

[0142] By extracting the time-domain and frequency-domain feature parameters mentioned above, the characteristics of the spray arm vibration signal can be comprehensively characterized from four dimensions: intensity, energy, duration, and frequency distribution. Moreover, all parameters have clear correlations with the tableware loading amount and distribution uniformity, providing rich and effective feature basis for the subsequent load identification model to accurately distinguish different load states.

[0143] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0144] Corresponding to the dishwasher control method described in the above embodiments, Figure 6 This is a schematic diagram of the structure of a dishwasher control device provided in an embodiment of this application. The device can be implemented as part or all of a computer device by software, hardware, or a combination of both. This computer device can be... Figure 7 The electronic device shown.

[0145] Reference Figure 6 The dishwasher's control unit includes: The acquisition unit 601 is used to acquire the vibration signal of the dishwasher's spray arm in response to the start of the dishwasher's washing program.

[0146] Extraction unit 602 is used to process the vibration signal to extract time-domain and / or frequency-domain feature parameters from the vibration signal.

[0147] The processing unit 603 is used to input time-domain feature parameters and / or frequency-domain feature parameters into the pre-trained load recognition model to obtain the load level and distribution uniformity level of the tableware loaded in the dishwasher output by the load recognition model.

[0148] The control unit 604 is used to adjust the washing parameters of the dishwasher according to the loading level and distribution uniformity level of the tableware.

[0149] The load identification model is trained on a sample dataset using a machine learning algorithm. The sample dataset contains vibration feature samples under different loads and vibration feature samples under different tableware distribution states. Each vibration feature sample is associated with a corresponding load level label and distribution uniformity level label.

[0150] It is understood that the dishwasher control device embodiments and any implementation thereof correspond to the dishwasher control method embodiments and any implementation thereof. The technical effects corresponding to the dishwasher control device embodiments and any implementation thereof can be found in the aforementioned technical effects corresponding to the dishwasher control method embodiments and any implementation thereof, and will not be repeated here.

[0151] This application also provides an electronic device, which includes one or more processors and a memory; The memory is coupled to one or more processors. The memory is used to store computer program code, which includes computer instructions. One or more processors call the computer instructions to cause the electronic device to perform the aforementioned control method for the dishwasher.

[0152] Figure 7This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device 700 can be a mobile phone, smart screen, tablet computer, wearable electronic device, in-vehicle electronic device, augmented reality (AR) device, virtual reality (VR) device, laptop computer, ultra-mobile personal computer (UMPC), netbook, personal digital assistant (PDA), projector, or a communication device such as a server, storage device, or base station, or a smart car, etc. This application embodiment does not impose any limitations on the specific type of electronic device.

[0153] The memory 701 can be used to store computer software programs 702 and modules. The processor 703 executes various functional applications and data processing of the electronic device by running the software programs and modules stored in the memory 701. The memory 701 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device (such as audio data, telephone directory, etc.). In addition, the memory 701 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0154] The processor 703 may include one or more processors such as a central processing unit (CPU), an application processor (AP), and a baseband processor. The processor can serve as the nerve center and command center of the wireless router. The processor 703 can generate operation control signals based on instruction opcodes and timing signals to control instruction fetching and execution. The memory 701 can be used to store executable program code, including instructions. The processor 703 executes various functional applications and data processing of the network device by running the instructions stored in the memory. The memory 701 may include a program storage area and a data storage area, such as storing data for audio signals to be played. For example, the memory may be Double Data Rate Synchronous Dynamic Random Access Memory (DDR) or Flash memory.

[0155] This application also provides a computer-readable storage medium storing computer instructions; when the computer-readable storage medium is used on an electronic device, it causes the electronic device to execute the aforementioned dishwasher control method.

[0156] The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or can include one or more data storage devices such as servers or data centers that can be integrated with media. The available medium can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media, or semiconductor media (e.g., solid-state disks (SSDs)).

[0157] This application also provides a computer program product containing computer instructions, which, when run on an electronic device, enables the electronic device to execute the aforementioned dishwasher control method.

[0158] The computer storage medium and computer program product provided in the embodiments of this application are used to execute the methods provided above. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects corresponding to the methods provided above, and will not be repeated here.

[0159] In the above embodiments, implementation can also be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, optical fiber, Digital Subscriber Line, DSL) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access, or a data storage device such as a server or data center that integrates one or more available media. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc., and the storage medium can also include combinations of the above types of memory.

[0160] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0161] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments claimed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0162] In the embodiments provided in this application, it should be understood that the disclosed apparatus / network devices and methods can be implemented in other ways. For example, the apparatus / network device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0163] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0164] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for controlling a dishwasher, characterized in that, include: In response to the dishwasher starting the washing program, the vibration signal of the dishwasher's spray arm is acquired; The vibration signal is processed to extract time-domain and / or frequency-domain feature parameters from the vibration signal; The time-domain feature parameters and / or the frequency-domain feature parameters are input into the pre-trained load recognition model to obtain the load level and distribution uniformity level of the tableware loaded in the dishwasher output by the load recognition model. The load recognition model is trained on a sample dataset using a machine learning algorithm. The sample dataset includes vibration feature samples under different load levels and vibration feature samples under different tableware distribution states. Each vibration feature sample is associated with a corresponding load level label and distribution uniformity level label. The washing parameters of the dishwasher are adjusted according to the loading level and distribution uniformity level of the tableware.

2. The method according to claim 1, characterized in that, The step of acquiring the vibration signal of the dishwasher's spray arm in response to the dishwasher starting the washing program includes: In response to the dishwasher starting the washing program, the spray arm is controlled to run at a preset speed for a set duration; After the spray arm has been running at the preset speed for the preset duration, the vibration sensor installed on the spray arm base of the dishwasher is controlled to collect the vibration signal of the spray arm. The vibration signal of the spray arm collected by the vibration sensor is obtained.

3. The method according to claim 1, characterized in that, The washing parameters include at least one of the following: washing pump speed, washing temperature, washing liquid volume, washing time, and washing water volume; The step of adjusting the washing parameters of the dishwasher according to the loading capacity level and the distribution uniformity level includes: If the distribution uniformity level is lower than a preset uniformity threshold, and / or the loading level exceeds a preset loading threshold, it is determined that the dishwasher is overloaded and / or the tableware distribution is uneven. If the dishwasher is overloaded, at least one of the following is adjusted based on the load level: the washing pump speed, the washing temperature, the washing duration, and the washing water volume. If the dishwasher has uneven distribution of tableware, the amount of detergent and the speed of the washing pump are adjusted based on the distribution uniformity level.

4. The method according to claim 3, characterized in that, The higher the loading capacity level, the higher the corresponding washing pump speed, the higher the corresponding washing temperature, the longer the corresponding washing time, and the more washing water volume. The lower the distribution uniformity level, the more washing liquid is used and the higher the washing pump speed.

5. The method according to claim 1, characterized in that, The process of constructing the sample dataset includes: The system acquires a variety of pre-set loading scenarios and a variety of different distribution scenarios under each loading scenario. Each loading scenario corresponds to a pre-set total weight range of tableware, and each distribution scenario corresponds to a pre-set tableware placement rule. After the dishwasher starts the washing program and controls the spray arm to run at a preset speed for a set time, the vibration signal of the spray arm is collected under each of the loading scenarios. Time-domain feature parameters and / or frequency-domain feature parameters are extracted from the collected vibration signals to form vibration feature samples corresponding to each loading scenario; Each vibration feature sample is labeled with a corresponding load level label and distribution uniformity level label, and the samples are then aggregated to form the sample dataset.

6. The method according to any one of claims 1 to 5, characterized in that, After inputting the time-domain feature parameters and / or the frequency-domain feature parameters into a pre-trained load recognition model to obtain the load quantity level and distribution uniformity level of the tableware loaded in the dishwasher output by the load recognition model, the method further includes: If the distribution uniformity level is lower than a preset uniformity threshold, and / or the loading level exceeds a preset loading threshold, it is determined that the dishwasher is overloaded and / or the tableware distribution is uneven. If the dishwasher is overloaded and / or the tableware is unevenly distributed, a prompt message will be output. The prompt message will be output through the control panel of the dishwasher and / or through a mobile terminal associated with the dishwasher. The prompt message will include at least instructions on how to rearrange the tableware.

7. The method according to any one of claims 1 to 5, characterized in that, After adjusting the washing parameters of the dishwasher according to the loading level and distribution uniformity level of the tableware, the method further includes: The dishwasher is controlled to run a preset cycle according to the adjusted washing parameters; The vibration signal of the spray arm is collected again and the feature parameter extraction and model recognition steps are repeated to obtain the loading level and distribution uniformity level identified in the second recognition. If, based on the loading level and distribution uniformity level obtained from the secondary identification, it is still determined that the dishwasher is overloaded and / or the tableware distribution is uneven, the washing parameters of the dishwasher are adjusted a second time. The dishwasher is controlled to operate according to the washing parameters after secondary adjustment.

8. The method according to any one of claims 1 to 5, characterized in that, The time-domain characteristic parameters include the peak amplitude, root mean square amplitude, and vibration duration of the vibration signal. The frequency-domain characteristic parameters include the main vibration frequency of the vibration signal, the proportion of spectral energy in the first preset frequency band, and the proportion of spectral energy in the second preset frequency band, wherein the frequency range of the second preset frequency band is greater than that of the first preset frequency band.

9. A dishwasher, characterized in that, It includes a spray arm base, a vibration sensor mounted on the spray arm base, a main controller, and an actuator; The main controller, connected to the vibration sensor, is configured to control the vibration sensor to collect vibration signals from the dishwasher's spray arms in response to the dishwasher starting the washing program. The main controller, connected to the execution component, is further configured to process the vibration signal, extract time-domain feature parameters and / or frequency-domain feature parameters; input the time-domain feature parameters and / or the frequency-domain feature parameters into a pre-trained load recognition model to obtain the loading amount level and distribution uniformity level of the tableware loaded in the dishwasher output by the load recognition model; and control the execution component to perform corresponding operations based on the loading amount level and distribution uniformity level to adjust the washing parameters of the dishwasher. The load identification model is trained on a sample dataset using a machine learning algorithm. The sample dataset includes vibration feature samples under different loads and vibration feature samples under different tableware distribution states. Each vibration feature sample is associated with a corresponding load level label and distribution uniformity level label.

10. The dishwasher according to claim 9, characterized in that, The washing parameters include at least one of the following: washing pump speed, washing temperature, washing liquid volume, washing time, and washing water volume; the execution components include at least one of the following: spray arm drive, washing liquid distributor, washing pump, and heater. The main controller controls the execution components to perform corresponding operations based on the loading capacity level and distribution uniformity level, in order to adjust the washing parameters of the dishwasher, specifically configured as follows: If the distribution uniformity level is lower than a preset uniformity threshold, and / or the loading level exceeds a preset loading threshold, it is determined that the dishwasher is overloaded and / or the tableware distribution is uneven. If the dishwasher is overloaded, at least one of the detergent dispenser, the washing pump, and the heater will be controlled to perform corresponding operations based on the load level to adjust at least one of the washing pump speed, the washing temperature, the washing time, and the washing water volume. If the dishwasher has uneven distribution of tableware, at least one of the detergent dispenser and the spray arm drive is controlled to perform corresponding operations based on the distribution uniformity level to adjust the amount of detergent used and the speed of the washing pump.

Images