Method for controlling operation of laundry treating apparatus and laundry treating apparatus

Abstract

The present disclosure relates to a method for controlling operation of a clothes processing apparatus and a clothes processing apparatus. The method comprises: collecting operation data of the clothes processing apparatus; performing feature extraction on the operation data of the clothes processing apparatus to obtain a multi-dimensional feature vector; inputting the multi-dimensional feature vector into a plurality of preset sub-prediction models to obtain sub-drying probability values respectively output by the sub-prediction models; performing fusion processing on the plurality of sub-drying probability values to obtain a clothes-dried fusion probability, wherein the clothes-dried fusion probability is used to represent a possibility of the clothes reaching a dry state; and controlling an operation state of the clothes processing apparatus according to the clothes-dried fusion probability. Based on the method provided in the present disclosure, the probability of misjudgment can be reduced, and thus the accuracy of clothes drying judgment can be improved.

Description

Technical Field

[0001] This disclosure relates to the field of smart home appliance control, and in particular to a method for controlling the operation of clothing processing equipment and a clothing processing device. Background Technology

[0002] In the final stage of drying, heat pump clothing processing equipment typically uses sensor information such as temperature and humidity to automatically determine the drying status of the clothes and controls the program to end accordingly.

[0003] However, existing judgment schemes mostly use a single physical quantity and its fixed rules for judgment, which is prone to misjudgment and has low accuracy. Summary of the Invention

[0004] To overcome the problems existing in the related technologies, this disclosure provides an operation control method for clothing processing equipment and a clothing processing device.

[0005] According to a first aspect of the present disclosure, a method for controlling the operation of a garment processing device is provided, comprising:

[0006] Collect operational data from garment processing equipment;

[0007] Feature extraction is performed on the operating data of the clothing processing equipment to obtain a multidimensional feature vector; the multidimensional feature vector is then input into multiple preset sub-prediction models to obtain the sub-drying probability value output by each sub-prediction model.

[0008] Multiple sub-drying probability values ​​are fused to obtain the clothing dry fusion probability, which is used to characterize the likelihood that the clothing has reached a dry state; the operating status of the clothing processing equipment is controlled according to the clothing dry fusion probability.

[0009] The beneficial effects of this embodiment are as follows: by determining the fusion probability of whether the clothes are dry based on the multi-dimensional feature vector, a more comprehensive and uncertain quantitative judgment on whether the clothes are dry can be made, which can reduce the probability of misjudgment and thus improve the accuracy of judging whether the clothes are dry.

[0010] In one embodiment of this disclosure, it further includes:

[0011] Multiple consecutive sliding windows are used to extract features from the operating data of the clothing processing equipment, resulting in multidimensional feature vectors corresponding to each sliding window.

[0012] The multidimensional feature vectors corresponding to each of the multiple sliding windows are input into multiple preset sub-prediction models to obtain the sub-drying probability values ​​corresponding to each sliding window output by each sub-prediction model.

[0013] The sub-drying probability values ​​corresponding to each sliding window output by each sub-prediction model are fused to obtain the fused probability of clothes being dry for each sliding window;

[0014] The operating status of the clothing processing equipment is controlled based on the fusion probability of the clothes being dry corresponding to each of the multiple consecutive sliding windows.

[0015] The beneficial effects of this embodiment are as follows: by using a sliding window to extract features from the operating data of the clothing processing equipment, the operating data of the clothing processing equipment can be decomposed into local time periods and retain the information on phased changes. Based on the multi-dimensional feature vectors of each sliding window, the dynamic change characteristics of the operating data of the clothing machine can be accurately reflected, making the subsequent input for judging the degree of dryness of the clothes more complete, thereby improving the accuracy of the subsequent judgment of the degree of dryness of the clothes.

[0016] In one embodiment of this disclosure, the multidimensional feature vector includes at least two of the following: time feature data, contact humidity feature data, drum outlet humidity feature data, exhaust temperature feature data, drum outlet temperature feature data, condenser feature data, heat pump system temperature difference feature data, and intelligent power module temperature feature data.

[0017] The beneficial effects of this embodiment are as follows: The multi-dimensional feature vector provided in this example can simultaneously reflect the linkage between temperature, humidity and component thermal state of the clothing processing equipment during operation. It can integrate the operating data of various aspects of the clothing processing equipment in the same feature space, which can ensure the integrity of the drying judgment input, thereby making the operation control results of the clothing processing equipment more consistent with the actual drying state.

[0018] In one embodiment of this disclosure, a plurality of preset sub-prediction models include a first type of gradient boosting tree model, a second type of gradient boosting tree model, and a third type of gradient boosting tree model;

[0019] The sub-drying probability values ​​include the first probability, second probability, and third probability output by the first type gradient boosting tree model, the second type gradient boosting tree model, and the third type gradient boosting tree model, respectively.

[0020] The beneficial effect of this embodiment is that by performing parallel determination and fusion output of multiple models on the same multidimensional feature vector, the sensitivity of a single model to local fluctuations can be balanced in the comprehensive result, thereby improving the determination stability under critical drying scenarios.

[0021] In one embodiment of this disclosure, the sub-drying probability values ​​corresponding to each sliding window output by each sub-prediction model are fused to obtain the fused probability of clothes being dry for each sliding window, including:

[0022] Determine the first weight, second weight, and third weight corresponding to the first type gradient boosting tree model, the second type gradient boosting tree model, and the third type gradient boosting tree model, respectively;

[0023] Based on the first weight, the second weight, and the third weight, the first probability, the second probability, and the third probability corresponding to each sliding window are weighted and summed to obtain the fusion probability of the clothes being dry for each sliding window.

[0024] The beneficial effects of this embodiment are as follows: the probability of clothes being dry no longer depends on the output of a single model, but is determined by multiple models working together according to their weights, thus reducing the impact of single-model bias on the final determination. Since the fusion result is directly used for subsequent operational status control, this method enables the clothing processing equipment to make a more stable determination of the clothes' dryness status in the later stages of drying.

[0025] In one embodiment of this disclosure, the operating state of the clothing processing device is controlled based on the drying probability of each of a plurality of consecutive sliding windows, including:

[0026] A first preset probability threshold and a second preset probability threshold are determined for the clothes to be dried, wherein the first preset probability threshold is greater than the second preset probability threshold;

[0027] If the probability of the clothes being dried and fused corresponding to the sliding window is greater than or equal to the first preset probability threshold, then the drying status corresponding to the sliding window is determined to be that the clothes are dried; if the probability of the clothes being dried and fused corresponding to the sliding window is less than or equal to the second preset probability threshold, then the drying status corresponding to the sliding window is determined to be that the clothes are not dried; if the probability of the clothes being dried and fused corresponding to the sliding window is greater than the second preset probability threshold and less than the first preset probability threshold, then the drying status corresponding to the sliding window is determined to be uncertain.

[0028] If multiple consecutive sliding windows show the clothes as dried, the clothing processing equipment will stop operating; if a sliding window shows the clothes as not dried or the status is uncertain, the clothing processing equipment will continue operating.

[0029] The beneficial effects of this embodiment are as follows: by combining dual threshold comparison with continuous sliding window confirmation, instantaneous fluctuations in a single sliding window will not directly trigger a shutdown command; the shutdown judgment is based on a continuously stable drying probability. Therefore, the garment processing equipment can terminate the program when the garments have indeed reached the drying condition, and the switching of operating states is more smooth and consistent.

[0030] In one embodiment of this disclosure, it further includes:

[0031] Collect time-series operating data of the garment processing equipment under various preset loads. The time-series operating data includes the operating data of the garment processing equipment at multiple sampling times, including the relative humidity of the drum air outlet.

[0032] For each preset load, determine the target sampling time when the relative humidity at the drum outlet equals the first preset threshold under that preset load;

[0033] Remove the running data before the target sampling time from the time-series running data to obtain the target time-series running data corresponding to each preset load;

[0034] Determine the drying time corresponding to the preset standard moisture content of the clothes under each preset load;

[0035] In the target time-series running data corresponding to each preset load, the drying status of the running data before the drying time corresponding to the preset load is marked as not dried, and the drying status of the running data after the drying time corresponding to the preset load is marked as dried, thus obtaining training sample data;

[0036] The training sample data is preprocessed to obtain the target training dataset;

[0037] The first, second, and third gradient boosting tree models are trained using the target training dataset to obtain the trained first, second, and third gradient boosting tree models.

[0038] The beneficial effects of this embodiment are as follows: using the relative humidity at the drum outlet as the screening boundary, data with low correlation to drying in the early stages is first removed, and then labels are generated at the time corresponding to the standard moisture content, making the training samples closer to the state changes in the later stages of drying. The preprocessed target training dataset can reduce the interference of noisy samples on the model. The three types of gradient boosting tree models learn different discriminative features based on the same data source, thereby improving the adaptability of the subsequent model to complex load conditions and the consistency of state recognition.

[0039] In one embodiment of this disclosure, the training sample data is preprocessed to obtain the target training dataset, including:

[0040] A sliding window is used to extract features from the training sample data to obtain at least one training multidimensional feature vector corresponding to the sliding window. The training multidimensional feature vector corresponding to each sliding window is then determined as the sample data of each window.

[0041] The window sample data is subjected to label purity screening to obtain the sample dataset;

[0042] The sample dataset is cleaned to obtain the target training dataset.

[0043] The beneficial effects of this embodiment are as follows: after label purity screening and data cleaning, the label consistency and sample reliability of the target training dataset can be improved, the model training can more completely represent the critical state and fluctuating conditions, and thus improve the stability and accuracy of the subsequent judgment of the probability that the clothes are dry.

[0044] In one embodiment of this disclosure, label purity screening is performed on the window sample data to obtain a sample dataset, including:

[0045] For each window sample data, determine the sampling time included in the window sample data, and the drying status marked on the running data corresponding to each sampling time;

[0046] Based on the drying status marked by the operational data at each sampling time, the percentage marked as dried is determined.

[0047] Determine the first window of sample data whose proportion is greater than or equal to the second preset threshold;

[0048] The set of sample data from each first window is defined as the sample dataset.

[0049] The beneficial effects of this embodiment are: the sliding window level constrains the consistency of labels, so that the retained sample dataset mainly contains samples in a stable and dried state, thereby improving the purity of the training data and reducing the interference of mixed labels within the sliding window on the training data.

[0050] In one embodiment of this disclosure, the sample dataset is cleaned to obtain the target training dataset, including:

[0051] Multi-fold cross-training is performed on the first window sample data in the sample dataset to obtain the unbiased prediction probability corresponding to each first window sample data.

[0052] Determine the preset cleaning confidence threshold and the preset confidence boundary value;

[0053] The confidence interval is determined based on the preset cleaning confidence threshold and the preset confidence boundary value;

[0054] The first window of sample data whose unbiased prediction probability falls within the confidence interval is determined as the second window of sample data;

[0055] The second window of sample data is removed from the sample dataset to obtain the target training dataset.

[0056] The beneficial effects of this embodiment are as follows: Before training, the samples with blurred boundaries are identified and removed, making the sample distribution in the target training dataset more stable, reducing the interference of label noise on the model parameter learning, and making the subsequent discriminant model built based on the running data of clothing processing equipment have a more consistent training basis, thereby improving the output stability of the clothing dry fusion probability.

[0057] According to a second aspect of the present disclosure, a garment processing apparatus is provided for performing the operation control method of the garment processing apparatus as described in the first aspect.

[0058] The technical solutions provided by the embodiments of this disclosure can include the following beneficial effects: The operation control method and clothing processing device of the clothing processing equipment provided by this disclosure can collect the operation data of the clothing processing equipment and extract features from the operation data to obtain multi-dimensional feature vectors, which can characterize the dynamic changes of the operation process of the clothing processing equipment and the drying state of the clothing from multiple dimensions. By determining the fusion probability of clothing drying based on the multi-dimensional feature vectors, a more comprehensive and uncertain quantification judgment on whether the clothing is dry can be made, which can reduce the probability of misjudgment and thus improve the accuracy of judging whether the clothing is dry.

[0059] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0060] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.

[0061] Figure 1 This is an example of a scenario covered by this disclosure;

[0062] Figure 2 This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 1 ;

[0063] Figure 3 This is a flowchart illustrating the processing of multidimensional feature vectors;

[0064] Figure 4 This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 2 ;

[0065] Figure 5 This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 3 ;

[0066] Figure 6This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 4 ;

[0067] Figure 7 This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 5 ;

[0068] Figure 8 This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 6 ;

[0069] Figure 9 This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 7 ;

[0070] Figure 10 This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 8 ;

[0071] Figure 11 This is a schematic diagram of the structure of the electronic device provided in this disclosure. Detailed Implementation

[0072] Some embodiments of this disclosure will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. Various changes, modifications, and equivalents of the methods, apparatus, and / or systems described herein will become apparent upon understanding this disclosure. For example, the order of operations described herein is merely illustrative and is not limited to those orders set forth herein, but can be changed as will become apparent upon understanding this disclosure, except for operations that must be performed in a particular order. Furthermore, for clarity and brevity, descriptions of features known in the art may be omitted.

[0073] First, the application scenarios of this disclosure will be explained. This disclosure relates to the field of smart home appliance control. Figure 1 This is an example of a scenario illustration related to this disclosure, such as... Figure 1As shown, the application scenario of this disclosure specifically involves the operation control of a heat pump clothing processing equipment. Essentially, a heat pump clothing processing equipment is a closed-loop heat and moisture treatment system, mainly composed of a compressor, condenser, evaporator, throttling device, fan, and drying drum. During system operation, air circulates between the drum, condenser, and evaporator: after being heated by the condenser, the air enters the drum, where it undergoes convective heat exchange with the wet clothes and removes moisture; subsequently, the high-humidity air enters the evaporator, is cooled below the dew point, and the water vapor condenses and is discharged. The dehumidified air is then reheated and enters the next cycle. Simultaneously, the refrigerant completes heat transfer during compression, condensation, throttling, and evaporation, providing a stable heat source for the air circulation. During this process, the moisture content of the clothes continuously decreases, and the system temperature, humidity, and heat load status change dynamically accordingly. From a control perspective, the drying process of the heat pump clothing processing equipment can be divided into a heating start-up stage, a stable dehumidification stage, and a drying finish-up stage.

[0074] Automatic drying judgment primarily occurs at the end of the drying process. It typically relies on sensor data such as temperature and humidity to automatically determine the drying status of garments and accordingly control the termination of the garment processing equipment's program. However, existing drying judgment schemes often employ fixed rules or single models for assessment. When faced with varying weights, materials, and mixed loads, their ability to represent the dynamic changes in the drying process is limited, making it difficult to reliably adapt to complex operating conditions. Furthermore, the lack of effective quantification of uncertainty in the judgment process easily leads to misjudgments, affecting the accuracy of the drying judgment.

[0075] To address the aforementioned problems, this disclosure provides an operation control method and a garment processing device for garment processing equipment. By collecting operational data from the garment processing equipment and extracting features from the data to obtain multi-dimensional feature vectors, the dynamic changes in the operation process of the garment processing equipment and the drying status of the garments can be characterized from multiple dimensions. By determining the fusion probability of garment dryness based on the multi-dimensional feature vectors, a more comprehensive and uncertain quantification of whether the garments are dry can be made, reducing the probability of misjudgment and thus improving the accuracy of garment dryness assessment.

[0076] The present disclosure will now be described with reference to specific embodiments. The embodiments described in the following examples do not represent all embodiments consistent with the present disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0077] Figure 2 This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 1 ,like Figure 2 As shown, it includes the following steps.

[0078] S201. Collect operational data from the garment processing equipment;

[0079] For example, operational data refers to the data set generated by the garment processing equipment during the drying process that reflects its current operating status, and is used as the basic input for subsequent feature extraction. The operational data of the garment processing equipment can be collected by sensors in the garment processing equipment, mainly including compressor exhaust temperature, condenser temperature, drum outlet temperature, drum outlet humidity, contact humidity strip humidity, and Intelligent Power Module (IPM) temperature, etc.

[0080] S202. Extract features from the operating data of the clothing processing equipment to obtain a multi-dimensional feature vector.

[0081] For example, feature extraction is the process of converting operational data into a structured representation that can be used for subsequent processing. The multidimensional feature vector is an input composed of multiple feature dimensions, used to reflect the operating status of the clothing processing equipment and the drying status of the clothing. After obtaining the operational data of the clothing processing equipment, various features used to characterize the operating status of the clothing processing equipment and the drying status of the clothing are extracted from it, and multidimensional feature vectors are formed according to a preset method.

[0082] S203. Input the multidimensional feature vector into multiple preset sub-prediction models to obtain the sub-drying probability value output by each sub-prediction model.

[0083] For example, the probability of clothing being dry is a probability value representing the likelihood that the clothing is in a dry state under the current operating condition, and is used as the basis for subsequent operation control. Determining the probability of clothing being dry based on multi-dimensional feature vectors allows the determination of the clothing's dryness state to be comprehensively determined by combining multi-dimensional state information, thus making the determination result more consistent with the actual state of the clothing.

[0084] Optionally, multiple preset sub-prediction models include a first type gradient boosting tree model, a second type gradient boosting tree model, and a third type gradient boosting tree model; the sub-drying probability values ​​include the first probability, the second probability, and the third probability output by the first type gradient boosting tree model, the second type gradient boosting tree model, and the third type gradient boosting tree model, respectively.

[0085] For example, the first type of gradient boosting tree model can be an extreme gradient boosting model, which processes the multi-dimensional feature vector and outputs the first probability that the clothes are dry. The second type of gradient boosting tree model can be a lightweight gradient boosting model, which processes the multi-dimensional feature vector and outputs the second probability that the clothes are dry. The third type of gradient boosting tree model can be a categorical feature gradient boosting model, which processes the multi-dimensional feature vector and outputs the third probability that the clothes are dry.

[0086] S204. The multiple sub-drying probability values ​​are fused to obtain the clothing dry fusion probability, where the clothing dry fusion probability is used to characterize the possibility that the clothing has reached a dry state.

[0087] Example, Figure 3 This is a flowchart illustrating the processing of multidimensional feature vectors, such as... Figure 3 As shown, the extreme gradient boosting model, the lightweight gradient boosting model, and the category feature gradient boosting model process the multi-dimensional feature vector in parallel, outputting the first probability, the second probability, and the third probability respectively. Then, the first probability, the second probability, and the third probability are fused to obtain the final fused probability of "clothes are dry".

[0088] All three models are suitable for structured feature modeling, possess strong nonlinear fitting capabilities and robustness, and exhibit some complementarity in feature splitting mechanisms and model generalization properties, making them suitable for joint discrimination under complex working conditions. Therefore, based on the method provided in this example, by performing parallel judgments on the same multi-dimensional feature vector using multiple models and fusing the outputs, the sensitivity of individual models to local fluctuations can be balanced in the comprehensive result, thereby improving the judgment stability under critical drying scenarios.

[0089] S205. Control the operating status of the garment processing equipment based on the probability of the garments drying and merging.

[0090] For example, the operating state of the clothing processing equipment is controlled based on the clothing dryness fusion probability determined in step S204 above. When the clothing dryness fusion probability indicates that the clothing has reached a preset degree of dryness, the clothing processing equipment is controlled to stop operating; conversely, when the clothing dryness fusion probability indicates that the clothing has not reached the preset degree of dryness, the clothing processing equipment is controlled to continue operating. Using the clothing dryness fusion probability to control the operating state of the clothing processing equipment allows the equipment to perform corresponding control based on the actual dryness state of the clothing.

[0091] This example collects operational data from a garment processing device and extracts features from this data to obtain a multi-dimensional feature vector. This vector can characterize the dynamic changes in the operation of the garment processing device and the drying status of the garments from multiple dimensions. By determining the fusion probability of garment dryness based on the multi-dimensional feature vector, a more comprehensive and uncertain quantification of whether garments are dry can be made, reducing the probability of misjudgment and thus improving the accuracy of garment dryness assessment.

[0092] Optional, Figure 4 This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 2 ,like Figure 4 As shown, it also includes:

[0093] S401. Use multiple consecutive sliding windows to extract features from the operating data of the clothing processing equipment, and obtain the multi-dimensional feature vectors corresponding to each sliding window.

[0094] For example, considering that the operating data of the clothing processing equipment at only a certain moment is insufficient to reflect the dynamic changes in the drying process, this example uses a sliding window for feature construction. Preset window length, sampling step size, number of sampling points per window, and overlap rate of adjacent windows are used to extract features from the operating data. The window length, sampling step size, number of sampling points per window, and overlap rate of adjacent windows can be determined according to actual conditions. For example, the window length can be 5 minutes, the sampling step size can be 30 seconds, each window can contain 10 sampling points, and the overlap rate of adjacent windows can be set to 90%. Based on this, feature extraction is performed on the operating data to obtain multi-dimensional feature vectors corresponding to one or more sliding windows.

[0095] S402. Input the multidimensional feature vectors corresponding to each of the multiple sliding windows into multiple preset sub-prediction models to obtain the sub-drying probability values ​​corresponding to each sliding window output by each sub-prediction model.

[0096] For example, based on the multidimensional feature vectors corresponding to the multiple sliding windows obtained above, the multidimensional feature vectors corresponding to each sliding window are input into the three models mentioned above: the extreme gradient boosting model, the lightweight gradient boosting model, and the categorical feature gradient boosting model. The three models process the multidimensional feature vectors corresponding to each sliding window in parallel and output the first probability, the second probability, and the third probability corresponding to each sliding window.

[0097] S403. The sub-drying probability values ​​corresponding to each sliding window output by each sub-prediction model are fused to obtain the fused probability of clothes being dry for each sliding window.

[0098] Example, based on Figure 3 The example process involves fusing the first, second, and third probabilities corresponding to each sliding window, using sliding windows as units, to obtain the fused probability of the clothes being dry for each sliding window.

[0099] S404. Control the operating status of the clothing processing equipment according to the drying probability of each of the multiple consecutive sliding windows.

[0100] For example, based on the drying probability of each of the multiple sliding windows obtained above, the operating status of the clothing processing equipment is controlled. Assuming that the clothing processing equipment is in operation, controlling the operating status of the clothing processing equipment mainly includes controlling the clothing processing equipment to stop or controlling the clothing processing equipment to continue operating.

[0101] By using a sliding window to extract features from the operating data of the garment processing equipment, the operating data can be decomposed into local time periods and retain information on phased changes. Based on the multi-dimensional feature vectors of each sliding window, the dynamic changes of the garment machine's operating data can be accurately reflected, making the subsequent input for judging the degree of dryness of the garments more complete and thus improving the accuracy of the subsequent judgment of the degree of dryness of the garments.

[0102] Optionally, the multidimensional feature vector includes at least two of the following: time feature data, contact humidity feature data, drum outlet humidity feature data, exhaust temperature feature data, drum outlet temperature feature data, condenser feature data, heat pump system temperature difference feature data, and intelligent power module temperature feature data.

[0103] Table 1

[0104]

[0105] For example, Table 1 shows the composition of the multidimensional feature vector. As shown in Table 1, the time feature data includes the cumulative drying time, which characterizes the dehydration duration of the drying process. The contact humidity feature data includes the maximum, mean, coefficient of variation, and range of the surface humidity of the clothing, which reflects the surface humidity level and fluctuations of the clothing. The drum outlet humidity feature data includes the maximum humidity inside the drum, the relative humidity at the drum outlet, and the humidity variance, which characterizes the humidity state inside the drum. The exhaust temperature feature data includes the maximum exhaust temperature, temperature rise, range, and variance, which reflects the changes in the system's heat load. The drum outlet temperature feature data includes the maximum temperature, temperature difference, and fluctuation period, which reflects the temperature state inside the drum. The condenser feature data includes the maximum condenser temperature and the preset temperature difference efficiency index, which characterizes the heat exchange capacity of the clothing processing equipment. The heat pump system temperature difference feature data includes the average temperature difference between the condenser and the exhaust, which reflects the evaporation driving force of the clothing processing equipment. The temperature characteristic data of the intelligent power module includes data such as relative temperature rise, range, and variance, which are used to characterize the system's operating load and stability.

[0106] The multidimensional feature vector provided in this example can simultaneously reflect the linkage between temperature, humidity and component thermal state during the operation of the garment processing equipment. This allows the operational data of various aspects of the garment processing equipment to be fused in the same feature space, ensuring the integrity of the drying judgment input and making the operation control results of the garment processing equipment more consistent with the actual drying state.

[0107] Figure 5 This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 3 ,like Figure 5 As shown, S403 includes:

[0108] S501. Determine the first weight, second weight, and third weight corresponding to the first type gradient boosting tree model, the second type gradient boosting tree model, and the third type gradient boosting tree model, respectively.

[0109] For example, the first, second, and third weights can be determined by pre-setting or based on the model's performance. A more stable model corresponds to a higher weight, while a model with greater fluctuations corresponds to a lower weight. The first, second, and third weights can be denoted as ω1, ω2, and ω3, respectively; for example, they could be 0.3, 0.2, and 0.5.

[0110] S502. Based on the first weight, the second weight, and the third weight, the first probability, the second probability, and the third probability corresponding to each sliding window are weighted and summed to obtain the fusion probability of the clothes being dry for each sliding window.

[0111] For example, the first probability, the second probability, and the third probability are processed by weighted summation to obtain the probability that the clothes have dried and fused. That is, the probability that the clothes have dried and fused is obtained by the following formula.

[0112]

[0113] Where P is the probability that the clothes have dried and fused. XGB P LGB and P Cat These are the first probability, the second probability, and the third probability, respectively, and ω1, ω2, and ω3 are the first weight, the second weight, and the third weight, respectively.

[0114] This fusion method integrates the output information of three types of models within the same sliding window and distinguishes the contributions of each model through weights. This allows the final clothing dryness fusion probability to simultaneously reflect the judgment results of multiple models and the reliability of each model. Based on the method provided in this example, the clothing dryness fusion probability no longer depends on the output of a single model, but is determined jointly by multiple models according to their weights, thus reducing the impact of single-model bias on the final judgment. Since the fusion result is directly used for subsequent operational status control, this method enables the clothing processing equipment to make more stable judgments about the dryness of clothing in the later stages of drying.

[0115] Figure 6 This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 4 ,like Figure 6 As shown, the operating status of the garment processing equipment is controlled based on the drying probability of each of the multiple sliding windows, including:

[0116] S601. Determine a first preset probability threshold and a second preset probability threshold for the clothes to be dried, wherein the first preset probability threshold is greater than the second preset probability threshold.

[0117] For example, in a real-world user clothing drying control scenario, the cost of misjudgment is significantly asymmetric: Type I error (mistaking dry for wet): misjudging "dry" as "wet" leads to prolonged drying time; Type II error (misjudging wet for dry): misjudging "wet" as "dry" leads to premature shutdown. Therefore, while ensuring high recall, the "wet misjudging dry" situation should be suppressed first. To this end, this example designs a dual-threshold confidence locking mechanism. The first preset probability threshold represents the probability that the clothes are dried, and the second preset probability threshold represents the probability that the clothes are not dried. The first and second preset probability thresholds can be determined according to the actual situation and can be preset values, for example, the first preset probability threshold is 0.8 and the second preset probability threshold is 0.2.

[0118] S602. If the probability of the clothes being dried and fused corresponding to the sliding window is greater than or equal to the first preset probability threshold, then the drying state corresponding to the sliding window is determined to be that the clothes are dried; if the probability of the clothes being dried and fused corresponding to the sliding window is less than or equal to the second preset probability threshold, then the drying state corresponding to the sliding window is determined to be that the clothes are not dried; if the probability of the clothes being dried and fused corresponding to the sliding window is greater than the second preset probability threshold and less than the first preset probability threshold, then the drying state corresponding to the sliding window is determined to be uncertain.

[0119] For example, at the end of each sliding window, the corresponding dry clothing fusion probability is read, and the dry clothing fusion probability is compared with a first preset probability threshold and a second preset probability threshold. Based on the relationship between the dry clothing fusion probability and the first preset probability threshold and the second preset probability threshold, the drying conversion corresponding to each sliding window is determined.

[0120] For example, the following decision is used to determine whether clothes are dry:

[0121]

[0122] Therefore, when the probability of the clothes being dry and fused corresponding to the sliding window is greater than or equal to the first preset probability threshold, it indicates that the clothes are dry under that sliding window, and the drying state corresponding to that sliding window can be recorded as "clothes are dry." Referring to the above decision, when the probability of the clothes being dry and fused corresponding to the sliding window is less than or equal to the second preset probability threshold, it indicates that the clothes are not dry under that sliding window, and the drying state corresponding to that sliding window can be recorded as "clothes are not dry." In other cases, that is, when the probability of the clothes being dry and fused corresponding to the sliding window is greater than the second preset probability threshold but less than the first preset probability threshold, it is impossible to accurately determine whether the clothes are dry, so the drying state corresponding to the sliding window can be determined as uncertain.

[0123] S603. If multiple consecutive sliding windows correspond to the drying status of "clothes are dried", then control the clothes handling equipment to stop running; if there is a sliding window corresponding to the drying status of "clothes are not dried" or "uncertain", then control the clothes handling equipment to continue running.

[0124] For example, the drying status of multiple adjacent sliding windows is continuously judged. When multiple consecutive sliding windows meet the condition that the clothes are dry, a stop command is output to the main control board of the clothes processing equipment to cut off the operation control signals of the heat pump cycle, drum drive and related execution components, thereby stopping the whole machine.

[0125] For example, if the drying status is determined to be that the clothes are not yet dry, the drying process can continue, with continued heating, air supply, and drum rotation control to ensure that evaporated moisture is continuously condensed and discharged. If the determination is uncertain, the original operating parameters are maintained and one or more sliding window judgment cycles are extended to obtain more operational data reflecting changes in the moisture content of the clothes. This example of processing allows the clothing handling equipment to continue judging the critical state at the end of the drying process, rather than immediately ending the program. This allows subsequent operational data to be supplemented for re-identifying the drying status, thereby improving the continuity and stability of the control results.

[0126] Based on the method provided in this example, by combining dual threshold comparison with continuous sliding window confirmation, instantaneous fluctuations in a single sliding window will not directly trigger a shutdown command. The shutdown decision is based on a continuously stable drying probability. Therefore, the garment processing equipment can terminate the program when the garments have indeed reached the drying condition, and the switching of operating states is smoother and more consistent.

[0127] Optional, Figure 7 This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 5 ,like Figure 7 As shown, it also includes:

[0128] S701. Collect time-series operating data of the clothing processing equipment under various preset loads. The time-series operating data includes the operating data of the clothing processing equipment at multiple sampling times, including the relative humidity of the drum air outlet.

[0129] For example, different weights of clothing require different loads for the garment processing equipment. Preset loads can be 0.5kg, 1kg, 2kg, 3kg, 4kg, 5kg, 7kg, etc. Multiple load conditions can be pre-configured in the garment processing equipment's controller, and a standard drying program can be run under each condition. Using temperature and humidity sensors within the garment processing equipment, operational data is collected at fixed sampling intervals during the complete drying process under various preset loads, yielding the corresponding time-series operational data. The sampling interval can be determined based on actual conditions, for example, it could be 5 seconds. The operational data of the garment processing equipment includes the relative humidity at the drum outlet.

[0130] S702. For each preset load, determine the target sampling time when the relative humidity at the drum outlet equals the first preset threshold under that preset load.

[0131] For example, the first preset threshold can be determined according to the actual situation, such as 45%. During the operation of the clothing processing equipment, the clothes gradually dry, so the relative humidity at the drum air outlet will gradually decrease. The sampling time when the relative humidity at the drum air outlet is equal to 45% under various preset loads will be determined as the target sampling time under that preset load.

[0132] S703. Remove the running data before the target sampling time from the time-series running data to obtain the target time-series running data corresponding to each preset load.

[0133] For example, when the relative humidity at the drum air outlet is greater than 45%, the clothes are usually not dry. Therefore, there is no need to determine whether the clothes are dry when the relative humidity at the drum air outlet is greater than 45%. Thus, the operating data collected during the stage when the relative humidity at the drum air outlet is greater than 45% can be discarded, and the effective data that is closer to the later stage of drying can be retained to obtain the target time-series operating data.

[0134] S704. Determine the drying time corresponding to the preset standard moisture content of the clothes under each preset load.

[0135] For example, the drying time for each type of load to reach the standard moisture content can be determined using the "open-cabinet weighing method." This method involves weighing the clothes initially before washing and drying to obtain their initial weight. Then, during the operation of the garment processing equipment, the equipment can be stopped periodically, and the clothes removed and weighed to obtain the real-time weight at multiple weighing times. When the difference between the real-time weight and the initial weight is -1 ± 0.5% of the initial weight, the garment is considered to have reached the standard moisture content, and this weighing time is determined as the drying time corresponding to the garment reaching the preset standard moisture content.

[0136] S705. In the target time-series running data corresponding to each preset load, the drying status of the running data before the drying time corresponding to the preset load is marked as not dried, and the drying status of the running data after the drying time corresponding to the preset load is marked as dried, thus obtaining training sample data.

[0137] For example, if the clothes are not dry before reaching the preset standard moisture content at the designated drying time, the drying status of the data collected before this time is marked as not dry. If the clothes are dry after reaching the preset standard moisture content at the designated drying time, the drying status of the data collected after this time is marked as dry, thus obtaining the training sample data.

[0138] S706. Preprocess the training sample data to obtain the target training dataset.

[0139] For example, preprocessing the obtained training sample data can make the training data more effective.

[0140] S707. Train the preset first-class gradient boosting tree model, second-class gradient boosting tree model and third-class gradient boosting tree model using the target training dataset to obtain the trained first-class gradient boosting tree model, second-class gradient boosting tree model and third-class gradient boosting tree model.

[0141] For example, as mentioned above, the first type of gradient boosting tree model, the second type of gradient boosting tree model, and the third type of gradient boosting tree model are respectively the extreme gradient boosting model, the lightweight gradient boosting model, and the categorical feature gradient boosting model. The target training dataset is input into the extreme gradient boosting model, the lightweight gradient boosting model, and the categorical feature gradient boosting model to complete the training of the extreme gradient boosting model, the lightweight gradient boosting model, and the categorical feature gradient boosting model using the target training dataset. After training, the model can output the corresponding probability that the clothes are dry in actual judgment.

[0142] Based on the method provided in this example, the relative humidity at the drum outlet is used as the screening boundary. First, data with low correlation to drying conditions in the early stages are removed. Then, labels are generated at the corresponding time points based on standard moisture content, making the training samples more closely reflect the state changes in the later stages of drying. The preprocessed target training dataset reduces the interference of noisy samples on the model. The three types of gradient boosting tree models learn different discriminative features based on the same data source, thereby improving the adaptability of subsequent models to complex load conditions and the consistency of state recognition.

[0143] Optional, Figure 8 This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 6 ,like Figure 8 As shown, S706 includes:

[0144] S801. Use a sliding window to extract features from the training sample data to obtain at least one training multidimensional feature vector corresponding to the sliding window, and determine the training multidimensional feature vector corresponding to each sliding window as the sample data of each window.

[0145] For example, the training sample data is first arranged in chronological order, and then a sliding window is used to extract the feature data of the training sample data within the corresponding time period. Specific feature data can be found in Table 1. The sliding window for feature extraction of the training sample data can use the same sliding window parameters as described above, namely, a window length of 5 minutes, a sampling step size of 30 seconds, each window containing 10 sampling points, and an overlap rate of 90% between adjacent windows. These sliding window parameters are used to extract features from the training sample data, thereby obtaining training multidimensional feature vectors corresponding to one or more sliding windows.

[0146] S802. Perform label purity screening on the window sample data to obtain the sample dataset.

[0147] For example, for each window sample data, the label of the drying state corresponding to the running data in the corresponding sliding window is determined, the consistency of the label is determined, and the window sample data whose labels meet the preset purity threshold is retained, so as to obtain the sample dataset after label purity screening.

[0148] Optional, Figure 9 This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 7 ,like Figure 9 As shown, S802 includes:

[0149] S901. For each window sample data, determine the sampling time included in the window sample data, and the drying status marked on the running data corresponding to each sampling time.

[0150] For example, as mentioned above, the drying status of the data collected at each sampling time before the clothes reach the preset standard moisture content is marked as "not dried," and the drying status of the data collected at each sampling time after the clothes reach the preset standard moisture content is marked as "dried." Therefore, each window of sample data includes data from multiple sampling times, as well as a label for each sampling time, which includes both "dried" and "not dried."

[0151] S902. Based on the drying status marked by the operating data corresponding to each sampling time, determine the proportion marked as dried.

[0152] For example, for sample data in each window, the Label Purity Score (LPS) is determined based on the labels of the running data at each sampling time. This LPS can be the percentage of samples that are "dried". Specifically, the drying status labels corresponding to each sampling time within each sliding window can be read, the number of sampling points marked as "dried" can be determined, and the ratio of the number of dried sampling points to the total number of sampling points within the sliding window can be calculated to obtain the percentage of dried samples.

[0153] S903, determine the first window sample data whose proportion is greater than or equal to the second preset threshold.

[0154] For example, the second preset threshold can be determined according to the actual situation, such as 90%. The window sample data with a drying ratio greater than or equal to 90% is determined as the first window sample data.

[0155] S904. Combine the sample data from each first window into a sample dataset.

[0156] For example, the sample data in the first window is retained, and the sample data in other windows is removed to obtain the sample dataset.

[0157] Based on the method provided in this example, label consistency is constrained at the sliding window level, so that the retained sample dataset mainly contains samples in a stable and dried state, thereby improving the purity of the training data and reducing the interference of mixed labels within the sliding window on the training data.

[0158] S803. Perform data cleaning on the sample dataset to obtain the target training dataset.

[0159] For example, data cleaning can further remove confounding samples from the sample dataset, thereby further improving the purity of the training data.

[0160] After label purity screening and data cleaning, the label consistency and sample reliability of the target training dataset can be improved, and the model training can more completely represent the critical state and fluctuating conditions, thereby improving the stability and accuracy of subsequent judgment of the probability that clothes are dry.

[0161] Optional, Figure 10 This is a flowchart illustrating an operation control method for a garment processing device according to some embodiments of the present disclosure. Figure 8 ,like Figure 10 As shown, S803 includes:

[0162] S1001. Perform multi-fold cross-training on the first window sample data in the sample dataset to obtain the unbiased prediction probability corresponding to each first window sample data.

[0163] For example, the first window samples in the purity-filtered dataset are divided into multiple mutually exclusive subsets, such as K mutually exclusive subsets. An extreme gradient boosting model is used as a surrogate model for K-fold cross-validation: in the k-th iteration, the model is trained using K-1 subsets, and the remaining k-th subset is processed. After traversing all subsets, the unbiased prediction probability of each first window sample can be obtained.

[0164] S1002. Determine the preset cleaning confidence threshold and the preset confidence boundary value.

[0165] For example, a confidence threshold for cleaning can be defined as ϵ. ϵ and the confidence boundary value can be obtained from historical labeled data or manually set to a numerical range suitable for the current training task. The confidence range is thus determined as the range for cleaning boundary samples. For example, ϵ = 0.2, and the confidence boundary value can be a preset value, such as 0.5.

[0166] S1003. Determine the confidence interval based on the preset cleaning confidence threshold and the preset confidence boundary value.

[0167] For example, the confidence interval can be (0.5−ϵ, 0.5+ϵ), so the constructed confidence interval can be (0.3, 0.7).

[0168] S1004. The first window sample data whose unbiased prediction probability is located in the fixed confidence interval is determined as the second window sample data.

[0169] For example, the first window of sample data with unbiased prediction probabilities falling in the interval (0.3, 0.7) is determined as the second window of sample data.

[0170] S1005. Remove the second window sample data from the sample dataset to obtain the target training dataset.

[0171] For example, if the unbiased predicted probability of the first window of sample data falls within the confidence interval (0.5−ϵ, 0.5+ϵ), then the first window of sample data is considered to have low feature discrimination or contains label noise and can be removed. After removing the second window of sample data, the remaining samples retain clearer class discrimination information and can be directly used for subsequent model training.

[0172] Based on the method provided in this example, blurred boundary samples are identified and removed before training, making the sample distribution in the target training dataset more stable, reducing the interference of label noise on model parameter learning, and providing a more consistent training basis for the subsequent discriminant model built based on the running data of clothing processing equipment, thereby improving the output stability of the clothing dry fusion probability.

[0173] Optionally, this disclosure also provides a garment processing apparatus for performing the aforementioned operation control method for garment processing equipment.

[0174] The specific implementation process of the clothing processing device provided in this embodiment can be found in the above method embodiment, and its implementation principle and technical effect are similar. Therefore, it will not be repeated here.

[0175] This disclosure also provides an electronic device, Figure 11 This is a schematic diagram of the structure of the electronic device provided in this disclosure. For example... Figure 11 As shown, it includes: a processor 501 and a memory 502 communicatively connected to the processor 501; the memory 502 stores computer-executed instructions; the processor 501 executes the computer-executed instructions stored in the memory 502 to implement the operation control method of the clothing processing equipment as described in any of the first aspects above.

[0176] Optionally, the electronic device also includes a communication component 503. The processor 501, memory 502, and communication component 503 are connected via a bus. In a specific implementation, at least one processor 501 executes computer execution instructions stored in the memory 502, causing at least one processor 501 to execute the aforementioned operation control method for the clothing processing device. The specific implementation process of the processor 501 can be found in the above-described method embodiments, and its implementation principle and technical effects are similar; therefore, it will not be repeated here.

[0177] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0178] The memory may include random access memory (RAM) and non-volatile memory (NVM), such as at least one disk storage device.

[0179] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0180] This disclosure also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0181] This disclosure also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0182] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory, electrically erasable programmable read-only memory, erasable programmable read-only memory, programmable read-only memory, read-only memory, magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0183] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0184] The division of units is merely a logical functional division; 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 indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0185] 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.

[0186] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0187] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0188] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0189] Those skilled in the art will also understand that the various illustrative logical blocks and steps listed in the embodiments of this disclosure can be implemented by electronic hardware, computer software, or a combination of both. Whether such functionality is implemented in hardware or software depends on the specific application and the overall system design requirements. Those skilled in the art can implement the functionality using various methods for each specific application, but such implementation should not be construed as exceeding the scope of protection of the embodiments of this disclosure.

[0190] In the above detailed description, reference has been made to the accompanying drawings, which illustrate specific aspects of this disclosure by way of illustration. In this regard, terms indicating direction or positional relationship, such as “center,” “longitudinal,” “lateral,” “length,” “width,” “thickness,” “upper,” “lower,” “front,” “rear,” “left,” “right,” “vertical,” “horizontal,” “top,” “bottom,” “inner,” “outer,” “clockwise,” “counterclockwise,” “axial,” “radial,” and “circumferential,” are used with reference to the orientation of the described figures. Since components of the described device can be positioned in multiple different orientations, directional terms are used for illustrative purposes and not for limitation. It should be understood that other aspects can be utilized and structural or logical changes can be made without departing from the concept of this disclosure. Therefore, the following detailed description should not be considered limiting.

[0191] It should be understood that, unless otherwise specifically indicated, features of various embodiments of this disclosure described herein can be combined with each other. As used herein, the term "and / or" includes any of the associated listed items and any combination of any two or more.

[0192] It should be understood that, unless otherwise expressly specified and limited, the terms "joining," "attaching," "installing," "connecting," "linking," "fixing," etc., used in the embodiments of this disclosure should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection, an electrical connection, or a connection that allows communication between them; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components, unless otherwise expressly limited. Those skilled in the art can understand the specific meaning of the above terms herein based on the specific circumstances.

[0193] Furthermore, the term "above" as used herein with respect to components, elements, or material layers formed or located "above" a surface may be used to indicate that the component, element, or material layer is "indirectly" positioned (e.g., placed, formed, deposited, etc.) on the surface such that one or more additional components, elements, or layers are arranged between the surface and the component, element, or material layer. However, the term "above" as used with respect to components, elements, or material layers formed or located "above" a surface may also optionally have a specific meaning: that the component, element, or material layer is "directly" positioned (e.g., placed, formed, deposited, etc.) on the surface, for example, in direct contact with the surface.

[0194] Although terms such as “first,” “second,” and “third” may be used herein to describe various components, parts, regions, layers, or sections, these components, parts, regions, layers, or sections are not limited to these terms. Rather, these terms are used only to distinguish one component, part, region, layer, or section from another. Therefore, without departing from the teachings of the examples described herein, the first component, part, region, layer, or section mentioned in the examples may also be referred to as the second component, part, region, layer, or section. Furthermore, the terms “first” and “second” are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as “first” or “second” may explicitly or implicitly include at least one of that feature. In the description herein, “a plurality” means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0195] It should be understood that spatial relative terms, such as “above,” “upper,” “below,” and “lower,” are used herein to describe the relationship between one element and another shown in the figures. In addition to the orientation depicted in the figures, these spatial relative terms are also intended to encompass different orientations of the device in use or operation. For example, if the device in the figures is flipped, an element described as “above” or “upper” relative to another element would be “below” or “lower” relative to that other element. Thus, depending on the spatial orientation of the device, the term “above” encompasses both above and below orientations. Devices may have other orientations (e.g., rotated 90 degrees or in other orientations), and the spatial relative terms used herein should be interpreted accordingly.

[0196] Furthermore, the term “exemplary” is used herein to mean serving as an example, instance, or illustration. Any aspect or design described herein as “exemplary” is not necessarily to be construed as advantageous compared to other aspects or designs. Rather, the use of the term “exemplary” is intended to present the concept in a concrete manner. As used herein, the term “or” is intended to mean an inclusive “or” rather than an exclusive “or.” That is, unless otherwise specified or clear from the context, “X applies A or B” is intended to mean any of the natural inclusive arrangements. That is, “X applies A or B” satisfies any of the foregoing instances if X applies A; X applies B; or both X applies A and B. Additionally, unless otherwise specified or clear from the context to refer to the singular form, the articles “a” and “an” as used in this disclosure and the appended claims are generally understood to mean “one or more.”

[0197] Similarly, although this disclosure has been shown and described with respect to one or more implementations, equivalent variations and modifications will occur to those skilled in the art upon reading and understanding this specification and the accompanying drawings. This disclosure includes all such modifications and variations and is limited only by the scope of the claims. In particular, with respect to the various functions performed by the components described above (e.g., elements, resources, etc.), unless otherwise indicated, the terminology used to describe such components is intended to correspond to any component (functionally equivalent) that performs the specific function of the described component, even if structurally not equivalent to the disclosed structure. Furthermore, although specific features of this disclosure may have been disclosed with respect to only one of several implementations, such features may be combined with one or more other features of other implementations, as may be desired and advantageous to any given or particular application. Moreover, with regard to the terms “comprising,” “owning,” “having,” “having,” or variations thereof as used in the detailed description or claims, such terms are intended to be inclusive in a manner similar to the term “including.”

[0198] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the following claims.

[0199] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.

Claims

1. A method for controlling the operation of a garment processing device, characterized in that, include: Collect operational data from garment processing equipment; Feature extraction is performed on the operating data of the garment processing equipment to obtain a multidimensional feature vector; The multidimensional feature vector is input into multiple preset sub-prediction models to obtain the sub-drying probability value output by each sub-prediction model. Multiple sub-drying probability values ​​are fused to obtain the clothing dry fusion probability, wherein the clothing dry fusion probability is used to characterize the probability that the clothing has reached a dry state; The operating status of the clothing processing equipment is controlled based on the probability that the clothing has dried and blended.

2. The method according to claim 1, characterized in that, include: The operating data of the clothing processing equipment is extracted using multiple consecutive sliding windows to obtain multidimensional feature vectors corresponding to each of the multiple sliding windows; The multidimensional feature vectors corresponding to each of the multiple sliding windows are respectively input into multiple preset sub-prediction models to obtain the sub-drying probability values ​​corresponding to each sliding window output by each of the sub-prediction models. The sub-drying probability values ​​corresponding to each sliding window output by each of the sub-prediction models are fused to obtain the fused probability of clothes being dry corresponding to each sliding window; The operating state of the clothing processing device is controlled based on the drying probability of each of the multiple consecutive sliding windows.

3. The method according to claim 1 or 2, characterized in that, The multidimensional feature vector includes at least two of the following: time feature data, contact humidity feature data, drum outlet humidity feature data, exhaust temperature feature data, drum outlet temperature feature data, condenser feature data, heat pump system temperature difference feature data, and intelligent power module temperature feature data.

4. The method according to claim 2, characterized in that, The multiple preset sub-prediction models include a first type of gradient boosting tree model, a second type of gradient boosting tree model, and a third type of gradient boosting tree model; The sub-drying probability value includes the first probability, second probability, and third probability output by the first type of gradient boosting tree model, the second type of gradient boosting tree model, and the third type of gradient boosting tree model, respectively.

5. The method according to claim 4, characterized in that, The process of fusing the sub-drying probability values ​​corresponding to each sliding window output by each of the sub-prediction models to obtain the fused probability of clothes being dry for each sliding window includes: Determine the first weight, second weight, and third weight corresponding to the first type of gradient boosting tree model, the second type of gradient boosting tree model, and the third type of gradient boosting tree model, respectively; Based on the first weight, the second weight, and the third weight, the first probability, the second probability, and the third probability corresponding to each sliding window are weighted and summed to obtain the clothing dryness fusion probability corresponding to each sliding window.

6. The method according to claim 4, characterized in that, The step of controlling the operating state of the clothing processing equipment based on the drying and blending probabilities of the clothing corresponding to each of the multiple consecutive sliding windows includes: A first preset probability threshold and a second preset probability threshold are determined for the clothes to be dried, wherein the first preset probability threshold is greater than the second preset probability threshold; If the probability of the clothes being dried and fused corresponding to the sliding window is greater than or equal to the first preset probability threshold, then the drying status corresponding to the sliding window is determined to be that the clothes are dried; if the probability of the clothes being dried and fused corresponding to the sliding window is less than or equal to the second preset probability threshold, then the drying status corresponding to the sliding window is determined to be that the clothes are not dried; if the probability of the clothes being dried and fused corresponding to the sliding window is greater than the second preset probability threshold and less than the first preset probability threshold, then the drying status corresponding to the sliding window is determined to be uncertain. If multiple consecutive sliding windows show the drying status as "clothes are dried", the clothing processing equipment will stop operating; if any sliding window shows the drying status as "clothes are not dried" or "uncertain", the clothing processing equipment will continue operating.

7. The method according to claim 4, characterized in that, Also includes: Collect time-series operating data of the garment processing equipment under various preset loads, wherein the time-series operating data includes operating data of the garment processing equipment at multiple sampling times, and the operating data includes the relative humidity of the drum air outlet; For each preset load, determine the target sampling time when the relative humidity at the drum outlet equals the first preset threshold under that preset load; The running data before the target sampling time is removed from the time-series running data to obtain the target time-series running data corresponding to each preset load; Determine the drying time corresponding to the preset standard moisture content of the clothes under each preset load; In the target time-series running data corresponding to each preset load, the drying status of the running data before the drying time corresponding to the preset load is marked as not dried, and the drying status of the running data after the drying time corresponding to the preset load is marked as dried, so as to obtain training sample data. The training sample data is preprocessed to obtain the target training dataset; The target training dataset is used to train the preset first type gradient boosting tree model, second type gradient boosting tree model and third type gradient boosting tree model to obtain the trained first type gradient boosting tree model, second type gradient boosting tree model and third type gradient boosting tree model.

8. The method according to claim 7, characterized in that, The preprocessing of the training sample data to obtain the target training dataset includes: The training sample data is used to extract features by a sliding window to obtain at least one training multidimensional feature vector under the sliding window, and the training multidimensional feature vector under each sliding window is determined as the sample data of each window. The window sample data is subjected to label purity screening to obtain a sample dataset; The sample dataset is cleaned to obtain the target training dataset.

9. The method according to claim 8, characterized in that, The step of performing label purity filtering on the window sample data to obtain a sample dataset includes: For each window sample data, determine the sampling time included in the window sample data, and the drying status marked on the running data corresponding to each sampling time; Based on the drying status marked by the operating data corresponding to each sampling time, the percentage marked as dried is determined; Determine the first window of sample data whose proportion is greater than or equal to the second preset threshold; The set of sample data from each of the first windows is defined as the sample dataset.

10. The method according to claim 9, characterized in that, The step of cleaning the sample dataset to obtain the target training dataset includes: Multi-fold cross-training is performed on each first window sample data in the sample dataset to obtain the unbiased prediction probability corresponding to each first window sample data. Determine the preset cleaning confidence threshold and the preset confidence boundary value; Based on the preset cleaning confidence threshold and the preset confidence boundary value, a confidence interval is determined; The first window sample data whose unbiased prediction probability is located in the confidence interval is determined as the second window sample data; The second window sample data is removed from the sample dataset to obtain the target training dataset.

11. A garment processing device, characterized in that, Used to perform the method as described in any one of claims 1 to 10.

Images