Washing machine eccentric anti-collision control method and washing machine
By acquiring key operating parameters of the washing machine motor and using a modified model and timing algorithm for eccentricity detection and control, the problems of low eccentricity detection accuracy and slow response speed in existing washing machines are solved. This achieves accurate detection and timely anti-collision measures, reducing vibration and collision damage to the washing machine.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-17
- Publication Date
- 2026-03-13
AI Technical Summary
Existing methods for detecting eccentricity in washing machines have low accuracy, slow response speed, and lack of self-adaptability. They cannot accurately determine the position and degree of eccentricity, leading to vibration and collision damage to the washing machine and surrounding items.
By acquiring key operating parameters of the motor after the washing machine starts, eccentricity detection and control are performed using a modified model and timing algorithm, including offline training and online real-time adjustment. Combined with cloud server data analysis, accurate detection and timely anti-collision measures are achieved.
It improves the accuracy and response speed of washing machine eccentricity detection, reduces damage to the washing machine and surrounding items caused by vibration and collision, and enhances the adaptability and reliability of the washing machine.
Smart Images
Figure CN121653931A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the technical field of textile washing equipment, specifically, it relates to a washing machine eccentricity anti-collision control method based on end-to-cloud fusion and a washing machine. Background Technology
[0002] With the improvement of people's living standards, washing machines have become one of the essential household appliances. However, the problem of eccentricity during the operation of washing machines has always been a difficult problem for both users and manufacturers. The vibrations and collisions caused by eccentricity not only affect the user experience but also increase maintenance costs and product failure rates. Therefore, how to effectively detect and solve the eccentricity problem of washing machines has become a research hotspot in the washing machine industry.
[0003] Existing washing machine eccentricity detection and anti-collision technologies primarily rely on sensor detection and simple logic control. Sensors are typically installed on the inner drum or outer casing of the washing machine to detect vibrations. When the detected vibration exceeds a certain threshold, the washing machine control system takes measures such as reducing the spin speed or pausing operation to mitigate the effects of eccentricity. For example, some washing machines use accelerometers to detect vibrations, automatically reducing the washing speed when the vibration exceeds a preset value.
[0004] The existing technology has the following main drawbacks: 1. Low detection accuracy: Existing sensor detection methods can only detect the overall vibration of the washing machine, and cannot accurately determine the specific location and degree of eccentricity. Because the vibration of a washing machine can be affected by various factors, such as the installation location and the flatness of the ground, single vibration detection cannot accurately reflect the eccentricity of the clothes. 2. Slow response speed: Anti-collision measures based on simple logic control usually only activate after the vibration exceeds a certain threshold. This means that before measures are taken, the washing machine has already experienced a period of severe vibration and collision, which may have already damaged the washing machine and surrounding items. 3. Lack of adaptability: Existing technology cannot adjust the detection and anti-collision strategies in real time according to different loads of clothes, washing modes, and the operating status of the washing machine, and cannot adapt to complex and changing actual usage situations. Summary of the Invention
[0005] In order to solve the technical problem of slow response speed in the existing washing machine eccentricity detection and adjustment, this invention proposes a washing machine eccentricity anti-collision control method, which can solve the above-mentioned problem.
[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: A method for eccentric anti-collision control of a washing machine includes: (11) Obtain the key operating parameters of the motor after the washing machine is started; (12) Detect eccentricity based on the key operating parameters. Eccentricity includes normal and abnormal. When the eccentricity is abnormal, proceed to the correction step. Input the key operating parameters into the correction model. The correction model calculates and outputs correction parameters. The correction parameters include motor power change value and / or motor power change rate. (13) Control and adjust the motor operation according to the corrected parameters.
[0007] In some embodiments, the modified model is obtained through offline training, and the training method includes: (21) Construct a corrected dataset, obtain the key operating parameters of the motor when eccentricity is known to occur, which are the original data, and obtain the initial segment data of the original data. The initial segment data is the data after the motor starts and before the eccentricity state detection is performed. (22) Generate multiple sets of correction parameters. When the same initial segment data is obtained, control the eccentricity state of the motor after it runs according to different correction parameters. (23) Take the initial segment of the original data as input, and take the correction parameter with the goal of correcting the eccentric state to normal as output, and train to obtain the correction model.
[0008] In some embodiments, the method further includes an offline training step for the detection model, wherein the training method includes: (31) Construct a detection dataset and obtain historical data of various operating parameters of the motor, including positive samples without eccentricity and negative samples with eccentricity. Obtain the initial segment data of the historical data. The initial segment is the data after the motor starts and before the eccentricity state detection is performed. The detection segment is the data within a set time after the eccentricity state detection is performed. The negative samples also include a correction segment, which is the data within a set time after the correction step is entered. (32) Using time series algorithms, the data in the detection dataset is used as input by taking N consecutive frames of data in the initial segment to predict the next frame of data, and the predicted data is compared with the actual data to adjust the model parameters and train the prediction model. (33) Find the threshold for judging the eccentric state.
[0009] In some embodiments, step (33) also includes obtaining the relevance of each operating parameter, finding the top M operating parameters in the relevance ranking, and using them as key operating parameters.
[0010] In some embodiments, in step (12), the key operating parameters are input to the detection model, the detection model outputs a predicted value for the next frame of data, and the predicted value is compared with the threshold to determine the eccentricity state of the washing machine.
[0011] In some embodiments, when the predicted values of several consecutive frames are determined to be abnormally eccentric, one of the frames is selected as the key frame for entering the correction step.
[0012] In some embodiments, the timing algorithm may employ, but is not limited to, ARIMA or LSTM algorithms.
[0013] In some embodiments, after step (13), the process further includes acquiring key operating parameters of the motor and returning to step (12).
[0014] In some embodiments, the acquired key operating parameters are uploaded to a cloud server, which then performs eccentricity detection and correction steps and sends the output correction parameters to the washing machine for control.
[0015] This invention also proposes a washing machine, including a control module configured to perform control according to the washing machine eccentric anti-collision control method described in any of the preceding claims; or, The control module communicates wirelessly with the cloud server. The control module is configured to acquire key operating parameters of the motor after the washing machine starts and send them to the cloud server. The cloud server performs eccentricity detection based on the key operating parameters. The eccentricity status includes normal and abnormal. When the eccentricity status is abnormal, it enters the correction step, inputs the key operating parameters into the correction model, and the correction model calculates and outputs correction parameters. The correction parameters include the motor power change value and / or the motor power change rate. The cloud server sends the correction parameters to the control module, which then controls and adjusts the motor operation based on the correction parameters.
[0016] Compared with the prior art, the advantages and positive effects of the present invention are as follows: The washing machine eccentricity anti-collision control method of the present invention can detect the eccentricity state based on the key operating parameters of the motor after the washing machine is started. It can issue an early warning when the washing machine is eccentric. When the eccentricity state is abnormal, it enters the correction step. The correction step can output correction parameters to solve the current eccentricity problem based on the current key operating parameters. It can quickly take corresponding anti-collision measures and reduce the damage of vibration and collision to the washing machine and surrounding items.
[0017] Other features and advantages of the present invention will become clearer after reading the detailed description of the embodiments of the present invention in conjunction with the accompanying drawings. Attached Figure Description
[0018] Figure 1 This is a flowchart of an embodiment of the washing machine eccentric anti-collision control method proposed in this invention; Figure 2This is a schematic diagram of an embodiment of the eccentric anti-collision control method for washing machines proposed in this invention. Detailed Implementation
[0019] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.
[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0021] Example 1, see Figure 1 , Figure 2 As shown, this embodiment proposes an eccentric anti-collision control method for washing machines, including: (11) Obtain the key operating parameters of the motor after the washing machine is started.
[0022] (12) Detect eccentricity based on the key operating parameters. Eccentricity includes normal and abnormal. When the eccentricity is abnormal, proceed to the correction step. Input the key operating parameters into the correction model. The correction model calculates and outputs correction parameters. The correction parameters include the motor power change value and / or the motor power change rate.
[0023] (13) Control and adjust the motor operation according to the corrected parameters.
[0024] The washing machine eccentricity anti-collision control method of this embodiment can detect the eccentricity state based on the key operating parameters of the motor after the washing machine is started. It can issue an early warning when the washing machine is eccentric. When the eccentricity state is abnormal, it enters the correction step and can quickly take corresponding anti-collision measures to reduce the damage to the washing machine and surrounding items caused by vibration and collision.
[0025] Key operating parameters include, but are not limited to, multi-dimensional data such as motor speed, vibration, temperature, and current, and are continuous in time, forming a multi-dimensional time-series data.
[0026] In some embodiments, the modified model is obtained through offline training, and the training method includes: (21) Construct a corrected dataset, obtain the key operating parameters of the motor when eccentricity is known to occur, and obtain the initial segment data of the original data. The initial segment data is the data after the motor starts and before the eccentricity state detection is performed.
[0027] (22) Generate multiple sets of correction parameters. When the same initial segment data is obtained, control the eccentricity state of the motor after running according to different correction parameters.
[0028] (23) Take the initial segment of the original data as input, and take the correction parameter with the goal of correcting the eccentric state to normal as output, and train to obtain the correction model.
[0029] This solution trains and corrects the model, adding correction parameters (such as power change value and rate of change). It can directly provide correction parameters for the eccentricity problem reflected by different key operating parameters, which helps the control module to quickly take corresponding anti-collision measures and reduce the damage of vibration and collision to the washing machine and surrounding items.
[0030] In some embodiments, when training the correction model, time-series algorithms such as ARIMA and LSTM are used. For example, using N frames of data as the main body, regression is performed on frames N+1 to N+5 to find the functional relationship between the correction parameters and the change in the result. This allows for a rapid response based on the key input operating parameters during online real-time detection, enabling timely correction and adjustment of the washing machine's operation to solve the eccentricity problem.
[0031] In step (22), multiple sets of correction parameters are generated and written into the test fixture, and the same initial segment data is input into the test fixture. When the test fixture implements the same initial segment data, the eccentricity state of the motor is detected after operation according to different correction parameters.
[0032] In some embodiments, the modified model also adds a dimension of result change values, such as vibration change values and rotational speed change values. The result change values can intuitively reflect the adjustment effect of the introduced correction parameters.
[0033] By combining key operating parameters, correction parameters, and result changes to form a correction dataset, and dividing the correction dataset into two groups—a training set and a test set—time series algorithms such as ARIMA and LSTM are used to perform regression on frames N+1 to N+5, with N frames as the main data, to find the functional relationship between the correction parameters and the result changes.
[0034] The functional relationships obtained above can be repeatedly verified and optimized in the test set to improve their accuracy.
[0035] In some embodiments, the method further includes an offline training step for the detection model, wherein the training method includes: (31) Construct a detection dataset and obtain historical data of various operating parameters of the motor, including positive samples without eccentricity and negative samples with eccentricity. Obtain the initial segment data and detection segment data of the historical data. The initial segment is the data after the motor starts and before the eccentricity state detection is performed. The detection segment is the data within a certain period of time after the eccentricity state detection is performed. The negative samples also include the correction segment, which is the data within a certain period of time after entering the correction step.
[0036] The data from the initial segment and the detection segment are aggregated into the detection algorithm's dataset to complete the creation of the detection dataset.
[0037] The detection dataset is divided into training and test sets according to a certain ratio (such as 7:3 or 8:2) for training.
[0038] (32) Using a time series algorithm, the data in the detection dataset is used as input by taking N consecutive frames of data in the initial segment to predict the next frame of data. The predicted data is compared with the actual data to adjust the model parameters and train the prediction model.
[0039] By training a prediction model, this invention can issue an early warning when the washing machine becomes eccentric and take corresponding anti-collision measures quickly, reducing the damage to the washing machine and surrounding items caused by vibration and collision.
[0040] (33) Find the threshold for judging the eccentric state.
[0041] For negative sample data, when the eccentric state occurs, it will inevitably be reflected in the operating parameters. Step (33) finds the critical value of the change in each operating parameter when the eccentricity occurs by inputting positive and negative samples, and uses it as the threshold for judging the eccentric state.
[0042] In some embodiments, step (33) also includes obtaining the relevance of each operating parameter, finding the top M operating parameters in the relevance ranking, and using them as key operating parameters.
[0043] The impact of eccentricity on various operating parameters is not fixed or uniform. Therefore, when constructing the detection dataset, it is necessary to take into account as many operating parameters as possible to comprehensively reflect the operating status and eccentricity of the washing machine, and provide rich information for accurate detection and decision-making.
[0044] This approach allows selection of one or more of the following operating parameters. During training, the correlation of each parameter is obtained, and the top M parameters by correlation are identified as key operating parameters. This saves computational resources during online real-time detection while accurately determining the eccentricity state.
[0045] The operating parameters may include, but are not limited to, the following parameters: Average rotation speed: This reflects the average level of the washing machine drum's rotation speed and is a basic indicator for judging operational stability. Abnormal values may indicate uneven load or motor failure.
[0046] Total effective value of vibration: an effective quantitative indicator that comprehensively measures the overall vibration intensity of the washing machine. It reflects the actual energy level of vibration through root mean square calculation and is used to assess the overall balance of the machine.
[0047] Average current: This monitors the average value of the motor's operating current, which is directly related to the load size and energy efficiency. Abnormal fluctuations may indicate increased mechanical resistance or electrical system problems.
[0048] Dominant frequency: The main frequency component identified from the vibration signal spectrum, which can correspond to the operating characteristics of specific mechanical parts (such as bearing frequency, unbalanced excitation frequency) and is used for fault tracing.
[0049] Rotational speed vibration correlation: This analyzes the correlation between rotational speed changes and vibration intensity. A high correlation may indicate that the vibration mainly originates from the imbalance of rotating components.
[0050] Total vibration energy: The total energy of vibration signals in all directions is calculated by integration, which comprehensively quantifies the mechanical vibration intensity of the washing machine during operation.
[0051] Load weight: The weight of clothing estimated by current or power directly affects the vibration amplitude and operational stability, and is an important background parameter for eccentricity judgment.
[0052] Comprehensive eccentricity index: A composite index constructed by combining multi-sensor data (such as vibration phase and energy distribution) to directly quantify the degree of eccentricity of clothing and provide a basis for decision-making in anti-collision control.
[0053] Average temperature: The average temperature of the motor or bearing, which indirectly reflects the mechanical friction condition. Continuous temperature rise may indicate overload or poor lubrication.
[0054] Energy per unit speed: The ratio of total vibration energy to the current speed. It is used to eliminate the influence of speed on vibration and to more purely characterize the degree of load imbalance.
[0055] Speed standard deviation: a statistical measure of the degree of speed fluctuation. The larger the value, the more unstable the speed, which may be caused by sudden load changes or abnormal control.
[0056] Vibration mean (X, Y, Z axes): This describes the average vibration level across the three axes, comprehensively reflecting the spatial distribution characteristics of vibration through vector synthesis or axial statistics.
[0057] Peak current: The maximum current value during the working cycle, which is of great significance for motor protection by capturing instantaneous overload or starting impact.
[0058] Load balance: A load distribution uniformity index calculated using initial shaking or acceleration signals. The lower the value, the more evenly the clothing is distributed.
[0059] Spectral bandwidth: The width of energy distribution around the dominant frequency, reflecting the frequency concentration of the vibration signal. Sudden changes in bandwidth may indicate loosening of components or impact.
[0060] In some embodiments, in step (12), key operating parameters are input to the detection model, the detection model outputs a predicted value for the next frame of data, the predicted value is compared with the threshold, and the eccentricity state of the washing machine is determined.
[0061] In some embodiments, when the predicted values of several consecutive frames are determined to be abnormally eccentric, one of the frames is selected as the key frame for entering the correction step.
[0062] In some embodiments, the timing algorithm may employ, but is not limited to, ARIMA or LSTM algorithms.
[0063] In some embodiments, multi-dimensional data such as the washing machine's rotation speed, vibration, temperature, and current are collected as key operating parameters of the motor.
[0064] In some embodiments, after step (13), the process further includes acquiring key operating parameters of the motor and returning to step (12). This solution continues to perform eccentricity state identification and judgment after correcting the motor control. When the eccentricity state returns to normal, the subsequent program of the washing machine continues to be executed. If it still cannot return to normal after several attempts, an alarm is triggered, and manual adjustment is required.
[0065] In some embodiments, the washing machine eccentric anti-collision control method can be deployed locally to achieve local anti-collision detection.
[0066] In some embodiments, the washing machine eccentricity anti-collision control method can also be deployed in the cloud. The key operating parameters obtained are uploaded to the cloud server, which performs eccentricity state detection and correction steps and sends the output correction parameters to the washing machine for control.
[0067] By leveraging the powerful computing and storage capabilities of cloud servers, massive amounts of historical and real-time data can be analyzed and processed, improving the accuracy of detection and the scientific basis of decision-making. Simultaneously, the computing power requirements at the washing machine end are reduced, helping to lower costs.
[0068] Machine learning models are used to train and analyze data when training the detection and correction models. This enables the model to automatically learn and adapt to different working conditions and eccentricity patterns, achieving accurate detection and real-time early warning of washing machine eccentricity.
[0069] The washing machine automatically adjusts its anti-collision decision strategy based on different operating conditions and eccentricity, enabling it to maintain stable operation under various circumstances and improving its adaptability and reliability.
[0070] By using multi-dimensional data collection and machine learning models, this invention can accurately determine the position and degree of eccentricity of the washing machine, improving the accuracy of detection and avoiding misjudgment and missed judgment.
[0071] This solution supports both self-deployment on the edge and edge-cloud integrated deployment, offering great flexibility in cost control and good adaptability.
[0072] Big data support: Cloud servers can store and analyze large amounts of historical data. By mining and analyzing this data, detection models and decision-making strategies can be continuously optimized, thereby improving the system's performance and efficiency.
[0073] Example 2: This example proposes a washing machine, including a control module, which is configured to perform control according to the washing machine eccentric anti-collision control method described in Example 1.
[0074] For specific methods of eccentric anti-collision control for washing machines, please refer to Example 1, which will not be elaborated here.
[0075] By deploying the eccentric anti-collision control method locally, the washing machine responds faster.
[0076] Example 3: This example proposes a washing machine, including a control module and a washing machine eccentric anti-collision control method deployed in the cloud. That is, the control module communicates wirelessly with the cloud server. The control module is configured to obtain key operating parameters of the motor after the washing machine starts and send them to the cloud server.
[0077] The cloud server performs eccentricity detection based on key operating parameters. Eccentricity includes normal and abnormal states. When the eccentricity is abnormal, a correction step is initiated. Key operating parameters are input into the correction model, which calculates and outputs correction parameters, including motor power change value and / or motor power change rate.
[0078] The cloud server sends the corrected parameters to the control module, which then controls and adjusts the motor operation based on the corrected parameters.
[0079] Multiple sensors, including but not limited to accelerometers, temperature sensors, and current sensors, are installed on the inner drum and outer shell of the washing machine to collect multi-dimensional data such as the washing machine's spin speed, vibration, temperature, and current in real time. This data can comprehensively reflect the washing machine's operating status and any eccentricity issues.
[0080] Other steps of the washing machine eccentric anti-collision control method can be found in Example 1, and will not be repeated here.
[0081] By deploying the eccentric anti-collision control method for washing machines in the cloud, the hardware cost of washing machines can be effectively reduced.
[0082] Of course, the above description is not intended to limit the present invention, and the present invention is not limited to the examples given above. Any changes, modifications, additions or substitutions made by those skilled in the art within the scope of the present invention should also fall within the protection scope of the present invention.
Claims
1. A method for eccentric anti-collision control in a washing machine, characterized in that, include: (11) Obtain the key operating parameters of the motor after the washing machine is started; (12) Detect eccentricity based on the key operating parameters. Eccentricity includes normal and abnormal. When the eccentricity is abnormal, proceed to the correction step. Input the key operating parameters into the correction model. The correction model calculates and outputs correction parameters. The correction parameters include motor power change value and / or motor power change rate. (13) Control and adjust the motor operation according to the corrected parameters.
2. The washing machine eccentric anti-collision control method according to claim 1, characterized in that, The modified model is obtained through offline training, and its training method includes: (21) Construct a corrected dataset, obtain the key operating parameters of the motor when eccentricity is known to occur, which are the original data, and obtain the initial segment data of the original data. The initial segment data is the data after the motor starts and before the eccentricity state detection is performed. (22) Generate multiple sets of correction parameters. When the same initial segment data is obtained, control the eccentricity state of the motor after it runs according to different correction parameters. (23) Take the initial segment of the original data as input, and take the correction parameter with the goal of correcting the eccentric state to normal as output, and train to obtain the correction model.
3. The washing machine eccentric anti-collision control method according to claim 2, characterized in that, It also includes the step of training the detection model offline, and the training methods include: (31) Construct a detection dataset and obtain historical data of various operating parameters of the motor, including positive samples without eccentricity and negative samples with eccentricity. Obtain the initial segment data and detection segment data of the historical data. The initial segment is the data after the motor starts and before the eccentricity state detection is performed, and the detection segment is the data within a certain period of time after the eccentricity state detection is performed. (32) Using time series algorithms, the data in the detection dataset is used as input by taking N consecutive frames of data in the initial segment to predict the next frame of data, and the predicted data is compared with the actual data to adjust the model parameters and train the prediction model. (33) Find the threshold for judging the eccentric state.
4. The washing machine eccentric anti-collision control method according to claim 3, characterized in that, Step (33) also includes obtaining the relevance of each operating parameter, finding the top M operating parameters in the relevance ranking, and using them as key operating parameters.
5. The washing machine eccentric anti-collision control method according to claim 3, characterized in that, In step (12), the key operating parameters are input into the detection model, and the detection model outputs a predicted value for the next frame of data. The predicted value is compared with the threshold to determine the eccentricity of the washing machine.
6. The washing machine eccentric anti-collision control method according to claim 5, characterized in that, When the predicted values of several consecutive frames are judged to be abnormally eccentric, one of the frames is selected as the key frame for entering the correction step.
7. The washing machine eccentric anti-collision control method according to claim 3, characterized in that, The timing algorithm is either ARIMA or LSTM.
8. The washing machine eccentric anti-collision control method according to any one of claims 1-7, characterized in that, After step (13), the process also includes obtaining the key operating parameters of the motor and returning to step (12).
9. The washing machine eccentric anti-collision control method according to any one of claims 1-7, characterized in that, The acquired key operating parameters are uploaded to the cloud server, which then performs eccentricity detection and correction steps and sends the output correction parameters to the washing machine for control.
10. A washing machine, characterized in that, The system includes a control module configured to perform control according to the washing machine eccentric anti-collision control method according to any one of claims 1-9. or, The control module communicates wirelessly with the cloud server. The control module is configured to acquire key operating parameters of the motor after the washing machine starts and send them to the cloud server. The cloud server performs eccentricity detection based on the key operating parameters. The eccentricity status includes normal and abnormal. When the eccentricity status is abnormal, it enters the correction step, inputs the key operating parameters into the correction model, and the correction model calculates and outputs correction parameters. The correction parameters include the motor power change value and / or the motor power change rate. The cloud server sends the correction parameters to the control module, which then controls and adjusts the motor operation based on the correction parameters.