Air conditioning system and control method of air conditioning system

Abstract

The invention discloses an air conditioning system and a control method of the air conditioning system, and relates to the technical field of air conditioners. The method comprises the steps that reservation information and air conditioner environment information corresponding to the air conditioner are obtained; the reserved temperature and the air conditioner environment information are input into a trained first time prediction model, time feature extraction is carried out through a Mangbar reasoning layer in the first time prediction model, and the first duration when the indoor temperature of the area where the air conditioner is located reaches the reserved temperature is determined according to the extracted time features; and the time of the first duration before the reserved starting time serves as the starting time, and when the starting time is up, the air conditioner is controlled to start running. According to the method, the starting time of the air conditioner can be determined according to the reservation information and the air conditioner environment information, so that the indoor temperature of the area where the air conditioner is located reaches the reservation temperature when the user enters the area where the air conditioner is located at the reservation starting time, and the use experience of the user is improved.

Description

Technical Field

[0001] This application relates to the field of air conditioner technology, and in particular to an air conditioning system and a control method for the air conditioning system. Background Technology

[0002] Air conditioners, as a common electrical appliance, are widely used in daily life and production. Air conditioners can change the indoor temperature to provide users with a comfortable environment. However, air conditioners usually need to run for a period of time to reach the user-set temperature.

[0003] In order to ensure that the indoor environment reaches the temperature set by the user at a certain time, the user often needs to manually control the air conditioner to start running in advance, so that the air conditioner changes the indoor ambient temperature in advance to reach the preset temperature.

[0004] However, manually controlling the air conditioner to start running in advance is not only cumbersome, but also leaves users with no way of knowing how far in advance they need to start the air conditioner, resulting in a poor user experience. Summary of the Invention

[0005] The embodiments of this application provide an air conditioning system and a control method for the air conditioning system, which can determine the start-up time of the air conditioner based on reservation information and air conditioning environment information, so that when the user arrives at the reserved start-up time and enters the area where the air conditioner is located, the indoor temperature of the area where the air conditioner is located reaches the reserved temperature, thereby improving the user experience.

[0006] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0007] In a first aspect, embodiments of this application provide a control method for an air conditioning system, comprising: acquiring reservation information and air conditioning environment information corresponding to the air conditioner; the reservation information includes a reserved start-up time and a reserved temperature, and the air conditioning environment information includes at least indoor environment data and outdoor environment data of the area where the air conditioner is located; inputting the reserved temperature and air conditioning environment information into a trained first-time prediction model, extracting time features through the Mamba inference layer in the first-time prediction model, determining the first duration for the indoor temperature of the area where the air conditioner is located to reach the reserved temperature based on the extracted time features; using the first duration before the reserved start-up time as the start-up time, and controlling the air conditioner to start running when the start-up time arrives.

[0008] The technical solution provided in this application provides at least the following beneficial effects: This technical solution uses the Mamba inference layer in the first-time prediction model to determine the first duration for the indoor temperature in the area where the air conditioner is located to reach the preset temperature based on the user-input preset temperature and air conditioning environment information. Then, based on the first duration and the preset start-up time, the start-up time of the air conditioner is determined, thereby controlling the air conditioner. Therefore, using the Mamba inference layer to determine the first duration improves the accuracy of the predicted first duration. Simultaneously, it automatically adjusts the start-up time of the air conditioner according to changes in the preset start-up time and air conditioning environment information, ensuring that when the user enters the area where the air conditioner is located at the preset start-up time, the indoor temperature in the area where the air conditioner is located reaches the preset temperature set by the user, thus improving the user experience.

[0009] In the first implementation of the first aspect, the Mamba inference layer includes an embedding module, a Mamba module, and an output module connected in sequence; the embedding module is used to convert the input discrete data into continuous initial vector features; the Mamba module includes multiple parallel Mamba sub-modules for extracting multiple levels of time features from the initial vector features; the output module is used to determine a first duration based on the multiple levels of time features and the initial vector features.

[0010] In the second implementation of the first aspect, the Mamba submodule is used to perform linear transformation, convolution, activation function operation and state space operation on the initial vector features to obtain temporal features. The state space operation is an operation based on the state space model.

[0011] In the third implementation of the first aspect, the above method further includes: acquiring a power-on training dataset, which includes multiple power-on training data, each power-on training data corresponding to a target duration; each power-on training data includes a target temperature and historical environmental information; inputting the target temperature and historical environmental information into an initial first-time prediction model, extracting time features through the Mamba inference layer in the first-time prediction model, and then determining the predicted duration for the indoor temperature in the area where the air conditioner is located to reach the target temperature based on the extracted time features; training the initial first-time prediction model based on the predicted duration and the target duration to obtain the first-time prediction model.

[0012] In the fourth implementation of the first aspect, obtaining the startup training dataset includes: obtaining the historical startup time points of the air conditioner and the historical environmental information corresponding to the historical startup time points; for each historical startup time point, determining the change in indoor temperature within a first preset duration after the historical startup time point; determining the correspondence between the time difference and temperature difference corresponding to the first preset time based on the change in indoor temperature; the correspondence between the time difference and temperature difference is used to characterize the difference in indoor temperature at any time point within the first preset duration compared to the historical startup time point; based on the historical startup time points and the correspondence between the time difference and temperature difference corresponding to the historical startup time points, obtaining the correspondence between the target duration and target temperature corresponding to the historical startup time points; and generating the startup training dataset based on the historical environmental information corresponding to the historical startup time points and the correspondence between the target duration and target temperature.

[0013] Secondly, embodiments of this application provide a control method for an air conditioning system, comprising: acquiring reservation information and air conditioning environment information corresponding to the air conditioner; the reservation information includes a reserved shutdown time, and the air conditioning environment information includes at least indoor environment data and outdoor environment data of the area where the air conditioner is located; inputting the air conditioning environment information into a trained second time prediction model, extracting time features through the Mamba inference layer in the second time prediction model, determining a second duration for the indoor temperature of the area where the air conditioner is located to reach a preset temperature based on the extracted time features; using the second duration before the reserved shutdown time as the shutdown time, and controlling the air conditioner to stop operating when the shutdown time arrives.

[0014] The technical solution provided in this application offers at least the following beneficial effects: The solution analyzes indoor and outdoor environmental data of the area where the air conditioner is located using the Mamba inference layer in the second time prediction model to determine the cooling time required for the indoor temperature to drop to the preset temperature. This, combined with the scheduled shutdown time, controls the air conditioner to shut down before the scheduled shutdown time, reducing energy waste. Simultaneously, when the scheduled shutdown time arrives, the indoor temperature in the area where the air conditioner is located remains at the preset temperature, avoiding a poor user experience. In other words, the technical solution provided in this application maintains a good user experience while saving energy.

[0015] In the first implementation of the second aspect, the method further includes: acquiring a shutdown training dataset, which includes multiple shutdown training data, each corresponding to a target duration; each shutdown training data includes a target temperature and historical environmental information; inputting the target temperature and historical environmental information into an initial second time prediction model, extracting time features through the Mamba inference layer in the second time prediction model, and then determining the predicted duration for the indoor temperature in the area where the air conditioner is located to reach the target temperature based on the extracted time features; and training the initial second time prediction model based on the predicted duration and the target duration to obtain the second time prediction model.

[0016] Thirdly, this application provides an air conditioning system, including: an air conditioner; and a controller configured to: acquire reservation information and air conditioning environment information corresponding to the air conditioner; the reservation information includes a reserved start-up time and a reserved temperature, and the air conditioning environment information includes at least indoor and outdoor environmental data of the area where the air conditioner is located; input the reserved temperature and air conditioning environment information into a trained first-time prediction model, extract time features through the Mamba inference layer in the first-time prediction model, determine the first duration for the indoor temperature of the area where the air conditioner is located to reach the reserved temperature based on the extracted time features; use the first duration before the reserved start-up time as the start-up time, and control the air conditioner to start running when the start-up time arrives.

[0017] In some embodiments, the Mamba inference layer includes an embedding module, a Mamba module, and an output module connected in sequence; the embedding module is used to convert the input discrete data into continuous initial vector features; the Mamba module includes multiple parallel Mamba sub-modules for extracting multiple levels of time features from the initial vector features; and the output module is used to determine a first duration based on the multiple levels of time features and the initial vector features.

[0018] Fourthly, this application provides an air conditioning system, including: an air conditioner; and a controller configured to: acquire reservation information and air conditioning environment information corresponding to the air conditioner; the reservation information includes a reserved shutdown time, and the air conditioning environment information includes at least indoor environment data and outdoor environment data of the area where the air conditioner is located; input the air conditioning environment information into a trained second time prediction model, extract time features through the Mamba inference layer in the second time prediction model, determine a second duration for the indoor temperature of the area where the air conditioner is located to reach a preset temperature based on the extracted time features; use the second duration before the reserved shutdown time as the shutdown time, and control the air conditioner to stop operating when the shutdown time arrives.

[0019] Fifthly, embodiments of this application provide a controller, including: one or more processors; one or more memories; wherein the one or more memories are used to store computer program code, the computer program code including computer instructions, and when the one or more processors execute the computer instructions, the controller executes any one of the control methods for an air conditioning system provided in the first or second aspect.

[0020] In a sixth aspect, embodiments of this application provide a computer-readable storage medium including computer instructions that, when executed on a computer, cause the computer to perform the methods provided in the first aspect and possible implementations of the first aspect or the second aspect and possible implementations of the second aspect.

[0021] In a seventh aspect, embodiments of the present invention provide a computer program product that can be directly loaded into a memory and contains software code. After being loaded and executed by a computer, the computer program product can implement the methods provided in the first aspect and its possible implementations or the second aspect and its possible implementations.

[0022] It should be noted that the aforementioned computer instructions may be stored, in whole or in part, on a computer-readable storage medium. This computer-readable storage medium may be packaged together with the controller's processor, or it may be packaged separately from the controller's processor; this application does not impose any limitations on this.

[0023] The beneficial effects described in aspects three through seven of this application can be referred to the analysis of the beneficial effects of aspects one or two, and will not be repeated here. Attached Figure Description

[0024] The accompanying drawings are provided to further understand the technical solutions of the present invention and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of the present invention and do not constitute a limitation on the technical solutions of the present invention.

[0025] Figure 1 This is a schematic diagram of the structure of an air conditioning system provided in an embodiment of this application;

[0026] Figure 2 A hardware configuration block diagram of an air conditioning system provided in an embodiment of this application;

[0027] Figure 3 This is a schematic flowchart of a control method for an air conditioning system provided in an embodiment of this application;

[0028] Figure 4 This is a schematic diagram of the structure of a first-time prediction model provided in an embodiment of this application;

[0029] Figure 5 A flowchart illustrating the processing of a Mamba submodule provided in an embodiment of this application;

[0030] Figure 6 A flowchart of an SSM algorithm provided in an embodiment of this application;

[0031] Figure 7 A schematic flowchart of another air conditioning system control method provided in this application embodiment;

[0032] Figure 8 A schematic flowchart of another air conditioning system control method provided in this application embodiment;

[0033] Figure 9 A schematic flowchart of another air conditioning system control method provided in this application embodiment;

[0034] Figure 10 A schematic flowchart of another air conditioning system control method provided in this application embodiment;

[0035] Figure 11 A schematic flowchart of another air conditioning system control method provided in this application embodiment;

[0036] Figure 12 This is a timing diagram illustrating a control method for an air conditioning system provided in an embodiment of this application. Detailed Implementation

[0037] The embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0038] In the description of this invention, it should be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0039] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0040] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. Furthermore, when describing pipelines or channels, the terms "connection" and "linking" used in this application have the meaning of enabling conduction. The specific meaning needs to be understood in conjunction with the context.

[0041] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0042] To facilitate understanding, we will first provide a brief introduction and explanation of some terms or basic concepts of technology involved in the embodiments of this application.

[0043] Terminology Explanation:

[0044] 1. Reversible instance normalization (RevIN) is a standardization processing method.

[0045] 2. The embedding layer is a commonly used network layer in deep learning and natural language processing. Its main function is to map discrete variables (such as words and characters) in high-dimensional space to continuous vector representations in low-dimensional space.

[0046] 3. The Mamba algorithm is a linear time series modeling method.

[0047] 4. Fully Connected Layer (FC) is a common layer structure in deep learning neural networks.

[0048] 5. The State-Space Model (SSM) is a statistical model used to model time-dependent data sequences. It describes the dynamic evolution of data by defining the relationships between a set of implicit state variables and observed variables.

[0049] Currently, air conditioners are a common type of electrical appliance and are widely used in daily production and life. In order to ensure that the air conditioner can provide users with a comfortable ambient temperature, it is often necessary to control the air conditioner to turn on in advance so that the ambient temperature reaches a certain set temperature when in use.

[0050] In related technologies, when users are active in the area where the air conditioner is located, they need to manually control the air conditioner to turn it on in advance so that the indoor temperature reaches the preset temperature at the scheduled start time.

[0051] However, manually controlling the air conditioner to start running in advance is not only cumbersome, but also leaves users with no way of knowing how far in advance they need to start the air conditioner, resulting in a poor user experience.

[0052] Based on this, embodiments of this application provide an air conditioning system and a control method for the air conditioning system. By using a Mamba inference layer in a first-time prediction model, based on the user-inputted preset temperature of the air conditioner and air conditioning environment information, the method obtains the first duration for the indoor temperature in the area where the air conditioner is located to reach the preset temperature. Then, based on the first duration and the preset start-up time, the method determines the start-up time of the air conditioner, thereby controlling the air conditioner. Thus, using the Mamba inference layer to determine the first duration improves the accuracy of the predicted first duration. Simultaneously, it automatically adjusts the start-up time of the air conditioner according to changes in the preset start-up time and air conditioning environment information, ensuring that when the user enters the area where the air conditioner is located at the preset start-up time, the indoor temperature in the area where the air conditioner is located has reached the preset temperature at the preset start-up time, improving the user experience.

[0053] The embodiments provided in this application will now be described in detail with reference to the accompanying drawings.

[0054] Figure 1 This is a schematic diagram of the structure of an air conditioning system provided in an embodiment of this application, such as... Figure 1 As shown, the air conditioning system 1 includes an air conditioner 2.

[0055] It should be noted that the air conditioner can be a multi-split air conditioner, a central air conditioner, etc., and this application does not limit the type of air conditioner.

[0056] As a feasible implementation method, the air conditioning system 1 and the remote control 5 are connected by communication.

[0057] As a feasible implementation method, the air conditioner 2 includes an outdoor unit 3, which is usually installed outdoors and used for heat exchange in the indoor environment.

[0058] In some embodiments, the air conditioner 2 includes an indoor unit 4.

[0059] As a feasible approach, indoor unit 4, taking wall-mounted units as an example, is typically installed on walls indoors (such as in bedrooms, living rooms, conference rooms, etc.). Another example is floor-standing units (…). Figure 1 (Not shown in the image) is also a type of indoor unit.

[0060] It should be noted that indoor unit 4 may include at least one indoor unit, and this application does not limit the number of indoor units of the air conditioner.

[0061] In some embodiments, the air conditioning system 1 includes an environmental information acquisition device 6 ( Figure 1 (Not shown in the image), used to collect air conditioning environment information of the area where air conditioner 2 is located.

[0062] For example, the environmental data acquisition device 6 can be a sensor, which can be installed on any component of the air conditioner 2. Alternatively, the air conditioning environmental information acquisition device 6 can be installed independently of the air conditioner 2 and communicate with the air conditioning system 1. This application does not limit the type and installation method of the environmental information acquisition device 6.

[0063] In some embodiments, the air conditioning system 1 includes a reservation information collection device 7. Figure 1 (Not shown in the image) is used to collect reservation information input by the user. It should be understood that reservation information may include: reservation temperature, reservation power-on time, reservation power-off time, etc., and this embodiment of the application does not limit this.

[0064] In some embodiments, the reservation information collection device 7 can be a terminal device, and the air conditioner 2 obtains reservation information by communicating with the terminal device. This application does not limit the type and setting method of the reservation information collection device 7.

[0065] Figure 2 This is a hardware configuration block diagram of an air conditioning system provided in an embodiment of this application. For example... Figure 2 As shown, the air conditioning system 1 may also include: a controller 1000.

[0066] In some embodiments, the controller 1000 is the control center of the air conditioning system 1, which connects various parts of the air conditioning system 1 through various interfaces and lines, and executes various programs of the air conditioning system by running or executing programs stored in the memory and calling data stored in the memory.

[0067] In some embodiments, the controller 1000 may include one or more processing units 1001.

[0068] In some embodiments, controller 1000 refers to a device that can generate operation control signals based on reservation information and air conditioning environment information, instructing the air conditioner 2 in the air conditioning system 1 to execute a control method. Exemplarily, controller 1000 can be a central processing unit (CPU), a network processor (NP), a digital signal processor (DSP), a microprocessor, a microcontroller, a programmable logic device (PLD), or any combination thereof. Controller 1000 can also be other devices with processing functions, such as circuits, devices, or software modules; this application embodiment does not impose any limitations on this.

[0069] In addition, the controller 1000 can be used to control the various components inside the air conditioning system 1 so that each component can perform the predetermined functions of the air conditioning system 1.

[0070] In some embodiments, such as Figure 2 As shown, the controller 1000 includes a communicator 1002.

[0071] In some embodiments, the communicator 1002 is used to establish communication connections with other network entities, such as establishing communication connections with terminal devices. The communicator 1002 may include a radio frequency (RF) module, a cellular module, a wireless fidelity (WIFI) module, and a GPS module, etc. Taking an RF module as an example, the RF module can be used for signal reception and transmission; specifically, it sends received information to the controller 1000 for processing; additionally, it transmits signals generated by the controller 1000. Typically, the RF circuit may include, but is not limited to, an antenna, at least one amplifier, a transceiver, a coupler, a low-noise amplifier (LNA), a duplexer, etc.

[0072] In some embodiments, such as Figure 2 As shown, the controller 1000 also includes a memory 1003.

[0073] In some embodiments, the memory 1003 may be used to store software programs and data. The controller 1000 executes various functions of the air conditioning system 1 and data processing by running the software programs or data stored in the memory 1003. The memory 1003 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. The memory 1003 stores an operating system that enables the air conditioning system 1 to run. In this application, the memory 1003 may store the operating system and various application programs, and may also store code that executes the control method of the air conditioning system 1 provided in the embodiments of this application.

[0074] Those skilled in the art will understand that Figure 2 The hardware structure shown does not constitute a limitation on the air conditioning system 1. The air conditioning system 1 may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0075] As a feasible implementation method, the air conditioning system 1 communicates with the remote controller 5 through the communicator 1002.

[0076] In some embodiments, the remote controller 5 has the function of communicating with the controller 1000, for example, using infrared or other communication methods. Users can set various operating parameters of the air conditioner 2 via the remote controller 5.

[0077] In other embodiments, the air conditioning system control method provided in this application is applied to the control device of the air conditioning system. The control device can be a personal computer (PC), laptop computer, mobile device, tablet computer, or other similar device. This application does not limit the specific form of the electronic device. Alternatively, the control device can be a single server or a server cluster consisting of multiple servers. In some implementations, the server cluster can be a distributed cluster server. This application does not impose any limitations on this.

[0078] Figure 3 This is a schematic flowchart of a control method for an air conditioning system provided in an embodiment of this application, as shown below. Figure 3 As shown, the method includes the following steps:

[0079] S301. Obtain the reservation information and air conditioning environment information corresponding to the air conditioner.

[0080] The reservation information includes the reserved start time and reserved temperature, and the air conditioning environment information includes at least the indoor and outdoor environmental data of the area where the air conditioner is located.

[0081] As a feasible approach, environmental information collection equipment can be installed in the area where the air conditioner is located. This equipment collects indoor and outdoor environmental data for the area where the air conditioner is located. The controller communicates with the environmental information collection equipment to obtain the indoor and outdoor environmental data for the area where the air conditioner is located.

[0082] As one feasible approach, the reservation information collection device collects the reservation information input by the user, and the controller communicates with the reservation information collection device to obtain at least one reservation message sent by the user.

[0083] The reservation information can be manually entered by the user, such as the user entering the reservation time and temperature via remote control; or it can be voice information entered by the user, such as the user saying "the temperature will reach 25 degrees Celsius at 10 o'clock".

[0084] In some embodiments, the reservation information only includes the reserved start-up time. The controller determines the reserved temperature based on the outdoor environmental data of the area where the air conditioner is located. This reserved temperature is a comfortable temperature for the user, determined by the controller based on the outdoor environmental data of the area on that day. For example, if the reservation information is "to use the area where the air conditioner is located at 10:00," the controller queries the area code and usage date of the area where the air conditioner is located to determine the outdoor environmental data of the area on that day. From the outdoor environmental data of that day, it determines the outdoor environmental data that is no more than 30 minutes different from 10:00, thereby determining the reserved temperature.

[0085] Furthermore, since the user sends multiple reservation messages, the controller can determine the latest reservation message based on the time the reservation message was sent, and use the latest reservation message as the reservation message for controlling the air conditioner.

[0086] Specifically, a user sends a reservation message saying "the temperature will reach 25 degrees Celsius at 10 o'clock," and after a period of time, the user sends another reservation message saying "the temperature will reach 20 degrees Celsius at 10 o'clock," using the latter reservation message as the reservation message for controlling the air conditioner.

[0087] Meanwhile, since the scheduled start-up time includes the scheduled usage date and the scheduled usage time, querying at least one reservation information yields at least one scheduled usage date. Based on the scheduled usage date, at least one reservation information is divided and stored in the database. When the scheduled usage date arrives, the reservation information in the database is called to control the air conditioner to work.

[0088] Specifically, users send reservation information such as "the temperature will reach 25 degrees Celsius at 10:00 on the 25th" and "the temperature will reach 25 degrees Celsius at 10:00 on the 26th". These two reservation information are saved in the database. On the 25th, the reservation information "the temperature will reach 25 degrees Celsius at 10:00 on the 25th" is retrieved, and on the 26th, the reservation information "the temperature will reach 25 degrees Celsius at 10:00 on the 26th" is retrieved.

[0089] S302. Input the reserved temperature and air-conditioning environment information into the trained first-time prediction model, extract time features through the Mamba inference layer in the first-time prediction model, and determine the first time duration for the indoor temperature of the area where the air conditioner is located to reach the reserved temperature based on the extracted time features.

[0090] It should be noted that the Mamba inference layer uses the Mamba algorithm to process the reservation temperature and air conditioning environment information. Because the Mamba algorithm is fast and performs well in processing large amounts of data, the first-time prediction model can not only process a large amount of reservation temperature and air conditioning environment information, but also improve the speed of determining the first duration.

[0091] S303: The start time is the first hour before the scheduled start time, and the air conditioner is controlled to start running when the start time arrives.

[0092] As a feasible implementation method, the controller determines the start time of the air conditioner based on the first duration output by the first time prediction model, which is the first duration before the scheduled start time. When the start time arrives, the controller controls the air conditioner to start running.

[0093] In some embodiments, since the air conditioner is divided into two modes, heating and cooling, the corresponding first-time prediction models for the two modes are also different. The reservation information may also include the working mode of the air conditioner. The controller calls the corresponding first-time prediction model according to the working mode to determine the first duration, and then controls the air conditioner to start running when the start time arrives and to work in heating or cooling mode.

[0094] In other embodiments, the controller can determine whether the air conditioner should be in heating or cooling mode based on outdoor environmental data, then call the corresponding first-time prediction model to determine the first duration, and then control the air conditioner to start running when the start-up time arrives.

[0095] Furthermore, when there are multiple air conditioners in the same area, the first-time prediction model for each air conditioner is different. The controller can determine which air conditioner to control based on its number, call the corresponding first-time prediction model to determine the first duration, and then control the air conditioner to start running when the start-up time arrives.

[0096] Meanwhile, in the case of multiple areas, the controller can first determine the area where the air conditioner is located, then determine which air conditioner to control based on the air conditioner's number, then call the corresponding first-time prediction model to determine the first duration, and then control the air conditioner to start running when the start time arrives, so that the indoor temperature in the area where the air conditioner is located reaches the preset temperature at the preset start time.

[0097] The technical solution provided in this application provides at least the following beneficial effects: By using the Mamba inference layer in the first-time prediction model, based on the user-inputted scheduled temperature of the air conditioner and the air-conditioning environment information, the first duration for the indoor temperature in the area where the air conditioner is located to reach the scheduled temperature is obtained. Then, based on the first duration and the scheduled start-up time, the start-up time of the air conditioner is determined, thereby achieving control of the air conditioner. Therefore, using the Mamba inference layer to determine the first duration improves the accuracy of the predicted first duration. Simultaneously, the start-up time of the air conditioner can be automatically adjusted according to changes in the scheduled start-up time and air-conditioning environment information, ensuring that the indoor temperature in the area where the air conditioner is located reaches the scheduled temperature at the scheduled start-up time, thus improving the user experience.

[0098] In some embodiments, to illustrate in detail the structure of the first-time prediction model, the structure of the Mamba inference layer, and the process by which the Mamba inference layer determines the first duration for the indoor temperature in the area where the air conditioner is located to reach the preset temperature based on the preset temperature and air conditioning environment information, this application combines... Figure 4 , Figure 5 and Figure 6 This section introduces the first-time prediction model.

[0099] Figure 4 This is a schematic diagram of the structure of a first-time prediction model provided in an embodiment of this application, as shown below. Figure 4 As shown, the Mamba inference layer includes an embedding module (embedding layer), a Mamba module (Mamba layer), and an output module (output layer) connected in sequence. The embedding module is used to convert the input discrete data into continuous initial vector features. The Mamba module includes multiple parallel Mamba sub-modules for extracting multiple levels of time features from the initial vector features. The output module is used to determine the first duration based on the multiple levels of time features and the initial vector features.

[0100] It should be noted that since the ability of a single Mamba submodule to extract time features from time series data is limited, in order to enhance the first-time prediction model's understanding of time series data, this application embodiment uses multiple parallel Mamba submodules to extract time features at multiple levels. Each Mamba submodule corresponds to one level, thereby improving the understanding of time series data.

[0101] The reserved temperature and air conditioning environment information input into the model are time series data. Since the air conditioning environment information is sampled based on time signals, the time series data input into the model is discrete.

[0102] As a feasible implementation method, the embedded module converts the discrete input data into continuous initial vector features. By converting the discrete data into continuous initial vector features, the Mamba submodule can better capture the temporal features in the discrete input data, thereby improving the accuracy and precision of the first-time prediction model in determining the first time interval for the indoor temperature in the area where the air conditioner is located to reach the preset temperature.

[0103] The embedding module algorithm converts discrete time series data into continuous initial vector features, and its calculation method is shown in formula (1):

[0104] X E =Embedding(X 0 ) Formula (1)

[0105] Among them, X 0 X represents the input discrete time series data. E This represents the characteristic of continuous initial vectors.

[0106] As a feasible implementation method, taking two parallel Mamba sub-modules as an example, the time features of the same continuous initial vector output by the embedded module are extracted hierarchically, and the calculation method is shown in formula (2):

[0107]

[0108] Among them, Y M1 Y represents the low-level temporal features extracted by the Mamba submodule. M2 This represents the high-level temporal features extracted by the Mamba submodule. The division of temporal feature levels is determined based on the dimensions of the time series data, and no specific limitations are made here.

[0109] As a feasible implementation method, the output module is used to determine the first duration based on the time characteristics and initial vector characteristics at multiple levels.

[0110] Specifically, the output module sums the time features and initial vector features at multiple levels, and the calculation process is shown in formulas (3) and (4):

[0111] Y M =Y M1 +Y M2 Formula (3)

[0112] Y E =Y M +XE Formula (4)

[0113] Among them, Y M Representing multi-level temporal features, Y E This indicates the time characteristics after residual processing.

[0114] It should be noted that the first-time prediction model is prone to losing input data when extracting deeper temporal features. By using residual processing to directly connect the input of the Mamba inference layer to the output, we can ensure that the Mamba inference layer extracts deeper temporal features without losing important information in the input data.

[0115] Furthermore, for Y E The format is converted according to formula (5) to obtain the first duration. The calculation process is shown in formula (5):

[0116] Y = FC(Y) E ) Formula (5)

[0117] Where Y represents the first duration and FC represents the linear transformation.

[0118] As a feasible implementation method, the first-time prediction model also includes an air conditioning data standardization layer. Before the Mamba inference layer extracts time features, the air conditioning data standardization layer standardizes the scheduled temperature and air conditioning environment information. By standardizing the scheduled temperature and air conditioning environment information, the accuracy of the first-time prediction model in determining the first time it takes for the indoor temperature in the area where the air conditioner is located to reach the scheduled temperature is improved.

[0119] Specifically, the air conditioning data standardization layer standardizes the input reserved temperature and air conditioning environment information based on RevIN to obtain standardized time series data. The calculation process is shown in formula (6):

[0120] X 0 =RevIN(X) Formula (6)

[0121] Where X represents the input reservation temperature and air conditioning environment information, X 0 This represents standardized time series data.

[0122] As a feasible implementation method, the first-time prediction model also includes a time prediction layer. The time prediction layer converts the format of the first duration output by the Mamba inference layer based on multiple multilayer perceptrons. The specific process can be shown in formula (7):

[0123]

[0124] Taking two multilayer perceptrons as an example, the output Y of the Mamba inference layer is processed twice. This indicates the first duration after format conversion.

[0125] In some embodiments, such as Figure 4 As shown, the first-time prediction model includes: an air conditioning data standardization layer, a Mamba inference layer, and a time prediction layer. The process of determining the first duration in the first-time prediction model is as follows: the air conditioning data standardization layer standardizes the time series data (air conditioning sequence data) to obtain standardized time series data; the Mamba inference layer extracts time features based on the standardized time series data and determines the first duration; and the time prediction layer converts the first duration into a format to obtain the format-converted first duration (on / off prediction time).

[0126] The Mamba inference layer comprises an embedding module (embedding layer), Mamba sub-modules (Mamba layers), an output module (output layer), a fully connected layer, and residual information. The embedding module (embedding layer) processes the standardized time-series data output from the air conditioning data standardization layer to obtain continuous initial vector features. The Mamba sub-modules (Mamba layers) extract multiple levels of time features from the initial vector features. The output module determines the first duration based on the multiple levels of time features and the initial vector features. The fully connected layer performs data format conversion in the Mamba inference layer, and the residual information represents the direct connection between the input and output of the Mamba inference layer through residual processing.

[0127] Furthermore, the Mamba submodule is used to perform linear transformation, convolution, activation function operations, and state-space operations on the initial vector features to obtain temporal features. The state-space operations are performed based on the state-space model.

[0128] Figure 5 A processing flowchart of a Mamba submodule provided in this application embodiment is shown below. Figure 5 As shown, the Mamba submodule includes: a linear transformation function, a convolution function, an activation function, and a state-space model. The processing procedure of the Mamba submodule can be shown in formula (8):

[0129] Mamba(X)=FC[Tanh(FC(X))×SSM(Tanh(Conv(FC(X))))] Formula (8)

[0130] Wherein, FC represents linear transformation, Tanh(X) represents activation function (Tanh) operation, Conv(X) represents convolution (Conv) operation, and SSM represents state space operation.

[0131] Specifically, the implementation process of the SSM algorithm can be shown in equations (9)-(14):

[0132] h'(t)=Ah(t)+Bx(t) Formula (9)

[0133] y(t)=Ch(t) formula (10)

[0134] Where A represents the state transition matrix, B represents the input matrix, C represents the observation matrix, x(t) represents the input feature, y(t) represents the output feature, and h(t) represents the intermediate state feature.

[0135] Since the input data may be discrete, the three matrices A, B, and C are discretized to obtain discrete matrices.

[0136]

[0137] Where Δ represents the sampling period, This represents the matrix after A is discretized. This represents the matrix after B is discretized.

[0138] Furthermore, after discretization, formulas (9) and (10) can be expressed by formulas (13) and (14):

[0139]

[0140] y(t)=Ch(t) Formula (14)

[0141] like Figure 6 As shown, Figure 6 A flowchart of an SSM algorithm provided in this application embodiment. Figure 6 This indicates that the previous time step h will be used. t-1 The hidden state is combined with the current input x(t) to create a new hidden state h. t .

[0142] In some embodiments, since the initial first-time prediction model has not been trained iteratively with a large amount of data, there may be inaccurate output results. Therefore, the first-time prediction model needs to be trained before using it to determine the first duration. Figure 7 A schematic flowchart of another air conditioning system control method provided in this application embodiment, which includes the following steps before step 301:

[0143] S701. Obtain the boot training dataset.

[0144] The boot training dataset includes multiple boot training datasets, each corresponding to a target duration; each boot training dataset includes target temperature and historical environmental information.

[0145] As a feasible implementation method, the controller can determine the historical environmental information of the area where the air conditioner is located; determine the historical environmental information and target temperature corresponding to the air conditioner based on the air conditioner's number; and use the historical environmental information and target temperature as the start-up training dataset for the air conditioner.

[0146] In some embodiments, the controller can also calculate the efficiency of the area where the air conditioner is located, and add the efficiency, historical environmental information, and target temperature to the air conditioner's startup training data. Since an area may include multiple air conditioners, the efficiency of multiple air conditioners within that area is the same.

[0147] Efficiency is used to distinguish different areas where the air conditioner is located. The calculation process is shown in formula (15):

[0148]

[0149] Among them, end time Indicates the air conditioner's shutdown time, begin time Indicates the air conditioner's operating time. This indicates the indoor ambient temperature when the air conditioner is turned off. This indicates the indoor ambient temperature when the air conditioner is turned on.

[0150] S702. Input the target temperature and historical environmental information into the initial first-time prediction model, extract time features through the Mamba inference layer in the first-time prediction model, and then determine the predicted time for the indoor temperature of the area where the air conditioner is located to reach the target temperature based on the extracted time features.

[0151] S703. Train the initial first-time prediction model based on the predicted duration and the target duration to obtain the first-time prediction model.

[0152] As a feasible approach, a loss function is determined based on the predicted duration and the target duration. The first-time prediction model is then iteratively trained based on the loss function until the iteration conditions are met, thus obtaining the first-time prediction model.

[0153] It should be noted that the iteration stopping condition can be the number of training iterations reaching a preset number, the training duration reaching a preset duration, or other possible iteration stopping conditions. This application does not limit the iteration stopping condition.

[0154] Furthermore, after the first-time prediction model is trained, the reserved temperature and air-conditioning environment information are input into the trained first-time prediction model. The time features are extracted through the Mamba inference layer in the first-time prediction model. Based on the extracted time features, the first time it takes for the indoor temperature in the area where the air conditioner is located to reach the reserved temperature is determined, and then the start-up time of the air conditioner is determined. When the start-up time arrives, the air conditioner is controlled to start running.

[0155] The technical solution provided in this embodiment brings at least the following beneficial effects: by obtaining the boot training dataset to train the initialized first-time prediction model, a well-trained first-time prediction model is obtained, thereby improving the accuracy of the output results of the first-time prediction model.

[0156] In some embodiments, in order to improve the training speed and accuracy of the first-time prediction model, it is necessary to screen the boot training dataset before training the first-time prediction model. Figure 8 This is a schematic flowchart of another air conditioning system control method provided in an embodiment of this application, as shown below. Figure 8 As shown, step S701 above includes the following steps:

[0157] S801. Obtain the historical start-up time of the air conditioner, as well as the historical environmental information corresponding to the historical start-up time.

[0158] As a feasible approach, based on the historical usage data of the air conditioner, the historical start-up time of the air conditioner and the corresponding historical environmental information are determined. The historical environmental information is then first filtered to remove data where the indoor temperature remains unchanged.

[0159] S802. For each historical power-on time point, determine the change in indoor temperature within the first preset time period after the historical power-on time point.

[0160] For example, the first preset duration is 25 minutes. Taking the operation process of an air conditioner on a certain day as an example, the start-up time and the stop-off time of the air conditioner are determined. The indoor temperature corresponding to the air conditioner between the start-up time and the stop-off time is the start-up training data. Since there is a lot of start-up training data between the start-up time and the stop-off time, it is necessary to perform a second screening on the start-up training data between the start-up time and the stop-off time to determine the indoor temperature corresponding to the air conditioner within 25 minutes after the start-up time.

[0161] Furthermore, the values ​​of constant indoor temperature in the 25-minute startup training data were subjected to a second filtering to obtain the filtered data.

[0162] Furthermore, the filtered data is interpolated at every minute interval to generate processed data. This interpolation process improves the continuity of indoor temperature data, thereby increasing the accuracy of the results output by the first-time prediction model.

[0163] S803. Determine the correspondence between the time difference and temperature difference corresponding to the first preset time based on the changes in indoor temperature.

[0164] The correspondence between time difference and temperature difference is used to characterize the difference in indoor temperature at any point within the first preset time period compared to the historical start-up time.

[0165] As a feasible implementation method, the time point and indoor temperature corresponding to the power-on time point are used as the benchmark. The time point and indoor temperature corresponding to the power-on time point are subtracted from the time point and indoor temperature corresponding to each minute to obtain the time difference and temperature difference between each minute and the power-on time point within the first preset time period, and then the correspondence between the time difference value and the temperature difference value is obtained.

[0166] S804. Based on the historical power-on time points and the corresponding time and temperature differences, the corresponding target duration and target temperature are obtained for each historical power-on time point.

[0167] S805. Generate a boot training dataset based on the historical environmental information corresponding to the historical boot time points and the correspondence between the target duration and the target temperature.

[0168] As a feasible approach, the indoor and outdoor environmental data are obtained within a second preset time range from the start-up time of the air conditioner, based on the start-up time. The correspondence between the target time and the target temperature, along with the indoor and outdoor environmental data, are combined to form the start-up training dataset.

[0169] The technical solution provided in this embodiment brings at least the following beneficial effects: by obtaining the historical environmental information of the air conditioner within a first preset time after the historical start-up time, and obtaining the correspondence between the time difference and the temperature difference, a start-up training dataset is generated based on the historical environmental information corresponding to the historical start-up time and the correspondence between the target time and the target temperature, thereby improving the accuracy of the start-up training data in the start-up training dataset.

[0170] Air conditioners, as a common electrical appliance, are closely related to people's daily lives and work. However, due to their high power consumption, they consume a significant amount of energy during use. Currently, users often wait until they are finished before turning off the air conditioner in its area. However, even after the air conditioner is turned off, the area remains at a relatively comfortable temperature for a period of time, easily leading to energy waste.

[0171] Based on this, this application proposes a control method for an air conditioning system. Please refer to [link / reference]. Figure 9 , Figure 9 This is a schematic flowchart of another air conditioning system control method provided in an embodiment of this application, as shown below. Figure 9 As shown, the method includes the following steps:

[0172] S901. Obtain the reservation information and air conditioning environment information corresponding to the air conditioner.

[0173] The reservation information includes the scheduled shutdown time, and the air conditioning environment information includes at least indoor and outdoor environmental data for the area where the air conditioner is located.

[0174] For example, the implementation process of S901 can be referred to S301, and will not be repeated here.

[0175] S902. Input the air conditioning environment information into the trained second time prediction model, extract time features through the Mamba inference layer in the second time prediction model, and determine the second duration for the indoor temperature of the area where the air conditioner is located to reach the preset temperature based on the extracted time features.

[0176] The preset temperature is determined in advance based on outdoor and indoor environmental data.

[0177] It should be noted that in some embodiments, the preset temperature can be adjusted as needed. In this case, the preset temperature and air conditioning environment information are input into the trained second time prediction model, and the time features are extracted through the Mamba inference layer in the second time prediction model. Based on the extracted time features, the second duration for the indoor temperature in the area where the air conditioner is located to reach the preset temperature is determined.

[0178] For example, the process of the second time prediction model determining the second duration can refer to the process of the first time prediction model determining the first duration in S302, and will not be repeated here.

[0179] S903: The shutdown time is set to the second time elapsed before the scheduled shutdown time, and the air conditioner is controlled to stop running when the shutdown time arrives.

[0180] For example, the implementation process of S903 can be referred to S303, and will not be repeated here.

[0181] The technical solution provided in this application provides at least the following beneficial effects: By analyzing the indoor and outdoor environmental data of the area where the air conditioner is located through the Mamba inference layer in the second time prediction model, the cooling time required for the indoor temperature in the area where the air conditioner is located to drop to the preset temperature is determined. This, combined with the scheduled shutdown time, controls the air conditioner to shut down before the scheduled shutdown time, reducing energy waste. Simultaneously, when the scheduled shutdown time arrives, the indoor temperature in the area where the air conditioner is located remains at the preset temperature, avoiding a poor user experience. In other words, the technical solution provided in this application can maintain a good user experience while saving energy.

[0182] In some embodiments, the Mamba inference layer of the second prediction model includes an embedding module, a Mamba module, and an output module connected in sequence; the embedding module is used to convert the input discrete data into continuous initial vector features; the Mamba module includes multiple parallel Mamba sub-modules for extracting multiple levels of time features from the initial vector features; and the output module is used to determine the target duration based on the multiple levels of time features and the initial vector features.

[0183] The Mamba submodule is used to perform linear transformation, convolution, activation function operations, and state-space operations on the initial vector features to obtain temporal features. The state-space operations are performed based on the state-space model.

[0184] The structure of the Mamba inference layer and the process by which the Mamba inference layer determines the second time it takes for the indoor temperature in the area where the air conditioner is located to reach the preset temperature based on the reserved temperature and air conditioning environment information can be referred to the description of the first-time prediction model in the above embodiment, and will not be repeated here.

[0185] In some implementations, the initial second-time prediction model may not have undergone extensive data iteration training, potentially leading to inaccurate output results. Therefore, before predicting the shutdown time, the second-time prediction model needs to be trained. The training process can refer to the training process of the first-time prediction model. Figure 10 This is a schematic flowchart of another air conditioning system control method provided in an embodiment of this application, as shown below. Figure 10 As shown, the specific steps include the following:

[0186] S1001. Obtain the shutdown training dataset.

[0187] The shutdown training dataset includes multiple shutdown training datasets, each corresponding to a target duration; each shutdown training dataset includes target temperature and historical environmental information.

[0188] S1002. Input the target temperature and historical environmental information into the initial second time prediction model, extract time features through the Mamba inference layer in the second time prediction model, and then determine the predicted time for the indoor temperature of the area where the air conditioner is located to reach the target temperature based on the extracted time features.

[0189] S1003. Train the initial second time prediction model based on the predicted duration and the target duration to obtain the second time prediction model.

[0190] The technical solution provided in this application provides at least the following beneficial effects: by obtaining the shutdown training dataset to train the initialized second time prediction model, a well-trained second time prediction model is obtained, thereby improving the accuracy of the output results of the second time prediction model.

[0191] In some embodiments, Figure 11 This is a schematic flowchart of another air conditioning system control method provided in an embodiment of this application, as shown below. Figure 11 As shown, the process of controlling the air conditioning system includes:

[0192] S1101. Determine the original data.

[0193] The raw data includes: indoor unit operating data, outdoor weather data, indoor unit equipment attributes, and historical air conditioner data. The raw data is equivalent to the historical start-up times and the corresponding historical environmental information mentioned above.

[0194] S1102, Data Preprocessing.

[0195] As a feasible approach, data preprocessing can involve filtering the original data to obtain filtered data, grouping the filtered data according to the air conditioner's on / off cycle to obtain grouped data, performing interpolation on the grouped data, constructing features from the interpolated data, and finally generating test and training sets.

[0196] S1103. Train the model based on the training set.

[0197] As a feasible approach, the process of training a model based on a training set can include: initializing the model based on the model parameters, training the initialized model based on the training set, validating the model and optimizing the gradient based on the predicted duration and the target duration, obtaining the trained model and outputting the model.

[0198] S1104. Based on the test set, call the trained model to perform time prediction and call the command to control the air conditioner to turn on and off.

[0199] As a feasible implementation method, the process of calling the trained model based on the test set to predict the time and calling the command to control the air conditioner to start or stop operation may include: calling the output model, inputting the test set data into the model to obtain the output predicted time, and calling the command according to the predicted time and the scheduled start / stop time to control the air conditioner to start or stop operation.

[0200] Figure 12 A timing diagram illustrating a control method for an air conditioning system provided in an embodiment of this application, as shown below. Figure 12 As shown, based on the process of controlling the air conditioner's on / off state mentioned in the above embodiment, for example, the air conditioner's scheduled start time is 10:00 AM and scheduled stop time is 11:00 AM, with a first duration of 10 minutes and a second duration of 5 minutes. The operating sequence of the air conditioner in the air conditioning system is as follows: before the air conditioner starts running, the indoor temperature is T1, the scheduled temperature is T2, and the preset temperature is T3. When the time point of 9:50 AM is reached, the air conditioner is controlled to start running, and the indoor temperature begins to decrease from T1, reaching T2 at 10:00 AM. When the time point of 10:55 AM is reached, the air conditioner is controlled to stop running, and the indoor temperature begins to rise from T2, reaching T3 at 11:00 AM.

[0201] Based on the control method of the air conditioning system mentioned in the above embodiments, this application also proposes an air conditioning system for controlling the start-up of the air conditioner, including: an air conditioner and a controller.

[0202] The controller is configured to: acquire reservation information and air conditioning environment information corresponding to the air conditioner; the reservation information includes the reserved start time and reserved temperature, and the air conditioning environment information includes at least indoor and outdoor environmental data of the area where the air conditioner is located; input the reserved temperature and air conditioning environment information into the trained first-time prediction model, extract time features through the Mamba inference layer in the first-time prediction model, determine the first time interval for the indoor temperature of the area where the air conditioner is located to reach the reserved temperature based on the extracted time features; use the first time interval before the reserved start time as the start time, and control the air conditioner to start running when the start time arrives.

[0203] In some embodiments, the Mamba inference layer includes an embedding module, a Mamba module, and an output module connected in sequence; the embedding module is used to convert the input discrete data into continuous initial vector features; the Mamba module includes multiple parallel Mamba sub-modules for extracting multiple levels of time features from the initial vector features; and the output module is used to determine a first duration based on the multiple levels of time features and the initial vector features.

[0204] In some embodiments, the Mamba submodule is used to perform linear transformation, convolution, activation function operation, and state space operation on the initial vector features to obtain temporal features. The state space operation is an operation based on the state space model.

[0205] In some embodiments, the controller is further configured to: acquire a power-on training dataset, which includes multiple power-on training data sets, each corresponding to a target duration; each power-on training data set includes a target temperature and historical environmental information; input the target temperature and historical environmental information into an initial first-time prediction model, extract time features through the Mamba inference layer in the first-time prediction model, and then determine the predicted duration for the indoor temperature in the area where the air conditioner is located to reach the target temperature based on the extracted time features; and train the initial first-time prediction model based on the predicted duration and the target duration to obtain the first-time prediction model.

[0206] In some embodiments, the controller is further configured to: acquire historical start-up times of the air conditioner and historical environmental information corresponding to those historical start-up times; for each historical start-up time, determine the change in indoor temperature within a first preset duration after the historical start-up time; determine the correspondence between time difference and temperature difference corresponding to the first preset time based on the change in indoor temperature; the correspondence between time difference and temperature difference is used to characterize the difference in indoor temperature at any time point within the first preset duration compared to the historical start-up time; based on the historical start-up time and the correspondence between time difference and temperature difference corresponding to the historical start-up time, obtain the correspondence between target duration and target temperature corresponding to the historical start-up time; and generate a start-up training dataset based on the historical environmental information corresponding to the historical start-up time and the correspondence between target duration and target temperature.

[0207] For example, the specific working process of the controller controlling the air conditioner to start can be referred to the above method embodiment, and will not be repeated here.

[0208] Based on the control method of the air conditioning system mentioned in the above embodiments, this application also proposes an air conditioning system for controlling the air conditioner to turn off, including: an air conditioner and a controller.

[0209] The controller is configured to: acquire reservation information and air conditioning environment information corresponding to the air conditioner; the reservation information includes the scheduled shutdown time, and the air conditioning environment information includes at least indoor and outdoor environmental data of the area where the air conditioner is located; input the air conditioning environment information into a trained second time prediction model, extract time features through the Mamba inference layer in the second time prediction model, determine the second duration for the indoor temperature of the area where the air conditioner is located to reach the preset temperature based on the extracted time features; use the second duration before the scheduled shutdown time as the shutdown time, and control the air conditioner to stop running when the shutdown time arrives.

[0210] In some embodiments, the Mamba inference layer includes an embedding module, a Mamba module, and an output module connected in sequence; the embedding module is used to convert the input discrete data into continuous initial vector features; the Mamba module includes multiple parallel Mamba sub-modules for extracting multiple levels of time features from the initial vector features; and the output module is used to determine a second duration based on the multiple levels of time features and the initial vector features.

[0211] In some embodiments, the Mamba submodule is used to perform linear transformation, convolution, activation function operation, and state space operation on the initial vector features to obtain temporal features. The state space operation is an operation based on the state space model.

[0212] In some embodiments, the controller is further configured to: acquire a shutdown training dataset, which includes multiple shutdown training data sets, each corresponding to a target duration; each shutdown training data set includes a target temperature and historical environmental information; input the target temperature and historical environmental information into an initial second time prediction model, extract time features through the Mamba inference layer in the second time prediction model, and then determine the predicted duration for the indoor temperature in the area where the air conditioner is located to reach the target temperature based on the extracted time features; and train the initial second time prediction model based on the predicted duration and the target duration to obtain the second time prediction model.

[0213] In some embodiments, the controller is further configured to: acquire historical shutdown times of the air conditioner and historical environmental information corresponding to those historical shutdown times; for each historical shutdown time, determine the change in indoor temperature within a first preset duration after the historical shutdown time; determine the correspondence between time difference and temperature difference corresponding to the first preset time based on the change in indoor temperature; the correspondence between time difference and temperature difference is used to characterize the difference in indoor temperature at any time point within the first preset duration compared to the historical shutdown time; based on the historical shutdown time and the correspondence between time difference and temperature difference corresponding to those historical shutdown times, obtain the correspondence between target duration and target temperature corresponding to the historical shutdown time; and generate a shutdown training dataset based on the historical environmental information corresponding to the historical shutdown time and the correspondence between target duration and target temperature.

[0214] For example, the specific working process of the controller controlling the air conditioner to turn off can be referred to the above method embodiment, and will not be repeated here.

[0215] This invention also provides a computer-readable storage medium, which includes computer-executed speech, such that when the computer-executed speech is run on a computer, the computer performs the method provided in the above embodiments.

[0216] This invention also provides a computer program product that can be directly loaded into a memory and contains software code. After being loaded and executed by a computer, the computer program product can implement the methods provided in the above embodiments.

[0217] Although this application has been described herein in conjunction with various embodiments, those skilled in the art, by reviewing the accompanying drawings, the disclosure, and the appended claims, will understand and implement other variations of the disclosed embodiments in carrying out the claimed application. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude multiple instances. A single processor or other unit can implement several functions listed in the claims. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce good results.

[0218] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely exemplary illustrations of this application as defined by the appended claims, and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from the spirit and scope of this application. Thus, if such modifications and modifications of this application fall within the scope of the claims of this application and their equivalents, this application is also intended to include such modifications and modifications.

[0219] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An air conditioning system, characterized in that, include: Air conditioner; The controller is configured as follows: Obtain the reservation information and air conditioning environment information corresponding to the air conditioner; The reservation information includes the reserved start-up time and reserved temperature, and the air conditioning environment information includes at least the indoor and outdoor environmental data of the area where the air conditioner is located. The reserved temperature and the air conditioning environment information are input into the trained first time prediction model. The time features are extracted through the Mamba inference layer in the first time prediction model. Based on the extracted time features, the first duration for the indoor temperature of the area where the air conditioner is located to reach the reserved temperature is determined. The start time is defined as the first duration preceding the scheduled start time, and the air conditioner is controlled to start operating when the start time arrives.

2. The air conditioning system according to claim 1, characterized in that, The Mamba inference layer includes an embedded module, a Mamba module, and an output module connected in sequence. The embedding module is used to convert the input discrete data into continuous initial vector features; The Mamba module includes multiple parallel Mamba sub-modules for extracting multiple levels of temporal features from the initial vector features. The output module is used to determine the first duration based on the time features of the multiple levels and the initial vector features.

3. An air conditioning system, characterized in that, include: Air conditioner; The controller is configured as follows: Obtain the reservation information and air conditioning environment information corresponding to the air conditioner; The reservation information includes the scheduled shutdown time, and the air conditioning environment information includes at least indoor and outdoor environmental data of the area where the air conditioner is located. The air conditioning environment information is input into the trained second time prediction model. The time features are extracted through the Mamba inference layer in the second time prediction model. Based on the extracted time features, the second duration for the indoor temperature of the area where the air conditioner is located to reach the preset temperature is determined. The second duration preceding the scheduled shutdown time is taken as the shutdown time, and the air conditioner is controlled to stop operating when the shutdown time arrives.

4. A control method for an air conditioning system, characterized in that, include: Obtain the reservation information and air conditioning environment information corresponding to the air conditioner; The reservation information includes the reserved start-up time and reserved temperature, and the air conditioning environment information includes at least the indoor and outdoor environmental data of the area where the air conditioner is located. The reserved temperature and the air conditioning environment information are input into the trained first time prediction model. The time features are extracted through the Mamba inference layer in the first time prediction model. Based on the extracted time features, the first duration for the indoor temperature of the area where the air conditioner is located to reach the reserved temperature is determined. The start time is defined as the first duration preceding the scheduled start time, and the air conditioner is controlled to start operating when the start time arrives.

5. The method according to claim 4, characterized in that, The Mamba inference layer includes an embedded module, a Mamba module, and an output module connected in sequence. The embedding module is used to convert the input discrete data into continuous initial vector features; The Mamba module includes multiple parallel Mamba sub-modules for extracting multiple levels of temporal features from the initial vector features. The output module is used to determine the first duration based on the time features of the multiple levels and the initial vector features.

6. The method according to claim 5, characterized in that, The Mamba submodule is used to perform linear transformation, convolution, activation function operation, and state space operation on the initial vector features to obtain the temporal features. The state space operation is based on the state space model.

7. The method according to claim 4, characterized in that, The method further includes: Obtain the boot-up training dataset, which includes multiple boot-up training data sets, each corresponding to a target duration; each boot-up training data set includes a target temperature and historical environmental information; The target temperature and the historical environmental information are input into the initial first time prediction model. Time features are extracted through the Mamba inference layer in the first time prediction model. Then, the predicted time for the indoor temperature in the area where the air conditioner is located to reach the target temperature is determined based on the extracted time features. The initial first time prediction model is trained based on the predicted duration and the target duration to obtain the first time prediction model.

8. The method according to claim 7, characterized in that, The process of obtaining the boot training dataset includes: Obtain the historical start-up time points of the air conditioner, and the historical environmental information corresponding to the historical start-up time points; For each historical power-on time point, determine the change in indoor temperature within a first preset time period after the historical power-on time point; The correspondence between the time difference and the temperature difference corresponding to the first preset time is determined based on the changes in the indoor temperature; the correspondence between the time difference and the temperature difference is used to characterize the difference in indoor temperature at any point within the first preset time period compared to the historical power-on time point; Based on the historical power-on time points and the corresponding time and temperature differences, the target duration and target temperature corresponding to the historical power-on time points are obtained. The boot training dataset is generated based on the historical environmental information corresponding to the historical boot time points and the correspondence between target duration and target temperature.

9. A control method for an air conditioning system, characterized in that, include: Obtain the reservation information and air conditioning environment information corresponding to the air conditioner; The reservation information includes the scheduled shutdown time, and the air conditioning environment information includes at least indoor and outdoor environmental data of the area where the air conditioner is located. The air conditioning environment information is input into the trained second time prediction model. The time features are extracted through the Mamba inference layer in the second time prediction model. Based on the extracted time features, the second duration for the indoor temperature of the area where the air conditioner is located to reach the preset temperature is determined. The second duration preceding the scheduled shutdown time is taken as the shutdown time, and the air conditioner is controlled to stop operating when the shutdown time arrives.

10. The method according to claim 9, characterized in that, The method further includes: Obtain a shutdown training dataset, which includes multiple shutdown training data sets, each corresponding to a target duration; each shutdown training data set includes a target temperature and historical environmental information; The target temperature and the historical environmental information are input into the initial second time prediction model. Time features are extracted through the Mamba inference layer in the second time prediction model. Then, the predicted time for the indoor temperature in the area where the air conditioner is located to reach the target temperature is determined based on the extracted time features. The initial second time prediction model is trained based on the predicted duration and the target duration to obtain the second time prediction model.

Images