Self-cleaning control method, model training method and device for air conditioner outdoor unit

CN116045446BActive Publication Date: 2024-12-03GREE ELECTRIC APPLIANCE INC OF ZHUHAI
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202211662762.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-12-23
Publication Date
2024-12-03
Estimated Expiration
2042-12-23

Smart Images

  • Figure CN116045446B_ABST
    Figure CN116045446B_ABST
Patent Text Reader

Abstract

The present invention provides a self-cleaning control method, a model training method and a device for an air-conditioning outdoor unit. The model training method comprises: obtaining reference data stored in a preset database, wherein the reference data at least comprises different dirtiness and blockage degrees of the air-conditioning outdoor unit and self-cleaning strategies corresponding to each dirtiness and blockage degree; when the air-conditioning is in a target state, obtaining a current dirtiness and blockage image of the outdoor unit; performing image analysis on the current dirtiness and blockage image of the outdoor unit to determine the corresponding current dirtiness and blockage degree, and determining the corresponding self-cleaning strategy as a target self-cleaning strategy based on the current dirtiness and blockage degree and the reference data; training a neural network model using the current dirtiness and blockage image of the outdoor unit and the target self-cleaning strategy to obtain a target neural network model. By using the model trained in this way to make self-cleaning decisions, the technical problem of poor intelligence of self-cleaning control of the air-conditioning outdoor unit in the related art can be effectively solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of air-conditioning control, and in particular to a self-cleaning control method, a model training method and a device for an outdoor unit of an air-conditioning unit. Background Art

[0002] At present, the outdoor unit of the air conditioner is installed outdoors. Due to the harsh outdoor environment, the entire outdoor unit is always exposed to the sun and rain without any protective measures. As the outdoor unit of the air conditioner is used for a long time, dust, lint, powder and other particles will continue to accumulate. The accumulation of these particles will increase the wind resistance of the air duct of the outdoor unit of the air conditioner, causing the power consumption of the outdoor fan to increase, affecting the air inlet heat exchange of the outdoor unit's condenser, resulting in a decrease in the outdoor unit's air volume, affecting the performance and service life of the outdoor unit's heat exchanger, causing abnormal thermal protection shutdown, and simultaneously reducing the service life of the compressor.

[0003] At present, the cleaning methods of outdoor heat exchangers of air conditioners mainly rely on two aspects. One is to rely on manual reminders for cleaning. Professional cleaning personnel are required to disassemble the casing of the outdoor unit and use professional tools to clean the particles in the condenser. This method can clean thoroughly, but since the outdoor unit is generally installed outdoors or on a high wall, it is difficult to disassemble. It is not only laborious, but also dangerous. The second method is to use the indoor unit to turn on the heating mode to let the outdoor unit frost, and then remove a certain amount of dust through defrosting. The self-cleaning air conditioners on the existing market require users to judge whether they need to be cleaned. This situation will lead to forgetting to clean or frequent cleaning, and the intelligence is poor. Summary of the invention

[0004] The purpose of the present invention is to overcome the above-mentioned technical deficiencies and provide a self-cleaning control method, a model training method and a device for an air-conditioning outdoor unit to solve the technical problem of poor intelligence of self-cleaning control of an air-conditioning outdoor unit in the related art.

[0005] In order to achieve the above technical objectives, the present invention adopts the following technical solutions:

[0006] According to one aspect of the present invention, there is provided a neural network model training method for self-cleaning control of an air-conditioning outdoor unit, the method comprising: obtaining reference data stored in a preset database, the reference data at least including different degrees of dirtiness and blockage of the air-conditioning outdoor unit and self-cleaning strategies corresponding to each degree of dirtiness and blockage; when the air-conditioning is in a target state, obtaining a current dirtiness and blockage image of the outdoor unit; performing image analysis on the current dirtiness and blockage image of the outdoor unit to determine the corresponding current degree of dirtiness and blockage, and determining the corresponding self-cleaning strategy as a target self-cleaning strategy based on the current degree of dirtiness and blockage and the reference data; and training a neural network model using the current dirtiness and blockage image of the outdoor unit and the target self-cleaning strategy to obtain a target neural network model.

[0007] Optionally, when the current degree of dirtiness and blockage is a specified degree of dirtiness and blockage, the method further includes: detecting changes in external unit pipe temperature parameters; and when the changes in the external unit pipe temperature parameters meet preset conditions, determining that the corresponding self-cleaning strategy is to start a self-cleaning action.

[0008] Optionally, the air conditioner is in the target state after the air conditioner is turned on or before the air conditioner is turned on.

[0009] Optionally, the reference data is obtained by: establishing a cloud database; analyzing and annotating feature information of images uploaded to the cloud database to obtain the reference data.

[0010] According to another aspect of the present invention, a self-cleaning control method for an air-conditioning outdoor unit is also provided, the method comprising: obtaining a dirty and blocked image of a target outdoor unit; inputting the dirty and blocked image of the target outdoor unit into a neural network model for self-cleaning control of the air-conditioning outdoor unit, and outputting a corresponding self-cleaning strategy, wherein the neural network model is trained according to any of the methods described above; and performing a corresponding self-cleaning operation on the air-conditioning outdoor unit according to the corresponding self-cleaning strategy.

[0011] Optionally, performing a corresponding self-cleaning operation on the air-conditioning outdoor unit according to the corresponding self-cleaning strategy includes: predicting a target time period for self-cleaning according to the corresponding self-cleaning strategy; and controlling the execution of the corresponding self-cleaning operation on the air-conditioning outdoor unit during the target time period.

[0012] According to another aspect of the present invention, there is also provided a neural network model training device for self-cleaning control of an air-conditioning outdoor unit, the device comprising: a reference data acquisition unit, for acquiring reference data stored in a preset database, the reference data at least including different degrees of dirt and blockage of the air-conditioning outdoor unit and self-cleaning strategies corresponding to each degree of dirt and blockage; a current outdoor unit dirt and blockage image acquisition unit, for acquiring a current outdoor unit dirt and blockage image when the air conditioner is in a target state; a target self-cleaning strategy determination unit, for performing image analysis on the current outdoor unit dirt and blockage image, determining the corresponding current degree of dirt and blockage, and determining the corresponding self-cleaning strategy as the target self-cleaning strategy based on the current degree of dirt and blockage and the reference data; a training unit, for training the neural network model using the current outdoor unit dirt and blockage image and the target self-cleaning strategy to obtain a target neural network model.

[0013] Optionally, when the current degree of dirtiness and blockage is a specified degree of dirtiness and blockage, the device further includes: an external unit pipe temperature parameter change detection unit, used to detect changes in the external unit pipe temperature parameters; and a self-cleaning action start control unit, used to determine the corresponding self-cleaning strategy to start a self-cleaning action when the change in the external unit pipe temperature parameters meets preset conditions.

[0014] According to another aspect of the present invention, a self-cleaning control device for an air-conditioning outdoor unit is also provided, the device comprising: a target outdoor unit dirty and blocked image acquisition unit, used to acquire a target outdoor unit dirty and blocked image; a self-cleaning strategy output unit, used to input the target outdoor unit dirty and blocked image into a neural network model for self-cleaning control of the air-conditioning outdoor unit, and output a corresponding self-cleaning strategy, wherein the neural network model is trained according to any one of the devices described above; and a self-cleaning operation execution unit, used to execute a corresponding self-cleaning operation on the air-conditioning outdoor unit according to the corresponding self-cleaning strategy.

[0015] Optionally, the self-cleaning operation execution unit is further used to: predict a target time period for self-cleaning according to the corresponding self-cleaning strategy; and control execution of a corresponding self-cleaning operation on the air conditioner outdoor unit during the target time period.

[0016] The neural network model training method for self-cleaning control of an air conditioner outdoor unit provided by the present invention obtains reference data stored in a preset database, wherein the reference data at least includes different dirty and blocked degrees of the air conditioner outdoor unit and the self-cleaning strategies corresponding to each dirty and blocked degree; when the air conditioner is in the target state, obtains the current dirty and blocked image of the outdoor unit; performs image analysis on the current dirty and blocked image of the outdoor unit to determine the corresponding current dirty and blocked degree, and determines the corresponding self-cleaning strategy as the target self-cleaning strategy based on the current dirty and blocked degree and the reference data; trains the neural network model using the current dirty and blocked image of the outdoor unit and the target self-cleaning strategy to obtain the target neural network model. Through adaptive dynamic learning, a neural network model capable of accurately predicting the self-cleaning strategy is obtained, and the self-cleaning strategy output and early warning can be realized more intelligently by using the model, effectively solving the technical problem of poor intelligence of the self-cleaning control of the air conditioner outdoor unit in the related art. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a schematic diagram of a neural network model training method for self-cleaning control of an air conditioner outdoor unit provided by an embodiment of the present invention;

[0018] Figure 2 is a schematic diagram of a self-cleaning control method for an air-conditioning outdoor unit provided by an embodiment of the present invention;

[0019] Figure 3is a schematic diagram of another self-cleaning control method for an air-conditioning outdoor unit provided by an embodiment of the present invention;

[0020] Figure 4 is a structural schematic diagram of a neural network model provided by an embodiment of the present invention;

[0021] Figure 5 is a schematic diagram of a neural network model training device for self-cleaning control of an air conditioner outdoor unit provided by an embodiment of the present invention;

[0022] Figure 6 It is a schematic diagram of a self-cleaning control device for an air-conditioning outdoor unit provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0023] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0024] Embodiment 1

[0025] According to an embodiment of the present invention, a neural network model training method for self-cleaning control of an air conditioner outdoor unit is provided. Figure 1 The method may be performed by a neural network model training device for self-cleaning control of an air conditioner outdoor unit. The method may include:

[0026] Step S101, obtaining reference data stored in a preset database, wherein the reference data at least includes different dirtiness and blockage degrees of an air-conditioning outdoor unit and a self-cleaning strategy corresponding to each dirtiness and blockage degree.

[0027] In this step, the preset database can be stored in the cloud.

[0028] In this step, the different degrees of dirt and blockage may be pre-classified dirt and blockage levels, for example, they may be divided into multiple levels according to the thickness of the accumulated dust or particles.

[0029] In this step, the self-cleaning strategy may include: whether to perform self-cleaning; and what specific means are used for self-cleaning. The specific means used for self-cleaning may be self-cleaning equipment, or manual cleaning by the user. The self-cleaning equipment may be provided on the air conditioner outdoor unit, or may be a device independent of the air conditioner outdoor unit, which is not specifically limited in this embodiment.

[0030] For example, when the dust thickness is 0 mm, the self-cleaning strategy is not to perform self-cleaning; when the dust thickness exceeds the maximum threshold, the self-cleaning strategy is manual cleaning by the user.

[0031] Step S102, when the air conditioner is in the target state, obtaining a current outdoor unit dirtiness and blockage image.

[0032] In this step, the current dirty and blocked image of the outdoor unit can be collected by the camera.

[0033] In this step, the air conditioner is in the target state, which may be after the air conditioner is turned on or before the air conditioner is turned on.

[0034] Step S103: performing image analysis on the current outdoor unit dirtiness and blockage image to determine the corresponding current dirtiness and blockage degree, and determining the corresponding self-cleaning strategy as the target self-cleaning strategy according to the current dirtiness and blockage degree and the reference data.

[0035] In this step, since the reference data stores the self-cleaning strategies corresponding to different degrees of dirt and blockage, after obtaining the current outdoor unit dirt and blockage image, a simple image analysis is first performed to determine the dirt and blockage degree, and then the corresponding self-cleaning strategy is searched in the reference data.

[0036] Step S104: training a neural network model using the current outdoor unit dirt and blockage image and the target self-cleaning strategy to obtain a target neural network model.

[0037] In this step, the neural network model is trained using the current outdoor unit dirt and blockage image and the target self-cleaning strategy. Through adaptive dynamic learning, the current dirt and blockage judgment and self-cleaning strategy are automatically saved, and the learning is continuously strengthened, so that the trained model effect becomes better and better.

[0038] Optionally, when the current degree of dirtiness and blockage is a specified degree of dirtiness and blockage, the method further includes: detecting changes in the temperature parameters of the external unit tube; and when the changes in the temperature parameters of the external unit tube meet a preset condition, determining that the corresponding self-cleaning strategy is to start a self-cleaning action. This embodiment can further improve the accuracy of the decision by detecting changes in the temperature parameters of the external unit tube to verify whether further self-cleaning of the external unit is required.

[0039] Optionally, the sensing data of a dust sensor with relatively low accuracy may be used to replace the change in the temperature parameters of the external unit pipe.

[0040] Optionally, the air conditioner is in the target state after the air conditioner is turned on or before the air conditioner is turned on. This embodiment supports execution before the air conditioner is turned on, which can effectively solve the problem that the heat exchanger of the outdoor unit is clogged with a lot of dust due to long-term non-startup, causing the whole system to be abnormal, resulting in poor heat exchange of the whole machine and poor user comfort.

[0041] Optionally, the reference data is obtained by: establishing a cloud database; analyzing and annotating feature information of images uploaded to the cloud database to obtain the reference data. In this embodiment, a database can be established in the cloud, and image analysis and annotation can be performed in the cloud, saving local storage space.

[0042] The neural network model training method for self-cleaning control of the air conditioner outdoor unit provided in this embodiment obtains reference data stored in a preset database, wherein the reference data at least includes different degrees of dirtiness and blockage of the air conditioner outdoor unit and the self-cleaning strategies corresponding to each degree of dirtiness and blockage; when the air conditioner is in the target state, obtains the current outdoor unit dirtiness and blockage image; performs image analysis on the current outdoor unit dirtiness and blockage image to determine the corresponding current degree of dirtiness and blockage, and determines the corresponding self-cleaning strategy as the target self-cleaning strategy based on the current degree of dirtiness and blockage and the reference data; trains the neural network model using the current outdoor unit dirtiness and blockage image and the target self-cleaning strategy to obtain the target neural network model. Through adaptive dynamic learning, a neural network model capable of accurately predicting the self-cleaning strategy is obtained, and the self-cleaning strategy output and early warning can be realized more intelligently using the model, effectively solving the technical problem of poor intelligence of the self-cleaning control of the air conditioner outdoor unit in the related technology.

[0043] Embodiment 2

[0044] Figure 2 is a schematic diagram of a self-cleaning control method for an air conditioner outdoor unit provided by an embodiment of the present invention, such as Figure 2 As shown, the method includes:

[0045] Step S201, obtaining a dirty and blocked image of a target outdoor unit.

[0046] In this step, a camera may be set on or near the outdoor unit to capture dirty or blocked images of the target outdoor unit.

[0047] Step S202: input the dirty and blocked image of the target outdoor unit into a neural network model for self-cleaning control of the air-conditioning outdoor unit, and output a corresponding self-cleaning strategy.

[0048] The neural network model is based on Figure 1 The method described above has been described in detail in the above content and will not be repeated here.

[0049] Step S203: performing a corresponding self-cleaning operation on the air conditioner outdoor unit according to the corresponding self-cleaning strategy.

[0050] In this step, the self-cleaning operation can be performed by a self-cleaning device or manually by a user. The self-cleaning device can be set on the air conditioner, or can be set independently of the air conditioner.

[0051] Optionally, performing a corresponding self-cleaning operation on the air conditioner outdoor unit according to the corresponding self-cleaning strategy includes: predicting a target time period for self-cleaning according to the corresponding self-cleaning strategy; and controlling the corresponding self-cleaning operation to be performed on the air conditioner outdoor unit during the target time period. This embodiment can provide accurate and effective early warning of the time for self-cleaning.

[0052] Optionally, it is possible to monitor whether the target time period is close to the current time. If it is close to the current time, a prompt message is output to the user so that the user can be informed in advance, thereby improving the user experience.

[0053] The self-cleaning control method of the air-conditioning outdoor unit provided in this embodiment obtains a dirty and blocked image of the target outdoor unit; inputs the dirty and blocked image of the target outdoor unit into a neural network model for self-cleaning control of the air-conditioning outdoor unit, and outputs a corresponding self-cleaning strategy, wherein the neural network model is trained according to the method described in any one of the above items; according to the corresponding self-cleaning strategy, a corresponding self-cleaning operation is performed on the air-conditioning outdoor unit, the degree of contamination of the air-conditioning outdoor unit can be analyzed and the corresponding self-cleaning operation can be accurately warned, the number of self-cleaning times can be optimized, and energy waste caused by forgetting to clean or frequently cleaning at scheduled times can be avoided.

[0054] Embodiment 3

[0055] Figure 3 is a schematic diagram of another self-cleaning control method for an air conditioner outdoor unit provided by an embodiment of the present invention, such as Figure 3 As shown, the method includes:

[0056] 1. Collect the image data of the contamination of the outdoor heat exchanger;

[0057] 2. Establish a cloud database;

[0058] 3. Classify the cloud data and summarize it through the image feature points;

[0059] 4. Establish data analysis on the degree of dirtiness and blockage of the two devices;

[0060] 5. Establish a scoring mechanism for the dirtiness and blockage of the two devices and a comparison mechanism for the temperature parameters of air conditioning pipes;

[0061] 6. Use video image comparison technology and convolutional neural network algorithm to compare and filter out classification levels;

[0062] 7. Determine whether the degree of dirtiness and blockage of the heat exchanger is greater than D1. If so, enter the self-cleaning mode of the outdoor unit.

[0063] Specifically, this embodiment optimizes the program of the self-cleaning start-up conditions through neural networks and deep learning, so as to set different self-cleaning parameters. The existing self-cleaning method of the external unit is a set from the inside to the outside. Under normal circumstances, we will use the naked eye to distinguish whether the external unit has reached a state of heavy dust accumulation, or when the current air conditioning cooling and heating effects are uncomfortable, we will manually perform the self-cleaning operation of the external unit. This embodiment calculates the thickness of the dust accumulation through the neural network and deep learning algorithm, and then predicts the time period when the external unit should be self-cleaned through the calculation results, thereby improving the efficiency of the external unit self-cleaning and improving the reliability of the whole machine. When detecting the thickness of the dust accumulation, it is necessary to install a camera recognition system device on the machine to detect the thickness of the dust accumulation. By adopting an intelligent control algorithm based on reinforcement learning based on the radial basis function (RBF) neural network, the deep learning algorithm is combined with real-time images for judgment, and adaptive dynamic learning is performed. Through continuous iteration, the optimal control strategy to maintain the current normal environment is found to clean the dust and keep the two devices clean.

[0064] First, after determining the control scheme of reinforcement learning, the input parameters of the neural network in the reinforcement learning algorithm are further determined. By collecting dirty pictures of the outdoor heat exchanger as the basic data source, a cloud database is created. In order to solve these problems, the network input and output parameters are determined by analyzing and sorting the collected training sample data and combining it with the working characteristics of the machine. The specific model network structure can be found in Figure 4 .

[0065] The specific scheme in the adaptive dynamic learning exploration process is as follows:

[0066] 1. During the air conditioner development phase, obtain pictures of different degrees of dirtiness and blockage of the outdoor heat exchanger, and divide the pictures into 0 to 10 levels. Please refer to the table below.

[0067] Table 1

[0068]

[0069]

[0070] 2. After the preliminary establishment of the comparison database, after the unit is running, the thickness of dust or particles can be detected in real time, matched with the relevant levels, and the execution of the external unit self-cleaning can be carried out after the comparison of the 0-10 level data. In the comparison process, the intelligent control algorithm based on reinforcement learning is based on the radial basis function (RBF) neural network, combined with the deep learning algorithm and real-time images for judgment, and adaptive dynamic learning is carried out. After a judgment is made, the current dirty and blocked pictures of the two devices are automatically saved, and iterative processing is continuously performed to find a more optimal state for dust removal.

[0071] 3. Under the comparison state of D4~D6, the change of the temperature parameter of the outdoor unit pipe can be further detected to verify whether further self-cleaning of the outdoor unit is required. 管 >53~55℃, the unit should start the external unit for self-cleaning, t 管 <50℃, the machine does not need to start self-cleaning. When 50℃≤t 管 ≤53℃, maintain the current status.

[0072] Optionally, a dust sensor with relatively low accuracy may be used to replace the parameter of the external pipe temperature which is greatly affected by the installation environment temperature for detection.

[0073] It should be noted that the calculation result does not need to be absolutely accurate, but relatively accurate. According to experience, parameters are compared item by item, and the thickness of dust accumulation and the change of related dust accumulation thickness are optimized during each power-on detection process. Compare and analyze before and after to determine the next adjustment direction and plan. This embodiment can effectively solve the problem that the heat exchanger of the outdoor unit is blocked by a large amount of dust accumulation due to long-term non-startup, resulting in abnormalities in the whole system, poor heat exchange of the whole machine, and poor user comfort.

[0074] This embodiment uses the above-mentioned technical means to realize the self-improvement and self-cleaning operating parameters of the air conditioner outdoor unit, and stores them in the database as experience. The next time the corresponding conditions are encountered, the optimal self-cleaning step is executed. At the same time, the thickness of the dirt and blockage is identified to give the prototype a self-cleaning identification signal, which can not only meet the ordinary human self-cleaning requirements, but also can be used for early warning; before the machine is turned on, after the machine was shut down for a long time, the heat exchanger of the outdoor unit has a lot of dust and dirt, and the early warning of the scheme can be used to perform the outdoor unit self-cleaning in advance. This embodiment can automatically remind the user to clean or connect the device for automatic cleaning, reducing the user's cumbersome operation actions and effectively improving the reliability of the whole system.

[0075] The above-mentioned embodiment collects data by identifying the degree of dirtiness and blockage of the unit itself through image recognition, and uses the convolutional neural network algorithm and the deep learning algorithm to warn and judge the dirtiness and blockage of the external unit condenser, optimize the number of self-cleaning times, and avoid the problem of energy waste caused by forgetting to clean or frequent cleaning at regular intervals. This embodiment can effectively warn in advance, thereby reducing the problem of increasing the wind resistance of the air duct of the air conditioner outdoor unit due to the accumulation of particulate matter, reducing the power consumption of the external fan, improving the efficiency of the air intake heat exchange of the outdoor unit condenser, increasing the air volume of the outdoor unit, improving the performance and service life of the outdoor unit heat exchanger, reducing abnormal thermal protection shutdowns, and also increasing the service life of the compressor.

[0076] Embodiment 4

[0077] Figure 5 is a schematic diagram of a neural network model training device for self-cleaning control of an air conditioner outdoor unit provided by an embodiment of the present invention, such as Figure 5 As shown, the device comprises:

[0078] A reference data acquisition unit 50 is used to acquire reference data stored in a preset database, wherein the reference data at least includes different dirtiness and blockage degrees of the air conditioner outdoor unit and the self-cleaning strategies corresponding to the dirtiness and blockage degrees;

[0079] A current outdoor unit dirty and blocked image acquisition unit 51 is used to acquire a current outdoor unit dirty and blocked image when the air conditioner is in a target state;

[0080] a target self-cleaning strategy determination unit 52, configured to perform image analysis on the current outdoor unit dirtiness and blockage image, determine the corresponding current dirtiness and blockage degree, and determine the corresponding self-cleaning strategy as the target self-cleaning strategy according to the current dirtiness and blockage degree and the reference data;

[0081] The training unit 53 is used to train the neural network model by using the current outdoor unit dirt and blockage image and the target self-cleaning strategy to obtain a target neural network model.

[0082] Optionally, when the current degree of dirtiness and blockage is a specified degree of dirtiness and blockage, the device further includes: an external unit pipe temperature parameter change detection unit, used to detect changes in the external unit pipe temperature parameters; and a self-cleaning action start control unit, used to determine the corresponding self-cleaning strategy to start a self-cleaning action when the change in the external unit pipe temperature parameters meets preset conditions.

[0083] The neural network model training device for the self-cleaning control of the air conditioner outdoor unit provided in this embodiment obtains reference data stored in a preset database, wherein the reference data at least includes different dirty and blocked degrees of the air conditioner outdoor unit and the self-cleaning strategies corresponding to each dirty and blocked degree; when the air conditioner is in the target state, obtains the current dirty and blocked image of the outdoor unit; performs image analysis on the current dirty and blocked image of the outdoor unit to determine the corresponding current dirty and blocked degree, and determines the corresponding self-cleaning strategy as the target self-cleaning strategy based on the current dirty and blocked degree and the reference data; trains the neural network model using the current dirty and blocked image of the outdoor unit and the target self-cleaning strategy to obtain the target neural network model. Through adaptive dynamic learning, a neural network model capable of accurately predicting the self-cleaning strategy is obtained, and the self-cleaning strategy output and early warning can be realized more intelligently using the model, effectively solving the technical problem of poor intelligence of the self-cleaning control of the air conditioner outdoor unit in the related technology.

[0084] Embodiment 5

[0085] Figure 6 Schematic diagram of a self-cleaning control device for an air conditioner outdoor unit provided by an embodiment of the present invention. Figure 6 As shown, the device comprises:

[0086] A target outdoor unit dirty and blocked image acquisition unit 60, used to acquire a target outdoor unit dirty and blocked image;

[0087] A self-cleaning strategy output unit 61 is used to input the dirty and blocked image of the target outdoor unit into a neural network model for self-cleaning control of the air-conditioning outdoor unit, and output a corresponding self-cleaning strategy; wherein the neural network model is trained according to any one of the above-mentioned devices, which will not be described in detail here;

[0088] The self-cleaning operation execution unit 62 is used to execute a corresponding self-cleaning operation on the air conditioner outdoor unit according to the corresponding self-cleaning strategy.

[0089] Optionally, the self-cleaning operation execution unit is further used to: predict a target time period for self-cleaning according to the corresponding self-cleaning strategy; and control execution of a corresponding self-cleaning operation on the air conditioner outdoor unit during the target time period.

[0090] The self-cleaning control device for the air-conditioning outdoor unit provided in this embodiment obtains a dirty and blocked image of a target outdoor unit; inputs the dirty and blocked image of the target outdoor unit into a neural network model for self-cleaning control of the air-conditioning outdoor unit, and outputs a corresponding self-cleaning strategy, wherein the neural network model is trained according to the method described in any one of the above items; according to the corresponding self-cleaning strategy, a corresponding self-cleaning operation is performed on the air-conditioning outdoor unit, the degree of dirtiness of the air-conditioning outdoor unit can be analyzed and an accurate warning can be given for the corresponding self-cleaning operation, thereby optimizing the number of self-cleaning times and avoiding energy waste due to forgetting to clean or frequently cleaning at scheduled times.

[0091] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0092] Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments, and this embodiment will not be described in detail here.

[0093] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0094] If the integrated units in the above embodiments are implemented in the form of software functional units and sold or used as independent products, they can be stored in the above computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for enabling one or more computer devices (which may be personal computers, servers, or network devices, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application.

[0095] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0096] In the several embodiments provided in the present application, it should be understood that the disclosed client can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0097] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0098] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0099] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A neural network model training method for self-cleaning control of an air conditioner outdoor unit, characterized in that: The method comprises: Acquire reference data stored in a preset database, wherein the reference data at least includes different dirtiness and blockage degrees of the air conditioner outdoor unit and self-cleaning strategies corresponding to the dirtiness and blockage degrees; When the air conditioner is in the target state, obtaining a current dirty and blocked image of the outdoor unit; Performing image analysis on the current outdoor unit dirtiness and blockage image to determine the corresponding current dirtiness and blockage degree, and determining the corresponding self-cleaning strategy as the target self-cleaning strategy according to the current dirtiness and blockage degree and the reference data; Using the current outdoor unit dirty and blocked image and the target self-cleaning strategy, a neural network model is trained to obtain a target neural network model; When the current dirtiness and congestion degree is a specified dirtiness and congestion degree, the method further includes: Detect changes in external unit pipe temperature parameters; When the change of the temperature parameter of the external unit tube meets the preset condition, determining the corresponding self-cleaning strategy to start the self-cleaning action; The air conditioner is in the target state after the air conditioner is turned on or before the air conditioner is turned on; The reference data is obtained by: Establish cloud database; Analyzing and annotating feature information of the image uploaded to the cloud database to obtain the reference data; The self-cleaning strategy includes: whether to perform self-cleaning; and the specific means used for self-cleaning, whether it is self-cleaning equipment that performs cleaning, or manual cleaning by the user, whether the self-cleaning equipment is set on the air conditioner outdoor unit, or whether it is a device independent of the air conditioner outdoor unit; The neural network model is trained using the current outdoor unit dirty blockage image and the target self-cleaning strategy. Through adaptive dynamic learning, the current dirty blockage judgment and self-cleaning strategy are automatically saved, and learning is continuously strengthened. According to the self-cleaning strategy, predicting a target time period for self-cleaning; and controlling the air conditioner outdoor unit to perform a corresponding self-cleaning operation during the target time period; Monitor whether the target time period is close to the current time, and if so, output a prompt message to the user so that the user can be informed in advance; Establish a dirt and blockage scoring mechanism and a comparison mechanism for air conditioning pipe temperature parameters; Video image comparison technology is used, and the convolutional neural network algorithm is used to compare and screen out the classification levels.

2. A self-cleaning control method for an air conditioner outdoor unit, characterized in that: The neural network model training method for self-cleaning control of an air conditioner outdoor unit according to claim 1 is adopted, and the method comprises: Obtain the dirty and blocked image of the target outdoor unit; Inputting the dirty and blocked image of the target outdoor unit into a neural network model for self-cleaning control of the air-conditioning outdoor unit, and outputting a corresponding self-cleaning strategy, wherein the neural network model is trained according to the method of claim 1; According to the corresponding self-cleaning strategy, a corresponding self-cleaning operation is performed on the air conditioner outdoor unit.

3. The method according to claim 2, characterized in that The performing a corresponding self-cleaning operation on the air conditioner outdoor unit according to the corresponding self-cleaning strategy includes: According to the corresponding self-cleaning strategy, predicting a target time period for self-cleaning; Controlling to perform a corresponding self-cleaning operation on the air conditioner outdoor unit during the target time period.

4. A neural network model training device for self-cleaning control of an air conditioner outdoor unit, characterized in that: The neural network model training method for self-cleaning control of an air conditioner outdoor unit according to claim 1 is adopted, and the device comprises: A reference data acquisition unit, used to acquire reference data stored in a preset database, wherein the reference data at least includes different dirtiness and blockage degrees of the air conditioner outdoor unit and self-cleaning strategies corresponding to each dirtiness and blockage degree; a current outdoor unit dirty and blocked image acquisition unit, used for acquiring a current outdoor unit dirty and blocked image when the air conditioner is in a target state; a target self-cleaning strategy determination unit, configured to perform image analysis on the current outdoor unit dirtiness and blockage image to determine the corresponding current dirtiness and blockage degree, and determine the corresponding self-cleaning strategy as the target self-cleaning strategy according to the current dirtiness and blockage degree and the reference data; The training unit is used to train the neural network model by using the current outdoor unit dirt and blockage image and the target self-cleaning strategy to obtain a target neural network model.

5. The device according to claim 4, characterized in that When the current dirtiness and blockage degree is a specified dirtiness and blockage degree, the device further includes: An external unit pipe temperature parameter change detection unit is used to detect changes in external unit pipe temperature parameters; The self-cleaning action start control unit is used to determine the corresponding self-cleaning strategy to start the self-cleaning action when the change of the external unit pipe temperature parameter meets the preset conditions.

6. A self-cleaning control device for an air conditioner outdoor unit, characterized in that: The self-cleaning control method of the air conditioner outdoor unit according to claim 2 is adopted, and the device comprises: A target outdoor unit dirty and blocked image acquisition unit, used to acquire a target outdoor unit dirty and blocked image; A self-cleaning strategy output unit, used for inputting the dirty and blocked image of the target outdoor unit into a neural network model for self-cleaning control of the air-conditioning outdoor unit, and outputting a corresponding self-cleaning strategy, wherein the neural network model is trained according to the device according to any one of claims 4-5; The self-cleaning operation execution unit is used to execute a corresponding self-cleaning operation on the air conditioner outdoor unit according to the corresponding self-cleaning strategy.

7. The device according to claim 6, characterized in that The self-cleaning operation execution unit is also used for: According to the corresponding self-cleaning strategy, predicting a target time period for self-cleaning; Controlling to perform a corresponding self-cleaning operation on the air conditioner outdoor unit during the target time period.