WO2022041988A1 - 空调器的智能控制方法与智能控制设备 - Google Patents

空调器的智能控制方法与智能控制设备 Download PDF

Info

Publication number
WO2022041988A1
WO2022041988A1 PCT/CN2021/102123 CN2021102123W WO2022041988A1 WO 2022041988 A1 WO2022041988 A1 WO 2022041988A1 CN 2021102123 W CN2021102123 W CN 2021102123W WO 2022041988 A1 WO2022041988 A1 WO 2022041988A1
Authority
WO
WIPO (PCT)
Prior art keywords
air conditioner
season
startup
operating
prediction model
Prior art date
Application number
PCT/CN2021/102123
Other languages
English (en)
French (fr)
Inventor
宋世芳
郭丽
吴丽琴
Original Assignee
青岛海尔空调器有限总公司
海尔智家股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 青岛海尔空调器有限总公司, 海尔智家股份有限公司 filed Critical 青岛海尔空调器有限总公司
Publication of WO2022041988A1 publication Critical patent/WO2022041988A1/zh

Links

Images

Classifications

    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F24—HEATING; RANGES; VENTILATING
    • F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
    • F24F11/00—Control or safety arrangements
    • F24F11/62—Control or safety arrangements characterised by the type of control or by internal processing, e.g. using fuzzy logic, adaptive control or estimation of values
    • F24F11/63—Electronic processing
    • F24F11/64—Electronic processing using pre-stored data
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F24—HEATING; RANGES; VENTILATING
    • F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
    • F24F11/00—Control or safety arrangements
    • F24F11/62—Control or safety arrangements characterised by the type of control or by internal processing, e.g. using fuzzy logic, adaptive control or estimation of values
    • F24F11/63—Electronic processing
    • F24F11/65—Electronic processing for selecting an operating mode
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F24—HEATING; RANGES; VENTILATING
    • F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
    • F24F2130/00—Control inputs relating to environmental factors not covered by group F24F2110/00
    • F24F2130/10—Weather information or forecasts
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02B—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO BUILDINGS, e.g. HOUSING, HOUSE APPLIANCES OR RELATED END-USER APPLICATIONS
    • Y02B30/00—Energy efficient heating, ventilation or air conditioning [HVAC]
    • Y02B30/70—Efficient control or regulation technologies, e.g. for control of refrigerant flow, motor or heating

Definitions

  • the present invention relates to intelligent home appliance control, in particular to an intelligent control method and an intelligent control device of an air conditioner.
  • One object of the present invention is to provide an intelligent control method and an intelligent control device for an air conditioner that at least to a certain extent solve any of the above-mentioned technical problems in the related art.
  • a further object of the present invention is to make the air conditioner use the startup prediction model to accurately determine startup conditions and startup parameters.
  • Another further objective of the present invention is that the air conditioner can intelligently provide a comfortable indoor environment and improve the user experience.
  • the present invention provides an intelligent control method for an air conditioner.
  • the intelligent control method for an air conditioner includes: acquiring environmental parameters of the operating environment of the air conditioner when the air conditioner is in a standby state;
  • the prediction model uses the power-on state of the air conditioner and corresponding environmental data as training samples to be trained by a machine learning algorithm; the environmental parameters are input into the power-on prediction model; the power-on prediction model is used to perform prediction and calculation to obtain the power-on conditions and the power-on parameters of the air conditioner; and After the start-up conditions are met, the air conditioner is controlled to start with the start-up parameters.
  • the method further includes: obtaining a manual adjustment instruction of the air conditioner; using the setting parameters in the manual adjustment instruction and the environmental parameters during the manual adjustment as training samples, and predicting the startup of the air conditioner.
  • the model is trained iteratively.
  • the step of obtaining the start-up prediction model of the air conditioner includes: determining the operating season of the air conditioner; selecting a start-up prediction model corresponding to the operation season from multiple candidate prediction models, and the multiple candidate prediction models are respectively used
  • the power-on status and environmental data of the air conditioner in each operating season are used as training samples to be trained by machine learning algorithms.
  • the step of determining the operating season in which the air conditioner is located includes: acquiring installation location information of the air conditioner; determining the climate law of the area where the air conditioner is located according to the installation location information; The climate law is matched to determine the operating season of the air conditioner.
  • the step of determining the operating season of the air conditioner includes: acquiring seasonal information of the area where the air conditioner is located published by the weather platform; and determining the operating season of the air conditioner according to the seasonal information.
  • the step of controlling the air conditioner to start with the boot parameters further includes: acquiring a self-adjustment model corresponding to the operating season, and the self-adjustment model uses the operating state of the air conditioner in the operating season and the corresponding environmental data as a training sample. Obtained through machine learning algorithm training; obtaining the operating state and environmental parameters of the air conditioner; inputting the operating state and environmental parameters into the self-adjusting model; using the self-adjusting model to predict and calculate the self-adjusting strategy of the air conditioner; and adjusting the air conditioner according to the self-adjusting strategy device to adjust.
  • the method further includes: in the case of a change between the determined operating season and the operating season determined by the last operation of the air conditioner, outputting prompt information for changing the operating season; The response operation for the prompt information of running season replacement fed back by the user, in the case of confirming the running season in response to the operation instruction, execute the step of acquiring the self-adjustment model corresponding to the running season.
  • the operating season includes any one or more of the following: cooling season, heating season, rainy season, and haze removal season, wherein in the cooling season, the priority operation mode of the air conditioner is the cooling mode; in the heating season, the air conditioner The priority operation mode of the air conditioner is the heating mode; in the rainy season, the priority operation mode of the air conditioner is the dehumidification mode; in the haze removal season, the priority operation mode of the air conditioner is the purification mode.
  • the step of acquiring the startup prediction model of the air conditioner includes: acquiring the operation record of the air conditioner; judging whether the air conditioner has invoked the startup prediction model according to the operation record; if so, acquiring the startup prediction model that has been invoked before; if not, acquiring The initial start-up prediction model configured for the area where the air conditioner is located.
  • an intelligent control device for an air conditioner includes: a processor; and a memory, where a machine executable program is stored in the memory, and the machine executable program is processed.
  • the intelligent control method for realizing any one of the above-mentioned air conditioners when the air conditioner is executed.
  • the intelligent control method of the air conditioner of the present invention uses the start-up state of the air conditioner and the corresponding environmental data as training samples to obtain a start-up prediction model, and uses the start-up prediction model to predict and calculate the environmental parameters obtained in the standby state to obtain the start-up of the air conditioner. conditions and start-up parameters.
  • the machine learning algorithm fully considers various influencing factors in the startup phase of the air conditioner, which can accurately meet the user's comfort requirements for the environment and improve the user's experience.
  • the intelligent control method of the air conditioner of the present invention also considers that a single startup prediction model may not meet the startup requirements under various conditions, and also selects operating seasons with clear use characteristics, and trains separately for each operating season.
  • the start-up prediction model fully considers the environmental adjustment needs of different operating seasons, which reduces the difficulty of prediction and calculation of the start-up prediction model.
  • one or more items of the cooling season, the heating season, the rainy season and the haze removal season can be selected as the operation season according to the installation area of the air conditioner.
  • it has a priority operation mode, which greatly reduces the uncertainty of the prediction calculation of the start-up prediction model.
  • the solution of the present invention is more intelligent and efficient, and improves the level of intelligence.
  • FIG. 1 is a schematic diagram of data interaction of an air conditioner according to an embodiment of the present invention.
  • FIG. 2 is a schematic diagram of an intelligent control device for an air conditioner according to an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of an intelligent control method for an air conditioner according to an embodiment of the present invention.
  • FIG. 4 is a flowchart of an application example of an intelligent control method for an air conditioner according to an embodiment of the present invention.
  • FIG. 1 is a schematic diagram of data interaction of an air conditioner 10 according to an embodiment of the present invention.
  • An air conditioner control application (App) or other control client (Client) is installed on the terminal device 20 (including but not limited to various mobile terminals).
  • the user of the air conditioner 10 can configure the functions and application scenarios of the air conditioner 10 through an application program or a client.
  • the network data platform 30 can be used to collect and record the operation data of the air conditioner 10, collect and record user behavior information, and the like.
  • the network data platform 30 can perform machine learning model training on the operation data of the air conditioner 10 and user behavior information, and use the model obtained by training to perform prediction calculation of the air conditioner 10 .
  • the decision conditions for the prediction calculation include the indoor and outdoor environments of the air conditioner 10 (temperature, humidity, pollution, wind, weather, etc.), user information (various physiological indicators, location, clothing index, etc.), and the predicted goals include: air conditioners On/off state (including power-on parameters), operation mode (cooling, heating, purification, dehumidification, etc.) of the appliance 10, setting parameters (wind power, wind direction, temperature, humidity, etc.).
  • the air conditioner 10 obtains its own operating state and indoor and outdoor environment data, and controls the air conditioner 10 to start up according to the start-up conditions and start-up parameters predicted by the network data platform 30.
  • the self-adjustment model can be further used for prediction and calculation to obtain the air conditioner. Self-tuning strategy.
  • the network data platform 30 can also send various reminder messages to the air conditioner 10 and the terminal device 20 , and receive information replied by the user through the air conditioner 10 and the terminal device 20 .
  • the machine learning model (self-adjusting model) used in this embodiment may be able to learn certain knowledge and capabilities from existing data (the operating state of the air conditioner 10 and environmental parameters) for processing new data, and may be designed It is used to perform various tasks, and in this embodiment, it is used to determine the control strategy of the air conditioner 10 .
  • machine learning models include, but are not limited to, various types of deep neural networks (DNNs), support vector machines (SVMs), decision trees, random forest models, and the like.
  • DNNs deep neural networks
  • SVMs support vector machines
  • decision trees random forest models
  • random forest models random forest models
  • the neural network control model can adopt various known network structures suitable for supervised learning, such as perceptron model, classifier model, Hopfield network and other basic neural network structures, and various corresponding mainstream training methods can also be used for Determination of model parameters in this embodiment.
  • Example machine learning models include neural networks or other multi-layer nonlinear models.
  • Example neural networks include feedforward neural networks, deep neural networks, recurrent neural networks, and convolutional neural networks.
  • the machine learning model may be included in or otherwise stored and implemented by a server computing system (network data platform 30 ) that operates according to a client-server relationship ( or application-server) communicates with the terminal device 20 or the air conditioner 10 .
  • a machine learning model may be implemented by a server computing system (network data platform 30) as part of a web service.
  • one or more models may be stored and implemented at the end device 20 and/or one or more models may be stored and implemented at the server computing system (network data platform 30).
  • the server computing system may include or otherwise be implemented by one or more server computing devices. Where a server computing system includes multiple server computing devices, such server computing devices may operate according to a sequential computing architecture, a parallel computing architecture, or some combination thereof.
  • the end device 20 or the air conditioner 10 and/or the server computing system may train the model via interaction with a training computing system communicatively coupled through the network.
  • the training computing system may be separate from the server computing system (network data platform 30), or may be part of the server computing system (network data platform 30).
  • Interaction between the terminal device 20 or air conditioner 10 and the server computing system network may be through any type of communication network, such as a local area network (eg, an intranet), a wide area network (eg, the Internet), or some combination thereof, and may include any number of wired or wireless link.
  • communication over a network can be via any type of wired and/or wireless connection, using various communication protocols (eg, TCP/IP, HTTP, SMTP, FTP), encoding or formats (eg, HTML, XML) and/or protection schemes (eg, VPN, secure HTTP, SSL).
  • Those skilled in the art can assign data processing and computing functions to the terminal device 20, the air conditioner 10, and the network data platform 30 as required. For example, certain preprocessing is performed on the data in the terminal device 20 and the air conditioner 10 to improve the efficiency of data transmission.
  • FIG. 2 is a schematic diagram of an intelligent control device 300 of an air conditioner according to an embodiment of the present invention.
  • the intelligent control device 300 may generally include: a memory 320 and a processor 310, wherein the memory 320 stores a machine-executable program 321, and when the machine-executable program 321 is executed by the processor 310, is used to implement the functions of this embodiment.
  • An intelligent control method for an air conditioner may be a central processing unit (central processing unit, CPU for short), or a digital processing unit or the like.
  • the processor 310 transmits and receives data through the communication interface.
  • the memory 320 is used to store programs executed by the processor 310 .
  • Memory 320 is any medium that can be used to carry or store the desired program code in the form of instructions or data structures and that can be accessed by a computer, and may also be a combination of multiple memories.
  • the above-mentioned machine-executable program 321 may be downloaded from a computer-readable storage medium to a corresponding computing/processing device or downloaded and installed to the intelligent control device 300 via a network such as the Internet, a local area network, a wide area network, and/or a wireless network.
  • the intelligent control device 300 can be arranged in the above-mentioned network data platform 30 according to the above.
  • the functions of the intelligent control device 300 can also be configured, combined, and divided among the network data platform 30 , the terminal 20 , and the air conditioner 10 due to the inherent flexibility of the computer-based system.
  • FIG. 3 is a schematic diagram of an intelligent control method for an air conditioner according to an embodiment of the present invention, and the intelligent control method for the air conditioner may include:
  • Step S302 obtaining environmental parameters of the operating environment of the air conditioner when the air conditioner is in a standby state.
  • Environmental parameters may include, but are not limited to, indoor and outdoor temperature, indoor and outdoor humidity, weather, air particle data, air composition, and the like. Further, the environmental parameters may also include the physical state of the user, such as physiological index data (body temperature, heart rate, etc.), location, and the like.
  • step S304 a startup prediction model of the air conditioner is obtained, and the startup prediction model is obtained by using the startup state of the air conditioner and corresponding environmental data as training samples through machine learning algorithm training.
  • the power-on state may include the power-on mode, power-on time, power-on parameters, etc. of the air conditioner.
  • Step S306 input the environmental parameters into the startup prediction model
  • Step S308 using the startup prediction model to perform prediction and calculation to obtain startup conditions and startup parameters of the air conditioner.
  • the startup conditions may include environmental conditions, time conditions, user conditions, etc.; startup parameters include startup setting parameters of the air conditioner, initial startup parameters of each component, and the like. It should be further explained that the startup condition is not only the threshold value of each parameter, but a condition of multiple influencing factors obtained by comprehensively considering the influencing factors of each parameter.
  • step S310 after the start-up conditions are satisfied, the air conditioner is controlled to start with the start-up parameters.
  • the manual adjustment instruction of the air conditioner can also be obtained; the set parameters in the manual adjustment instruction and the environmental parameters during the manual adjustment are used as training samples to iteratively train the startup prediction model. Iterative training of the boot prediction model with manual adjustment instructions and environmental parameters during manual adjustment can better match the user's personalized needs. For example, the recent manual adjustment instructions (eg, 5 times, 10 times) and the environmental parameters during the manual adjustment period can be used as training samples.
  • the self-adjustment model of the air conditioner can also be called to obtain the operating state and environmental parameters of the air conditioner, input the operating state and environmental parameters into the self-adjustment model, and use the self-adjustment model to perform prediction and calculation to obtain the self-adjustment of the air conditioner.
  • the air conditioner is adjusted according to the self-adjustment strategy, so that the self-adjustment model can be used to continuously adjust the air conditioner in time after the environmental parameters change, so as to meet the comfort requirements of users.
  • the parameters of the air conditioner start-up stage have a great influence on the operation state of the subsequent air conditioners, and the air conditioner also needs to avoid frequent changes of the operation state as much as possible, and all the state and target parameters need to be set in the air conditioner start-up stage. That is to say, the start-up control of the air conditioner is the most important part of the air conditioner control.
  • the intelligent control method of the air conditioner of this embodiment performs model training for the operating seasons with typical usage characteristics, so that the startup prediction model corresponds to the operating seasons, and the obtained startup strategy basically conforms to the characteristics of the operating seasons , which improves the user experience.
  • the step of obtaining the start-up prediction model of the air conditioner includes: determining the operation season in which the air conditioner is located, and selecting a prediction model corresponding to the operation season from a plurality of candidate prediction models.
  • the power-on prediction model, multiple candidate prediction models are obtained by using the power-on state of the air conditioner in each operating season and the environmental data as training samples through machine learning algorithm training.
  • the above-mentioned operating season is determined according to the operating state of the air conditioner, and can generally be a time period when the air conditioner has a typical application environment.
  • the operating season includes any one or more of the following: cooling season, heating season, rainy season (or called rainy season or wet season), and haze removal season (or called purification season).
  • the climatic characteristics of the region and the needs of users for the environment configure the operating season. For example, for tropical areas, you can configure only cooling season and wet season; for areas with distinct four seasons, you can configure cooling season and heating season; for areas with monsoon climate and harsh environment, you can add haze removal season.
  • the priority operation mode of the air conditioner is the cooling mode; in the heating season, the priority operation mode of the air conditioner is the heating mode; in the rainy season, the priority operation mode of the air conditioner is the dehumidification mode; in the haze removal season , the priority operation mode of the air conditioner is the purification mode.
  • the intelligent control method for an air conditioner in this embodiment uses the power-on state of the air conditioner in each operating season and environmental data as training samples to obtain multiple candidate prediction models through machine learning algorithm training.
  • Multiple alternative prediction models correspond one-to-one with the operating season, and the selected start-up prediction model is adapted to the current operation season, thereby reducing the difficulty of the prediction and calculation of the start-up prediction model, and the obtained start-up adjustment and start-up parameters are more in line with the operating season.
  • the environment adjusts the demand, thereby improving the user experience.
  • An optional way to determine the operating season of the air conditioner is: obtain the installation location information of the air conditioner; determine the climate law of the area where the air conditioner is located according to the installation location information; The climate law is matched to determine the operating season of the air conditioner.
  • the installation location information of the air conditioner can be determined through the sales and maintenance records of the air conditioner, or through reporting by the user of the air conditioner, or through the location of the terminal bound to the air conditioner. For example, by matching the environmental data of the previous 5 to 10 days with the climate law, the operating season of the air conditioner can be determined.
  • the operating season can be determined by the climatic laws. As described above, different regions can set corresponding operating seasons according to the climate. For example, the outdoor environment data of the air conditioner within the set time (for example, 5 days to 10 days) can be matched with the above climatic laws. For example, for the North China region, if the daily average temperature is lower than 10°C for 5 consecutive days, it can be considered that heating is in use. Season; the daily average temperature for 5 consecutive days is higher than 22 °C, which can be considered as entering the cooling season.
  • Another optional way to determine the operating season of the air conditioner is to obtain the seasonal information of the area where the air conditioner is located published by the weather platform, and determine the operating season of the air conditioner according to the seasonal information. For example, for the Beijing area, mid-November to mid-March of the following year can be the heating season, and mid-June to mid-March can be the cooling season; for example, for the Shanghai area, mid-June to early July is the rainy season, and 12 The heating season can be from mid-March to the end of February.
  • the weather platform can announce the seasonal information according to the weather data, so that the operating season of the air conditioner can be determined according to the seasonal information in the announcement.
  • the manner of determining the operating season of the air conditioner is not limited to the foregoing manner, and may also be determined by manual setting in some embodiments. For example, when the season is approaching, a reminder message can be provided to the user, and the user can manually set the operating season.
  • the operation season replacement prompt information can also be output; the response operation for the operation season replacement prompt information fed back by the user is obtained, and the response operation instruction
  • the start-up prediction model can be adjusted according to the characteristics of the operating season.
  • the air conditioner start-up prediction model corresponding to the operation season preferably uses the start-up prediction model used in the same operation season of the previous year. Since the startup prediction model used in the same operation season of the previous year is generally iteratively trained using the actual operation data of the air conditioner, it is more in line with the actual needs of the user of the air conditioner.
  • the step of obtaining the start-up prediction model of the air conditioner may include: obtaining the operation record of the air conditioner, and judging whether the air conditioner has invoked the start-up prediction model according to the operation record; The initial boot prediction model configured in your region.
  • the above operation records can be used to record various operation data of the air conditioner, including but not limited to: power-on/off records, parameter adjustment records, model usage records, model training records, user manual adjustment records, environmental data, and the like.
  • the operation record of the air conditioner can determine the start-up prediction model used in the same operation season of the previous year, or whether the currently determined start-up prediction model has been invoked by the user.
  • the boot prediction model used is preferably used, which can better satisfy the user's usage habits.
  • the method further includes: acquiring the initial start-up prediction model of the operation season configured for the area where the air conditioner is located.
  • the initial start-up prediction model in the running season can be obtained by training the start-up data of the area where the air conditioner is located, and the training of using big data samples preferentially can fully reflect the climate characteristics of the region and the air conditioner usage preferences of users in the area. That is to say, if the controlled air conditioner has not called the start-up prediction model corresponding to the operating season before, using the initial start-up prediction model of the area where it is located to perform the prediction calculation of the self-adjusting model, it can be highly probable Meet the comfort requirements of most users.
  • the self-adjustment model used for self-adjustment of the air conditioner may also correspond to the operating season.
  • the step of controlling the air conditioner to start up with the start-up parameters further includes: obtaining and running the air conditioner.
  • the self-adjusting model corresponding to the season the self-adjusting model uses the operating state of the air conditioner and the corresponding environmental data in the operating season as a training sample and is trained by a machine learning algorithm to obtain the operating state and environmental parameters of the air conditioner.
  • the parameters are input into the self-adjustment model, and the self-adjustment strategy of the air conditioner is obtained by using the self-adjustment model for prediction and calculation; and the air conditioner is adjusted according to the self-adjustment strategy.
  • the above self-adjustment model is trained separately for a specific operation season, so that the self-adjustment model can also correspond to the operation season, and the obtained control strategy basically conforms to the characteristics of the operation season, which improves the user experience.
  • the self-adjustment model is mainly aimed at adjusting the state of the air conditioner according to changes in environmental parameters and user states after the air conditioner is turned on.
  • the determination method of the self-adjusting model can be the same as that of the startup prediction model. It is also preferable to select the previously used model.
  • the big data in the area where the air conditioner is located can be selected for training to obtain the initial model.
  • the power-on predictive model and the self-tuning model may be determined simultaneously.
  • FIG. 4 is a flowchart of an application example of an intelligent control method for an air conditioner according to an embodiment of the present invention. The following steps can be included in this application example:
  • Step S402 determining the operating season of the air conditioner, and the determining method may include passing weather laws, receiving broadcast messages from a climate platform, and judging time.
  • Step S404 judging whether there is a change in the operating season, that is, judging whether the determined operating season and the operating season determined by the last operation of the air conditioner have changed, so as to determine whether there is a change of seasons. For example, if the daily average temperature is higher than 22°C for 5 consecutive days in mid-June, it is determined that the cooling season is entered.
  • Step S406 after determining the change of the operating season, output the prompt information of the replacement of the operating season.
  • Step S408 it is judged whether a confirmation response to the prompt information of running season replacement feedback from the user is obtained.
  • step S410 after obtaining the confirmation response from the user, it is determined whether the air conditioner has invoked the start-up prediction model and the self-adjustment model corresponding to the operating season.
  • Step S412 if not called, obtain the initial startup prediction model and the initial self-adjustment model configured for the area where the air conditioner is located;
  • Step S414 if called, obtain the boot prediction model and the self-adjustment model corresponding to the running season that have been called before.
  • Step S416 obtain the operating state and environmental parameters of the air conditioner, input the operating state and environmental parameters into the startup prediction model, and use the startup prediction model to perform prediction and calculation to obtain the startup conditions and startup parameters of the air conditioner.
  • step S4108 after the start-up conditions are satisfied, the air conditioner is controlled to start with the start-up parameters.
  • Step S420 Obtain the environmental parameters and operating state of the air conditioner after it is turned on, input the operating state and environmental parameters into the self-adjustment model, and use the self-adjustment model to perform prediction and calculation to obtain the self-adjustment strategy of the air conditioner.
  • Step S422 adjust the air conditioner according to the self-adjustment strategy
  • Step S424 obtaining a manual adjustment instruction of the air conditioner
  • Step S426 using the manual adjustment records and the environmental parameters during the manual adjustment as training samples, iteratively trains the boot prediction model and/or the self-tuning model, wherein if the manual adjustment occurs within a set time after booting, it can be used for Iterative training is performed on the boot prediction model. If the manual adjustment occurs after the set time after booting, it can be used for iterative training of the self-tuning model.
  • the above process is only an application example, and the execution order of the steps and some steps may be added or deleted based on the introduction of the intelligent control method of the air conditioner in this embodiment.
  • the method of this embodiment may also preferably provide the user with a manual control option or configure a general-purpose model.

Landscapes

  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • Fuzzy Systems (AREA)
  • Mathematical Physics (AREA)
  • Chemical & Material Sciences (AREA)
  • Combustion & Propulsion (AREA)
  • Mechanical Engineering (AREA)
  • General Engineering & Computer Science (AREA)
  • Air Conditioning Control Device (AREA)

Abstract

一种空调器的智能控制方法与智能控制设备。其中空调器的智能控制方法包括：在空调器处于待机状态下获取空调器运行环境的环境参数(S302)；获取空调器的开机预测模型(S304)，开机预测模型利用空调器的开机状态及对应的环境数据作为训练样本通过机器学习算法训练得到；将环境参数输入开机预测模型(S306)；利用开机预测模型进行预测计算得到空调器的开机条件以及开机参数(S308)；以及在满足开机条件后，控制空调器以开机参数进行启动(S310)。

Description

空调器的智能控制方法与智能控制设备 技术领域
本发明涉及智能家电控制，特别是涉及空调器的智能控制方法与智能控制设备。
背景技术
随着生活水平的日益提高，消费者对家电的选择不再是单单注重产品的质量，而是更注重产品能够带来的使用体验。
随着人工智能、机器学习等技术的快速发展，在空调器中使用相关智能技术也逐渐成为技术研究热点。然而现有应用人工智能技术的空调器的智能控制方法的使用结果还不能完全满足用户的使用需求。由于开机启动阶段对用户的使用感受影响最大，而且空调器的开机参数直接影响了后续状态调整的次数，因此空调器的开机控制是空调器智能控制中最为重要的部分。
虽然现有技术中已经提供了许多空调器智能开机方案，但是这些方案经常受到用户的质疑，部分用户甚至反馈智能空调器的开机条件不准确，反而带来了更多困扰。
发明内容
本发明的一个目的是要提供一种至少在一定程度上解决上述相关技术中的技术问题任一方面的空调器的智能控制方法与智能控制设备。
本发明一个进一步的目的是要空调器利用开机预测模型准确地确定开机条件以及开机参数。
本发明另一个进一步的目的是要空调器可以智能地提供舒适的室内环境，提高用户的使用体验。
特别地，本发明提供了一种空调器的智能控制方法，该空调器的智能控制方法包括：在空调器处于待机状态下获取空调器运行环境的环境参数；获取空调器的开机预测模型，开机预测模型利用空调器的开机状态及对应的环境数据作为训练样本通过机器学习算法训练得到；将环境参数输入开机预测模型；利用开机预测模型进行预测计算得到空调器的开机条件以及开机参数；以及在满足开机条件后，控制空调器以开机参数进行启动。
可选地，在控制空调器以开机参数进行启动的步骤之后还包括：获取空 调器的手动调整指令；将手动调整指令中的设定参数以及手动调整期间的环境参数作为训练样本，对开机预测模型进行迭代训练。
可选地，获取空调器的开机预测模型的步骤包括：确定空调器所处的运行季；从多个备选预测模型中选取与运行季对应的开机预测模型，多个备选预测模型分别利用空调器在各个运行季中的开机状态以及环境数据作为训练样本通过机器学习算法训练得到。
可选地，确定空调器所处的运行季的步骤包括：获取空调器的安装位置信息；根据安装位置信息确定空调器所在区域的气候规律；将运行环境此前设定时间段内的环境数据与气候规律进行匹配，以确定得出空调器所处的运行季。
可选地，确定空调器所处的运行季的步骤包括：获取由天气平台发布的空调器所在区域的季节信息；根据季节信息确定空调器所处的运行季。
可选地，在控制空调器以开机参数进行启动的步骤之后还包括：获取与运行季对应的自调整模型，自调整模型利用运行季中的空调器的运行状态及对应的环境数据作为训练样本通过机器学习算法训练得到；获取空调器的运行状态以及环境参数；将运行状态和环境参数输入自调整模型；利用自调整模型进行预测计算得到空调器的自调整策略；以及按照自调整策略对空调器进行调整。
可选地，在确定空调器所处的运行季的步骤之后还包括：在确定出的运行季与空调器上次运行确定的运行季出现变化的情况下，输出运行季更替提示信息；获取由用户反馈的运行季更替提示信息的响应操作，在响应操作指示确认运行季的情况下，执行获取与运行季对应的自调整模型的步骤。
可选地，运行季包括以下任意一项或多项：制冷季节、采暖季节、梅雨季节、除霾季节，其中在制冷季节中，空调器的优先运行模式为制冷模式；在采暖季节中，空调器的优先运行模式为制热模式；在梅雨季节中，空调器的优先运行模式为除湿模式；在除霾季节中，空调器的优先运行模式为净化模式。
可选地，获取空调器的开机预测模型的步骤包括：获取空调器的运行记录；根据运行记录判断空调器是否调用过开机预测模型；若是，获取之前调用过的开机预测模型；若否，获取为空调器所在区域配置的初始开机预测模型。
根据本发明的另一个方面，还提供了一种空调器的智能控制设备，该空调器的智能控制设备包括：处理器；以及存储器，存储器内存储有机器可执行程序，机器可执行程序被处理器执行时用于实现上述任一种的空调器的智能控制方法。
本发明的空调器的智能控制方法，利用空调器的开机状态及对应的环境数据作为训练样本训练得到开机预测模型，通过开机预测模型对待机状态下获取的环境参数进行预测计算得到空调器的开机条件以及开机参数。机器学习算法充分考虑了空调器开机阶段的各种影响因素，可以准确地满足用户对环境的舒适性要求，提高了用户的使用体验。
进一步地，本发明的空调器的智能控制方法，还考虑到单一的开机预测模型可能无法满足各种条件下的开机需求，还选取具有明确使用特征的运行季，针对每种运行季分别训练得到开机预测模型，充分考虑了不同运行季的环境调节需求，降低了开机预测模型的预测计算难度。
更进一步地，本发明的空调器的智能控制方法，还可以根据空调器的安装区域选择制冷季节、采暖季节、梅雨季节、除霾季节中的一项或多项作为运行季，由于这些运行季一般具有优先运行模式，大大降低了开机预测模型的预测计算的不确定性。相对于智能家电(智慧家电)及智能空调(智慧空调)等领域中的现有技术，本发明的方案更加智能高效，提高了智能化水平。
根据下文结合附图对本发明具体实施例的详细描述，本领域技术人员将会更加明了本发明的上述以及其他目的、优点和特征。
附图说明
后文将参照附图以示例性而非限制性的方式详细描述本发明的一些具体实施例。附图中相同的附图标记标示了相同或类似的部件或部分。本领域技术人员应该理解，这些附图未必是按比例绘制的。附图中：
图1是根据本发明一个实施例的空调器的数据交互示意图；
图2是根据本发明一个实施例的空调器的智能控制设备的示意图；
图3是根据本发明一个实施例的空调器的智能控制方法的示意图；以及
图4是根据本发明一个实施例的空调器的智能控制方法应用实例的流程图。
具体实施方式
图1是根据本发明一个实施例的空调器10的数据交互示意图。终端设备20(包括但不限于各种移动终端)上安装有空调器控制应用程序(App)或者其他控制客户端(Client)。空调器10的用户可以通过应用程序或者客户端来配置空调器10的功能以及应用场景。
网络数据平台30可以用于采集并记录空调器10运行数据、采集并记录用户行为信息等。网络数据平台30可以通过对空调器10运行数据以及用户行为信息进行机器学习模型的训练，并利用训练得到的模型进行空调器10的预测计算。其中预测计算的决策条件包括空调器10的室内外环境(温度、湿度、污染情况、风力、天气等)、用户信息(各项生理指标、位置、穿衣指数等)，预测的目标包括：空调器10的开关状态(包括开机参数)、运行模式(制冷、制热、净化、除湿等)、设定参数(风力、风向、温度、湿度等)。
空调器10获取自身运行状态以及室内外环境数据，并按照网络数据平台30预测得到的开机条件以及开机参数对空调器10进行开机控制，此外还可以进一步利用自调整模型进行预测计算得到空调器的自调整策略。
另外，网络数据平台30还可以向空调器10和终端设备20发送各种提醒消息，并接收用户通过空调器10以及终端设备20回复的信息。
本实施例中使用的机器学习模型(自调整模型)可以是能够从已有数据(空调器10的运行状态以及环境参数)中学习到一定的知识和能力用于处理新数据，并可以被设计用于执行各种任务，在本实施例中用于对空调器10控制策略的确定。机器学习模型的示例包括但不限于各类深度神经网络(DNN)、支持向量机(SVM)、决策树、随机森林模型等等。在实施例中，机器学习模型也可以被称为“学习网络”。其中神经网络控制模型可以采用各种已知的适合有监督学习的网络结构，例如感知器模型，分类器模型，Hopfield网络等基本的神经网络结构，各种相应的主流训练方法也都可以用于本实施例的模型参数的确定。示例机器学习模型包括神经网络或其他多层非线性模型。示例神经网络包括前馈神经网络、深度神经网络、递归神经网络和卷积神经网络。
机器学习模型可以包括在服务器计算系统(网络数据平台30)中或以其他方式由服务器计算系统(网络数据平台30)存储和实现，服务器计算系统(网络数据平台30)根据客户端-服务器关系(或者应用程序-服务器)与终 端设备20或者空调器10通信。例如，机器学习模型可以由服务器计算系统(网络数据平台30)实现为web服务的一部分。因此，可以在终端设备20处存储和实现一个或多个模型和/或可以在服务器计算系统(网络数据平台30)处存储和实现一个或多个模型。
服务器计算系统(网络数据平台30)可以包括一个或多个服务器计算设备或以其他方式由一个或多个服务器计算设备实现。在服务器计算系统包括多个服务器计算设备的情况下，这样的服务器计算设备可以根据顺序计算架构、并行计算架构或其一些组合来操作。
终端设备20或者空调器10和/或服务器计算系统(网络数据平台30)可以经由与通过网络通信地耦接的训练计算系统的交互来训练模型。训练计算系统可以与服务器计算系统(网络数据平台30)分离，或者可以是服务器计算系统(网络数据平台30)的一部分。
终端设备20或者空调器10和服务器计算系统网络之间可以通过任何类型的通信网络进行交互，诸如局域网(例如内联网)、广域网(例如因特网)或其一些组合，并且可以包括任何数量的有线或无线链路。通常，通过网络的通信可以经由任何类型的有线和/或无线连接，使用各种通信协议(例如，TCP/IP、HTTP、SMTP、FTP)、编码或格式(例如、HTML、XML)和/或保护方案(例如、VPN、安全HTTP、SSL)来承载。
本领域技术人员可以根据需要在终端设备20、空调器10、网络数据平台30分配数据的处理和运算功能。例如在终端设备20和空调器10中对数据进行一定的预处理，以提高数据传输的效率。
图2是根据本发明一个实施例的空调器的智能控制设备300的示意图。该智能控制设备300可以包括一般性地可以包括：存储器320以及处理器310，其中存储器320内存储有机器可执行程序321，机器可执行程序321被处理器310执行时用于实现本实施例的空调器的智能控制方法。处理器310可以是一个中央处理单元(central processing unit，简称CPU)，或者为数字处理单元等等。处理器310通过通信接口收发数据。存储器320用于存储处理器310执行的程序。存储器320是能够用于携带或存储具有指令或数据结构形式的期望的程序代码并能够由计算机存取的任何介质，也可以是多个存储器的组合。上述机器可执行程序321可以从计算机可读存储介质下载到相应计算/处理设备或者经由网络(例如因特网、局域网、广域网和/或无线网 络)下载并安装到智能控制设备300。
智能控制设备300可根据布置于上述网络数据平台30中。此外，智能控制设备300的功能也可以基于计算机的系统的固有灵活性允许在网络数据平台30、终端20、空调器10之间进行配置、组合以及划分。
图3是根据本发明一个实施例的空调器的智能控制方法的示意图，该空调器的智能控制方法可以包括：
步骤S302，在空调器处于待机状态下获取空调器运行环境的环境参数。环境参数可以包括但不限于：室内外温度、室内外湿度、天气、空气颗粒数据、空气成分等。进一步地，环境参数还可以包括用户的身体状态，例如生理指标数据(体温、心率等)、所在位置等。
步骤S304，获取空调器的开机预测模型，开机预测模型利用空调器的开机状态及对应的环境数据作为训练样本通过机器学习算法训练得到。开机状态可以包括空调器的开机模式、开机时间、开机参数等。
步骤S306，将环境参数输入开机预测模型；
步骤S308，利用开机预测模型进行预测计算得到空调器的开机条件以及开机参数。开机条件可以包括环境条件、时间条件、用户条件等；开机参数包括空调器的开机设定参数、各部件的开机初始参数等。需要进一步说明的是，开机条件并非仅仅为各项参数的阈值，而是综合考虑各项参数的影响因素得到的多影响因子的条件。
步骤S310，在满足开机条件后，控制空调器以开机参数进行启动。在空调器启动后，还可以获取空调器的手动调整指令；将手动调整指令中的设定参数以及手动调整期间的环境参数作为训练样本，对开机预测模型进行迭代训练。通过利用手动调整指令以及手动调整期间的环境参数对开机预测模型进行迭代训练可以更好地匹配用户的个性化需求。例如可以将最近若干次(例如5次、10次)的手动调整指令以及手动调整期间的环境参数作为训练样本。
在开启空调器之后，还可以调用空调器的自调整模型，获取空调器的运行状态以及环境参数，将运行状态和环境参数输入自调整模型，利用自调整模型进行预测计算得到空调器的自调整策略，以及按照自调整策略对空调器进行调整，从而利用自调整模型在环境参数出现变化后及时对空调器进行持续调整，满足用户的舒适性要求。
空调器开启阶段的参数对于后续空调器的运行状态影响较大，并且空调器也要求尽量避免频繁变更运行状态，而且空调器启动阶段需要对全部状态及目标参数进行设定。也就是说空调器的开机控制是空调器控制中最为重要的部分。
现有智能空调器的可实现的功能越来越多，仅仅以送风为例，除了风力、风向之外，还增加了自然风、循环风、新风、无风感等多种送风模式，而针对不同季节，用户的舒适度感受也存在差别。例如对于春季和秋季，虽然温度相差不大，但是用户的需求确明显不同。不同季节中出现同一环境数据的情况下，某些季节可能需要开启空调器，而某些季节可能并不需要开启空调器。使用同一开机预测模型进行预测计算，显然无法满足用户的要求。基于这一问题，本实施例的空调器的智能控制方法对于具有典型使用特征的运行季分别进行模型训练，从而使得开机预测模型与运行季相对应，得出的开机策略基本符合运行季的特点，提高了用户的使用体验。
在使用与运行季相对应的开机预测模型的实施例中，获取空调器的开机预测模型的步骤包括：确定空调器所处的运行季，从多个备选预测模型中选取与运行季对应的开机预测模型，多个备选预测模型分别利用空调器在各个运行季中的开机状态以及环境数据作为训练样本通过机器学习算法训练得到。
上述运行季根据空调器的运行状态进行确定，一般可为空调器具有典型应用环境的时间段。例如运行季包括以下任意一项或多项：制冷季节、采暖季节、梅雨季节(或称为雨季或潮湿季节)、除霾季节(或称为净化季)，本领域技术人员可以根据空调器所在区域的气候特点以及用户对环境的需求，配置运行季。例如对于热带区域，可以仅配置制冷季节以及潮湿季节；对于四季分明的区域，可以配置制冷季节、采暖季节；对于具有季风气候且环境较为恶劣的区域，可以增加设置除霾季节。
在制冷季节中，空调器的优先运行模式为制冷模式；在采暖季节中，空调器的优先运行模式为制热模式；在梅雨季节中，空调器的优先运行模式为除湿模式；在除霾季节中，空调器的优先运行模式为净化模式。
本实施例的空调器的智能控制方法，分别利用空调器在各个运行季中的开机状态以及环境数据作为训练样本通过机器学习算法训练得到多个备选预测模型。多个备选预测模型与运行季一一对应，选择出的开机预测模型与 当前的运行季相适配，从而减小了开机预测模型的预测计算难度，得到的开机调节和开机参数更加符合运行季的环境调节需求，从而提高了用户的使用体验。
确定空调器所处的运行季的一种可选方式为：获取空调器的安装位置信息；根据安装位置信息确定空调器所在区域的气候规律；将运行环境此前设定时间段内的环境数据与气候规律进行匹配，以确定得出空调器所处的运行季。空调器的安装位置信息可以通过空调器的销售及维护记录确定，也可以通过空调器的用户上报确定，还可以通过与空调器绑定的终端的位置确定。例如运行环境此前5天至10天的环境数据与气候规律进行匹配，可以确定得出空调器所处的运行季。
由于不同的区域，其气候规律存在较大的差别，可以通过气候规律来确定运行季，如上文所介绍的，不同的区域可以根据气候设置相应运行季。例如可以将空调器设定时间内(例如5天至10天)的室外环境数据与上述气候规律进行匹配，例如对于中国华北区域，连续5天日平均气温低于10℃，即可认为进入采暖季节；连续5天日平均气温高于22℃，即可认为进入制冷季节。
确定空调器所处的运行季的另一种可选方式为：获取由天气平台发布的空调器所在区域的季节信息，根据季节信息确定空调器所处的运行季。例如对于北京区域，11月中旬至次年3月中旬可以为采暖季节，而6月中旬至8月中旬可以为制冷季节；又例如对于上海区域，6月中旬至7月上旬是梅雨季节，12月中旬至2月底可以为采暖季节。天气平台可以根据气象数据对季节信息进行通告，从而根据通告中的季节信息可以确定空调器所处的运行季。
确定空调器所处的运行季的方式并不局限于上述方式，在一些实施例中还可以通过人工设定方式来确定。例如在临近换季时间，可以向用户提供提醒消息，由用户人工设定运行季。
在确定出的运行季与空调器上次运行确定的运行季出现变化的情况下，还可以输出运行季更替提示信息；获取由用户反馈的针对运行季更替提示信息的响应操作，在响应操作指示确认运行季的情况下，确定运行季更换。也就是说在出现换季的情况下，可以向用户提醒，在得到用户的确认后，按照该运行季的特点调整开机预测模型。
与运行季对应的空调器开机预测模型优选使用在上一年度的相同的运行季中使用过的开机预测模型。由于上一年度的相同的运行季中使用过的开机预测模型一般利用空调器的实际运行数据进行了迭代训练，更加符合该空调器的用户的实际需求。获取空调器的开机预测模型的步骤可以包括：获取空调器的运行记录，根据运行记录判断空调器是否调用过开机预测模型，若是，获取之前调用过的开机预测模型，若否，获取为空调器所在区域配置的初始开机预测模型。
上述运行记录可以用于记录空调器的各项运行数据，包括但不限于：开关机记录、参数调整记录、模型使用记录、模型训练记录、用户手动调整记录、环境数据等。空调器的运行记录可以确定出上一年度的相同的运行季中使用过的开机预测模型，或者当前确定出的开机预测模型是否被该用户调用过。优选采用使用的开机预测模型，可以更加满足用户的使用习惯。
在空调器未调用过与运行季对应的开机预测模型的情况下还包括：获取为空调器所在区域配置的运行季初始开机预测模型。运行季初始开机预测模型可以利用空调器所在区域的开机数据进行训练得到，优先利用大数据样本训练充分可以反映地域的气候特点以及该区域内用户的空调器使用偏好。也就是说，在被控的空调器之前未调用过与运行季对应的开机预测模型的情况下，使用其所在区域的初始开机预测模型进行自调整模型的预测计算，其在很大概率上可以满足大多数用户的舒适性要求。
进一步地，在空调器开机运行后，用于对空调器进行自调整的自调整模型同样也可以与运行季相对应，例如在控制空调器以开机参数进行启动的步骤之后还包括：获取与运行季对应的自调整模型，自调整模型利用运行季中的空调器的运行状态及对应的环境数据作为训练样本通过机器学习算法训练得到，获取空调器的运行状态以及环境参数，将运行状态和环境参数输入自调整模型，利用自调整模型进行预测计算得到空调器的自调整策略；以及按照自调整策略对空调器进行调整。
上述自调整模型针对特定运行季分别进行模型训练，从而也可以使得自调整模型与运行季相对应，得出的控制策略基本符合运行季的特点，提高了用户的使用体验。自调整模型主要针对于空调器开启后根据环境参数以及用户状态的变化对空调器的状态进行调整。自调整模型的确定方式与开机预测模型的确定方式可以相同，同样优选选择此前使用的模型，其次可以选择空 调器所在区域的大数据进行训练得到初始模型。在一些实施例中，可以同时确定开机预测模型以及自调整模型。
图4是根据本发明一个实施例的空调器的智能控制方法应用实例的流程图。在该应用实例中可以包括以下步骤：
步骤S402，确定空调器所处的运行季，确定方式可以包括通过气候规律、接收气候平台的广播消息、时间判断等。
步骤S404，判断运行季是否出现变化，也即判断确定出的运行季与空调器上次运行确定的运行季是否出现变化，从而判断是否出现换季的情况。例如若在6月中旬出现连续5天日平均气温高于22℃的情况，则判断进入制冷季节。
步骤S406，在确定运行季变化后，输出运行季更替提示信息。
步骤S408，判断是否获取到用户反馈的针对运行季更替提示信息的确认响应。
步骤S410，在得到用户的确认相应后，判断空调器是否调用过与运行季对应的开机预测模型以及自调整模型。
步骤S412，若未调用过，获取为空调器所在区域配置的初始开机预测模型以及初始自调整模型；
步骤S414，若调用过，获取之前调用过的与运行季对应的开机预测模型以及自调整模型。
步骤S416，获取空调器的运行状态以及环境参数，将运行状态和环境参数输入开机预测模型，利用开机预测模型进行预测计算得到空调器的开机条件以及开机参数。
步骤S418，在满足开机条件后，控制空调器以开机参数进行启动。
步骤S420，获取空调器开启后的环境参数以及运行状态，将运行状态和环境参数输入自调整模型，利用自调整模型进行预测计算得到空调器的自调整策略。
步骤S422，按照自调整策略对空调器进行调整；
步骤S424，获取空调器的手动调整指令；
步骤S426，将手动调整记录以及手动调整期间的环境参数作为训练样本，对开机预测模型和/或自调整模型进行迭代训练，其中如果手动调整发生于开机后的设定时间内，则可以用于对开机预测模型进行迭代训练，若手动 调整发生于开机后设定时间后的阶段，则可以用于对自调整模型进行迭代训练。
本领域技术人员应该了解上述流程仅为一个应用实例，可以在本实施例对空调器的智能控制方法的介绍的基础上调整步骤的执行顺序以及增删部分步骤。对于空调器并不处于某一具有典型特点的运行季中或者不存在运行季或对应的开机预测模型和/或自调整模型的情况下(例如天气变化较为频繁且随机或者空调器开机率较小的情况)，本实施例的方法还可以优选为用户提供手动控制选项或者配置通用型的模型。
至此，本领域技术人员应认识到，虽然本文已详尽示出和描述了本发明的多个示例性实施例，但是，在不脱离本发明精神和范围的情况下，仍可根据本发明公开的内容直接确定或推导出符合本发明原理的许多其他变型或修改。因此，本发明的范围应被理解和认定为覆盖了所有这些其他变型或修改。

Claims (10)

  1. 一种空调器的智能控制方法，包括：
    在所述空调器处于待机状态下获取所述空调器运行环境的环境参数；
    获取所述空调器的开机预测模型，所述开机预测模型利用空调器的开机状态及对应的环境数据作为训练样本通过机器学习算法训练得到；
    将所述环境参数输入所述开机预测模型；
    利用所述开机预测模型进行预测计算得到所述空调器的开机条件以及开机参数；以及
    在满足所述开机条件后，控制所述空调器以所述开机参数进行启动。
  2. 根据权利要求1所述的空调器的智能控制方法，其中，在控制所述空调器以所述开机参数进行启动的步骤之后还包括：
    获取所述空调器的手动调整指令；
    将所述手动调整指令中的设定参数以及手动调整期间的环境参数作为训练样本，对所述开机预测模型进行迭代训练。
  3. 根据权利要求1所述的空调器的智能控制方法，其中，所述获取所述空调器的开机预测模型的步骤包括：
    确定所述空调器所处的运行季；
    从多个备选预测模型中选取与所述运行季对应的开机预测模型，多个所述备选预测模型分别利用空调器在各个运行季中的开机状态以及环境数据作为训练样本通过机器学习算法训练得到。
  4. 根据权利要求3所述的空调器的智能控制方法，其中，所述确定所述空调器所处的运行季的步骤包括：
    获取所述空调器的安装位置信息；
    根据所述安装位置信息确定所述空调器所在区域的气候规律；
    将所述运行环境此前设定时间段内的环境数据与所述气候规律进行匹配，以确定得出所述空调器所处的运行季。
  5. 根据权利要求3所述的空调器的智能控制方法，其中所述确定所述空 调器所处的运行季的步骤包括：
    获取由天气平台发布的所述空调器所在区域的季节信息；
    根据所述季节信息确定所述空调器所处的运行季。
  6. 根据权利要求3所述的空调器的智能控制方法，其中，在控制所述空调器以所述开机参数进行启动的步骤之后还包括：
    获取与所述运行季对应的自调整模型，所述自调整模型利用所述运行季中的空调器的运行状态及对应的环境数据作为训练样本通过机器学习算法训练得到；
    获取所述空调器的运行状态以及环境参数；
    将所述运行状态和所述环境参数输入所述自调整模型；
    利用所述自调整模型进行预测计算得到所述空调器的自调整策略；以及
    按照所述自调整策略对所述空调器进行调整。
  7. 根据权利要求3所述的空调器的智能控制方法，其中，在确定所述空调器所处的运行季的步骤之后还包括：
    在确定出的所述运行季与所述空调器上次运行确定的运行季出现变化的情况下，输出运行季更替提示信息；
    获取由用户反馈的所述运行季更替提示信息的响应操作，在所述响应操作指示确认所述运行季的情况下，执行获取与所述运行季对应的自调整模型的步骤。
  8. 根据权利要求3所述的空调器的智能控制方法，其中，
    所述运行季包括以下任意一项或多项：制冷季节、采暖季节、梅雨季节、除霾季节，其中
    在所述制冷季节中，所述空调器的优先运行模式为制冷模式；
    在所述采暖季节中，所述空调器的优先运行模式为制热模式；
    在所述梅雨季节中，所述空调器的优先运行模式为除湿模式；
    在所述除霾季节中，所述空调器的优先运行模式为净化模式。
  9. 根据权利要求1所述的空调器的智能控制方法，其中，所述获取所述空调器的开机预测模型的步骤包括：
    获取所述空调器的运行记录；
    根据所述运行记录判断所述空调器是否调用过开机预测模型；
    若是，获取之前调用过的所述开机预测模型；
    若否，获取为所述空调器所在区域配置的初始开机预测模型。
  10. 一种空调器的智能控制设备，包括：
    处理器；以及
    存储器，所述存储器内存储有机器可执行程序，所述机器可执行程序被所述处理器执行时用于实现根据权利要求1至9中任一项所述的空调器的智能控制方法。
PCT/CN2021/102123 2020-08-28 2021-06-24 空调器的智能控制方法与智能控制设备 WO2022041988A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010889187.0A CN112128936B (zh) 2020-08-28 2020-08-28 空调器的智能控制方法与智能控制设备
CN202010889187.0 2020-08-28

Publications (1)

Publication Number Publication Date
WO2022041988A1 true WO2022041988A1 (zh) 2022-03-03

Family

ID=73847587

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/102123 WO2022041988A1 (zh) 2020-08-28 2021-06-24 空调器的智能控制方法与智能控制设备

Country Status (2)

Country Link
CN (1) CN112128936B (zh)
WO (1) WO2022041988A1 (zh)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114992814A (zh) * 2022-06-20 2022-09-02 南京天加环境科技有限公司 一种空调舒适性的控制方法、系统及空调
CN115065709A (zh) * 2022-03-10 2022-09-16 青岛海尔空调器有限总公司 一种信息处理方法、装置及系统
CN115200155A (zh) * 2022-07-26 2022-10-18 珠海格力电器股份有限公司 空调的控制方法、装置、空调及存储介质
CN117216505A (zh) * 2023-11-09 2023-12-12 广州视声智能股份有限公司 基于智慧家居使用记录的用户习惯预测方法及系统
CN117308288A (zh) * 2023-09-21 2023-12-29 中国工业互联网研究院 一种制丝空调启停预测控制方法以及装置

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112128936B (zh) * 2020-08-28 2022-08-19 青岛海尔空调器有限总公司 空调器的智能控制方法与智能控制设备
CN112665139A (zh) * 2021-01-19 2021-04-16 青岛海尔空调器有限总公司 空调器的控制方法、终端设备、服务器、空调器控制系统
US20240119163A1 (en) * 2021-02-10 2024-04-11 Digital Arts Inc. Information processing system, information processing method, and information processing program
CN113847681B (zh) * 2021-08-31 2023-04-28 青岛海尔空调电子有限公司 空调器及其控制方法
CN113790503A (zh) * 2021-11-11 2021-12-14 成都联帮医疗科技股份有限公司 一种智能调节供氧状态的弥散供氧终端机及其控制方法
CN114090091A (zh) * 2021-11-26 2022-02-25 联影(常州)医疗科技有限公司 设备自动开机方法、装置、电子设备及存储介质

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2010133589A (ja) * 2008-12-03 2010-06-17 Panasonic Corp 空気調和機
JP2010151337A (ja) * 2008-12-24 2010-07-08 Daikin Ind Ltd 空調システム
CN106547897A (zh) * 2016-10-31 2017-03-29 广东美的制冷设备有限公司 基于地理位置匹配空调器的方法及装置
CN108361927A (zh) * 2018-02-08 2018-08-03 广东美的暖通设备有限公司 一种基于机器学习的空调器控制方法、装置以及空调器
CN110030699A (zh) * 2019-04-04 2019-07-19 珠海格力电器股份有限公司 一种空调设备控制方法、空调和存储介质
CN110454930A (zh) * 2018-05-08 2019-11-15 中国科学院理化技术研究所 一种基于人体最佳热舒适度估算的空调控制方法与装置
CN112128936A (zh) * 2020-08-28 2020-12-25 青岛海尔空调器有限总公司 空调器的智能控制方法与智能控制设备

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108278737B (zh) * 2017-12-20 2020-12-25 珠海格力电器股份有限公司 一种空调的控制方法、装置、存储介质、空调及遥控器

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2010133589A (ja) * 2008-12-03 2010-06-17 Panasonic Corp 空気調和機
JP2010151337A (ja) * 2008-12-24 2010-07-08 Daikin Ind Ltd 空調システム
CN106547897A (zh) * 2016-10-31 2017-03-29 广东美的制冷设备有限公司 基于地理位置匹配空调器的方法及装置
CN108361927A (zh) * 2018-02-08 2018-08-03 广东美的暖通设备有限公司 一种基于机器学习的空调器控制方法、装置以及空调器
CN110454930A (zh) * 2018-05-08 2019-11-15 中国科学院理化技术研究所 一种基于人体最佳热舒适度估算的空调控制方法与装置
CN110030699A (zh) * 2019-04-04 2019-07-19 珠海格力电器股份有限公司 一种空调设备控制方法、空调和存储介质
CN112128936A (zh) * 2020-08-28 2020-12-25 青岛海尔空调器有限总公司 空调器的智能控制方法与智能控制设备

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115065709A (zh) * 2022-03-10 2022-09-16 青岛海尔空调器有限总公司 一种信息处理方法、装置及系统
CN114992814A (zh) * 2022-06-20 2022-09-02 南京天加环境科技有限公司 一种空调舒适性的控制方法、系统及空调
CN114992814B (zh) * 2022-06-20 2023-12-15 南京天加环境科技有限公司 一种空调舒适性的控制方法、系统及空调
CN115200155A (zh) * 2022-07-26 2022-10-18 珠海格力电器股份有限公司 空调的控制方法、装置、空调及存储介质
CN115200155B (zh) * 2022-07-26 2023-11-14 珠海格力电器股份有限公司 空调的控制方法、装置、空调及存储介质
CN117308288A (zh) * 2023-09-21 2023-12-29 中国工业互联网研究院 一种制丝空调启停预测控制方法以及装置
CN117308288B (zh) * 2023-09-21 2024-04-19 中国工业互联网研究院 一种制丝空调启停预测控制方法以及装置
CN117216505A (zh) * 2023-11-09 2023-12-12 广州视声智能股份有限公司 基于智慧家居使用记录的用户习惯预测方法及系统
CN117216505B (zh) * 2023-11-09 2024-03-19 广州视声智能股份有限公司 基于智慧家居使用记录的用户习惯预测方法及系统

Also Published As

Publication number Publication date
CN112128936B (zh) 2022-08-19
CN112128936A (zh) 2020-12-25

Similar Documents

Publication Publication Date Title
WO2022041988A1 (zh) 空调器的智能控制方法与智能控制设备
WO2022041987A1 (zh) 空调器的智能控制方法与智能控制设备
KR102553299B1 (ko) 데이터 학습 서버 및 이의 학습 모델 생성 및 이용 방법
US11137161B2 (en) Data learning server and method for generating and using learning model thereof
WO2021179958A1 (zh) 空调器的智能控制方法与空调器的智能控制设备
CN108052010B (zh) 智能电器自调节的方法、装置、计算机设备及存储介质
WO2017056403A1 (ja) 空調制御方法、空調制御装置及び空調制御プログラム
CN112432344B (zh) 空调器及其控制方法、计算机存储介质
CN112128935B (zh) 空调器的智能控制方法与智能控制设备
CN112162492A (zh) 家居设备的控制方法、装置、边缘计算网关及存储介质
WO2019196488A1 (zh) 一种控制家用电器执行控制指令的方法及装置
WO2015193032A1 (en) Controlling a target system
CN113377018A (zh) 空调器的智能控制方法与空调器的智能控制设备
CN112432343A (zh) 空调器及其开机模式的控制方法、存储介质
CN114442697B (zh) 一种温度控制方法、设备、介质及产品
CN113852657B (zh) 一种基于边缘计算的智慧家庭本地控制方法及系统
CN108419439A (zh) 家用设备学习方法、及服务器
CN111007731A (zh) 用户操作预测方法、装置、电子设备及可读存储介质
EP3352027A1 (en) Method and system of smart climatization management
WO2019084784A1 (zh) 智能温控方法、装置、存储介质、中央温度控制器和服务器
RU2772237C2 (ru) Сервер обучения на основе данных и способ формирования и использования его обучающей модели
CN118012546A (zh) 用户界面的显示方法、装置、设备及存储介质
US20220381472A1 (en) Air conditioner, air conditioner control method, and air conditioner control system
CN107436015A (zh) 空调寄生装置的控制方法、空调及空调系统
EP3026962A1 (en) Method and system for regulating power consumption

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21859830

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21859830

Country of ref document: EP

Kind code of ref document: A1