Air conditioning method, apparatus, storage medium, and program product

Abstract

This invention discloses an air conditioning method, device, storage medium, and program product, relating to the field of air conditioning technology. The air conditioning method includes: receiving a user's voice control command and determining the control requirements of the voice control command through a preset large language model; generating a control task corresponding to the control requirements based on the control requirements and scene environment data of the IoT scenario corresponding to the acquired voice control command, wherein the control task includes a target control device and control sub-tasks of the target control device; and controlling the target control device of the IoT scenario according to the control task. This invention solves the technical problem of low intelligence level in voice control.

Description

Technical Field

[0001] This invention relates to the field of air conditioning equipment technology, and in particular to an air conditioning method, equipment, storage medium, and program product. Background Technology

[0002] With the development of smart home technology, users' demand for controlling various devices in their homes via voice commands is increasing. For example, voice control of air conditioning devices usually requires users to predefine scene names and configure the control strategies and trigger command words corresponding to those scene names. When the user's voice hits the trigger command word, the air conditioning device will execute the control strategy corresponding to that trigger command word to achieve the corresponding scene control.

[0003] However, because this voice control method relies on fixed trigger command words and fixed control strategies, air conditioning equipment lacks flexibility when facing users' actual control needs and is difficult to meet those needs. Therefore, there is currently a technical problem of low intelligence in voice control.

[0004] The above content is only used to help understand the technical solutions of the embodiments of the present invention, and does not represent an admission that the above content is prior art. Summary of the Invention

[0005] The main objective of this invention is to provide an air conditioning method, device, storage medium, and program product, aiming to solve the technical problem of low intelligence level in voice control.

[0006] To achieve the above objectives, the present invention provides an air conditioning method, the air conditioning method comprising:

[0007] Receive user's voice control commands and determine the control requirements of the voice control commands through a preset large language model;

[0008] Based on the control requirements and the scene environment data of the IoT scenario corresponding to the acquired voice control command, a control task corresponding to the control requirements is generated, wherein the control task includes a target control device and a control sub-task of the target control device.

[0009] Control the target control device in the IoT scenario according to the control task.

[0010] In one embodiment, the step of determining the control requirements of the voice control command through a preset large language model includes:

[0011] Convert the voice data in the voice control commands into text data;

[0012] Based on the text data, determine the instruction type of the voice control command;

[0013] When the instruction type is a preset pre-planned type, the text data is input into a preset large language model, and the text data is semantically analyzed by the preset large language model to obtain the control requirements.

[0014] In one embodiment, the step of determining the instruction type of the voice control instruction based on the text data includes:

[0015] The text data is input into a preset instruction classification model, and the text data is classified by the instruction classification model to obtain the instruction type of the voice control instruction.

[0016] In one embodiment, the step of generating a control task corresponding to the control requirement based on the control requirement and the scene environment data of the IoT scene corresponding to the acquired voice control command includes:

[0017] By using a pre-defined large language model and the corresponding thought chain, the requirement keywords are extracted from the control requirements, and the scene environment data of the Internet of Things scenario is obtained through the pre-defined large language model.

[0018] Based on the aforementioned requirement keywords, the target control device is identified in the scenario environment data, and the control sub-task of the target control device is determined.

[0019] In one embodiment, the scene environment data includes a space, an air conditioning device within the space, and the device functions of the air conditioning device;

[0020] The step of obtaining the scene environment data of the IoT scenario through the preset large language model includes:

[0021] Obtain the IoT account of the voice acquisition device that collects the voice control commands;

[0022] Using the preset large language model, the device query component is invoked to find the space corresponding to the IoT account, the air conditioning device corresponding to each space, and the device function of each air conditioning device.

[0023] If no environmental control value is found in the required keywords of the control requirements, the environmental query component is invoked through a preset large language model to find the regional environmental parameters of the area where the air conditioning device is located, and it is determined that the scene environmental data includes the regional environmental parameters.

[0024] In one embodiment, the control subtask includes an advance subtask and an agreement subtask;

[0025] The steps of determining the target control device from the scene environment data based on the required keywords, and determining the control sub-task of the target control device, include:

[0026] By using a pre-set large language model, the target control space corresponding to the required keywords is determined in each space of the IoT account, and the target control device is determined in each air conditioning device in the target control space.

[0027] For each target control device, if the scene environment data does not include regional environment parameters, then based on the preset large language model, the device function of the target control device, and the requirement keywords, the advance sub-tasks and agreed sub-tasks of the target control device are generated.

[0028] For each target control device, if the scene environment data includes regional environment parameters, then based on the preset large language model, the device function of the target control device, the requirement keywords, and the regional environment parameters, the advance sub-tasks and agreed sub-tasks of the target control device are generated.

[0029] The advance subtask includes an advance start time and advance control parameters, while the agreed subtask includes an agreed start time and agreed control parameters, with the advance start time being earlier than the agreed start time.

[0030] In one embodiment, the method includes:

[0031] The system identifies additional requirements for generating control tasks that are not included in the control requirements by using a pre-defined large language model.

[0032] By calling the context component through the preset large language model, based on the acquisition time of the voice control command, related voice data is obtained in the first preset duration segment before the acquisition time and the second preset duration segment after the acquisition time. The keywords to be supplemented are obtained from the related voice data and the keywords to be supplemented are added to the required keywords.

[0033] In one embodiment, the step of controlling the target control device in the IoT scenario according to the control task includes:

[0034] The expected execution effect of the control task is determined by pre-setting a large language model;

[0035] The expected execution effect is verified. If the verification passes, the control subtask of the target control device is sent to the target control device via the Internet of Things protocol so that the target control device can execute the control subtask.

[0036] In one embodiment, the step of verifying the expected execution effect includes:

[0037] By using a pre-defined large language model, the desired effect of the control requirement expression is determined, and the expected execution effect is evaluated to determine whether the desired effect is met.

[0038] By using a pre-defined large language model, the target device functions required for the target control device to execute the control sub-task are determined, and it is determined whether the device functions of the target control device include the target device functions.

[0039] If the expected execution effect meets the effect requirements, and the device function of the target control device includes the target device function, then the verification of the expected execution effect is determined to be successful.

[0040] If the expected execution effect does not meet the effect requirements, and / or the device function of the target control device does not include the target device function, then it is determined that the verification of the expected execution effect fails, and the execution step is returned: determine the control requirements from the text data through the preset large language model until the verification of the expected execution effect passes.

[0041] The present invention also provides an air conditioning device, the device comprising:

[0042] The requirement determination module is used to receive the user's voice control command and determine the control requirement of the voice control command through a preset large language model.

[0043] The task generation module is used to generate a control task corresponding to the control requirement based on the control requirement and the scene environment data of the IoT scene corresponding to the obtained voice control command. The control task includes a target control device and a control sub-task of the target control device.

[0044] An execution module is used to control the target control device in the IoT scenario according to the control task.

[0045] The present invention also provides an air conditioning device, including a main body and an air conditioning unit disposed within the main body; the air conditioning unit includes a memory, a processor, and an air conditioning program stored in the memory and executable on the processor, wherein when the air conditioning program is executed by the processor, it performs the steps of the air conditioning method described above.

[0046] In addition, to achieve the above objectives, embodiments of the present invention also provide a computer program product, including a computer program that, when executed by a processor, implements the steps of the air conditioning method described above.

[0047] The present invention also provides a computer-readable storage medium storing an air conditioning program that can run on a processor, the air conditioning program being invoked by the processor to implement the steps of the air conditioning method described above.

[0048] This invention provides an air conditioning method that achieves at least the following technical effects: This invention can receive user voice control commands and determine the control requirements of the voice control commands through a preset large language model. This allows for semantic analysis of the voice control commands using the preset large language model, thereby understanding the control requirements expressed by the user in the voice. Furthermore, this invention can generate control tasks corresponding to the control requirements based on the control requirements and the scene environment data of the IoT scenario corresponding to the acquired voice control commands. The control tasks include a target control device and control sub-tasks for the target control device. This achieves the generation of control tasks by combining actual control requirements and actual scene environment data, and performs corresponding control on the target control device according to the control tasks, rather than controlling based on fixed trigger command words and fixed control strategies. Therefore, this invention improves the intelligence level of voice control. Attached Figure Description

[0049] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0050] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 This is a flowchart illustrating an embodiment of the air conditioning method of the present invention;

[0052] Figure 2 This is a flowchart illustrating another embodiment of the air conditioning method of the present invention;

[0053] Figure 3 This is a schematic flowchart of another embodiment of the air conditioning method of the present invention;

[0054] Figure 4 This is a schematic diagram of the module structure of the air conditioning device according to an embodiment of the present invention;

[0055] Figure 5 This is a schematic diagram of the hardware operating environment involved in an embodiment of the present invention.

[0056] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0057] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0058] With the development of smart home technology, users' demand for controlling various devices in their homes via voice commands is increasing. For example, voice control of air conditioning devices usually requires users to pre-define scene names in the corresponding application software (APP) of the air conditioning device and configure the control strategy and trigger command words corresponding to the scene name. After the user submits the above configuration in the application software, when the user's voice hits the trigger command word, the air conditioning device will execute the control strategy corresponding to the trigger command word to achieve the corresponding scene control.

[0059] However, this voice control method has the following problems:

[0060] 1. The implementation method is rigid, requiring pre-setting of scenarios, defining trigger command words and configuring control strategies; otherwise, it is impossible to enter the corresponding scenario and achieve control. 2. The triggering method is singular; triggering the corresponding scenario control must be done through a pre-set single trigger command. 3. The control strategy is fixed; it will only control according to the pre-set control strategy.

[0061] Therefore, current voice control methods cannot flexibly adjust control strategies, nor can they adapt to the actual environment; they can only be triggered by a single command word. Consequently, there is a technical problem with the low level of intelligence in the control of air conditioning equipment.

[0062] To overcome the aforementioned shortcomings, this invention provides an air conditioning method that receives voice control commands collected by an air conditioning device. It can determine the control requirements of the voice control commands through a preset large language model, thereby enabling semantic analysis of the voice control commands to understand the control needs expressed by the user in the voice. Furthermore, this invention can generate control tasks corresponding to the control requirements based on the control needs and the acquired scene environment data of the IoT scenario where the air conditioning device is located. The control tasks include a target control device and control sub-tasks for the target control device. This achieves the generation of control tasks by combining actual control needs and actual scene environment data, and then controls the target control device accordingly, rather than controlling it based on fixed trigger command words and fixed control strategies. Therefore, this invention improves the intelligence level of the air conditioning device.

[0063] Based on this, the present invention proposes an air conditioning method according to a first embodiment, please refer to... Figure 1 The air conditioning method includes steps S10 to S30:

[0064] Step S10: Receive the user's voice control command and determine the control requirements of the voice control command through a preset large language model;

[0065] It should be noted that voice control commands represent voice data collected by the air conditioning device, and these commands can be used to instruct the air conditioning device to perform corresponding controls. The voice data can be user voice commands, etc. In this embodiment, the executing entity can be a cloud server. The voice control commands collected by the voice acquisition device can be sent to the cloud server. The voice acquisition device is a device with voice acquisition capabilities, such as a voice-activated air conditioner, mobile phone, smart speaker, home control device, etc. This embodiment does not impose specific limitations on this.

[0066] A pre-defined large language model is used to determine the control requirements corresponding to voice control commands. This model can be used for semantic analysis to identify these requirements. Control requirements can describe the user's control intent, such as the user's desired operating status of the air conditioning equipment.

[0067] For example, a voice acquisition device can acquire voice control commands, a server can receive the voice control commands acquired by the voice acquisition device, and a cloud server can determine the control requirements of the voice control commands through a preset large language model.

[0068] In a feasible embodiment, step S10 further includes steps S11 to S13:

[0069] Step S11: Convert the voice data in the voice control command into text data;

[0070] Step S12: Determine the command type of the voice control command based on the text data;

[0071] It should be noted that text data is the written representation of speech data. Converting speech data into text data facilitates the subsequent determination of control requirements corresponding to voice control commands by a pre-set large language model. Air conditioning equipment can convert speech data into digital signals and transmit them to a cloud server. The cloud server can then use Speech-to-Text technology to convert the digital signals back into text data.

[0072] The command types can include direct control, entertainment needs, casual conversation, preset / planned commands, and others. Direct control commands are those explicitly requested by the user to control the air conditioning device at the current time, such as controlling the air conditioning device to operate at a specific temperature. This embodiment does not specifically limit this. Entertainment needs commands are those requesting entertainment content or services, such as playing relaxing music. Casual conversation commands are those expressing the user's desire for non-control tasks, such as casual conversation about the weather. This embodiment does not specifically limit this. Preset / planned commands are those indicating the user's desire to control the air conditioning device at a future time. For example, the text data for a preset / planned command might be: "Create a comfortable environment for a gathering of seven or eight people at 8 am tomorrow." This embodiment does not specifically limit this.

[0073] For example, the voice data in the voice control command is converted into text data, and the command type corresponding to the text data is determined.

[0074] In one embodiment, step S12 further includes step S121: inputting text data into a preset instruction classification model, classifying the text data through the instruction classification model, and obtaining the instruction type of the voice control instruction.

[0075] It should be noted that the preset instruction classification model can also be a large language model. The instruction classification model can be used to identify the instruction type corresponding to text data. For example, a cloud server can input text data into a preset instruction classification model, which can then classify the text data to obtain the instruction type. This embodiment determines the corresponding instruction type through a preset instruction classification model, thereby facilitating the subsequent determination of the control requirements of pre-planned voice control instructions by a preset large language model, and enabling corresponding control, thus improving the intelligence level of the air conditioning equipment.

[0076] Step S13: When the instruction type is a preset pre-planned type, input the text data into the preset large language model, perform semantic analysis on the text data through the preset large language model, and obtain the control requirements.

[0077] It should be noted that the preset large language model only determines the control requirements and corresponding control tasks for text data when the instruction type is a preset pre-planned type. Since the preset large language model requires some time to generate control tasks, it may be difficult to determine the corresponding control tasks for direct control type voice control instructions in a timely manner, potentially leading to control lag. However, preset language planning type voice control instructions represent control requirements for future time periods. Therefore, this embodiment can utilize the preset large language model to perform semantic analysis on the text data when the instruction type is a pre-planned type to obtain the control requirements, thus facilitating the subsequent determination of the corresponding control tasks without affecting the timely execution of the control tasks.

[0078] For example, when the instruction type is a preset pre-planned type, the text data is input into a preset large language model, and the text data is semantically analyzed by the preset large language model to analyze the user's control intention and obtain control requirements.

[0079] This embodiment utilizes the powerful semantic understanding capabilities of a pre-set large language model to perform semantic analysis on text data, thereby determining the user's actual control needs and improving the intelligence level of the air conditioning equipment.

[0080] Step S20: Based on the control requirements and the scene environment data of the IoT scene corresponding to the acquired voice control command, generate the control task corresponding to the control requirements. The control task includes the target control device and the control sub-task of the target control device.

[0081] It should be noted that the IoT scenario corresponding to the voice control command can be: the IoT scenario corresponding to the voice acquisition device that collects the voice control command, for example, the scenario corresponding to the IoT account to which the voice acquisition device belongs. For example, the IoT scenario can be a smart home scenario, etc., in which there may be multiple air conditioning devices, and each air conditioning device can be interconnected through IoT technology, etc., and all air conditioning devices in this IoT scenario belong to the same IoT account.

[0082] Scene environment data describes the environment of the IoT scenario in which the air conditioning device resides. The target control device is the air conditioning device within the IoT scenario that needs to be controlled. Control subtasks are the tasks that the target control device needs to perform. A control task can include multiple target control devices, each with its own corresponding control subtask. The target control devices in a control task are the devices present in the scene environment data. For example, a preset large language model can be invoked to generate control tasks based on control requirements and scene environment data.

[0083] Step S30: Control the target control device in the IoT scenario according to the control task.

[0084] It should be noted that each target control device can be controlled according to the control task to meet the control requirements. In this embodiment, a preset large language model can also be used to confirm whether the control task should be executed. If the control task is confirmed to be executed, the target control task is then controlled according to the control task. For example, the preset large language model can be used to determine whether the control task can meet the control requirements and whether the target control device can execute the control sub-task. If the control task meets the control requirements and the target control device can execute the control sub-task, the target control device in the IoT scenario is controlled according to the control task.

[0085] This invention can receive user voice control commands and determine the control requirements of the voice control commands through a preset large language model. This allows for semantic analysis of the voice control commands using the preset large language model, thereby understanding the control requirements expressed by the user in the voice. Furthermore, this invention can generate control tasks corresponding to the control requirements based on the control requirements and the scene environment data of the IoT scenario corresponding to the acquired voice control commands. The control tasks include a target control device and control sub-tasks for the target control device. This achieves the generation of control tasks by combining actual control requirements and actual scene environment data, and performs corresponding control on the target control device according to the control tasks, rather than controlling based on fixed trigger command words and fixed control strategies. Therefore, this invention improves the intelligence level of voice control.

[0086] Further, you can refer to Figure 2 In another feasible embodiment, step S20 further includes steps S21 to S22:

[0087] Step S21: Extract requirement keywords from control requirements through the preset large language model and the corresponding thought chain of the preset large language model, and obtain scene environment data of IoT scenario through the preset large language model.

[0088] It should be noted that multiple requirement keywords can be extracted from the control requirements, and these keywords can represent the user's actual needs. The thought chain can be pre-stored in a preset large language model; the thought chain (CoT) is a reasoning process. The preset large language model can progressively generate control tasks based on the thought chain.

[0089] For example, by using a pre-defined large language model's thought process chain, the following steps are executed sequentially: extracting key requirements from control needs; acquiring environmental data of the IoT scenario where the air conditioning device is located using the pre-defined large language model; identifying the target control device and its control sub-tasks. For instance, when the control requirement is "to create a comfortable environment for a dinner gathering of seven or eight people in the living room at 8 pm tonight, with special consideration for those with colds and those on their periods," the key requirements extracted using the pre-defined large language model could include "8 pm tonight, dinner gathering in the living room, seven or eight people, special requirements - one person with a cold and one person on their period."

[0090] In one feasible embodiment, the scene environment data includes space, air conditioning equipment within the space, and the device functions of the air conditioning equipment; step S21 further includes steps S211 to S213:

[0091] Step S211: Obtain the IoT account of the voice acquisition device that collects voice control commands;

[0092] Step S212: Using a preset large language model, call the device query component to find the space corresponding to the IoT account, the air conditioning device corresponding to each space, and the device function of each air conditioning device.

[0093] It should be noted that the IoT account refers to the IoT account corresponding to the IoT scenario where the voice acquisition device that collects voice control commands resides. The cloud server can record the IoT account of the voice acquisition device that collects voice control commands. The IoT account can be represented by an identifier such as an account ID (IDentity). The voice acquisition device can be an air conditioning device with voice acquisition function, or other devices with voice acquisition function, such as mobile phones, smart speakers, etc. This embodiment does not specifically limit this. Multiple air conditioning devices can exist under the same IoT account, such as air conditioners, air purifiers, etc. This embodiment does not specifically limit this, and multiple spaces may also exist under the same IoT account, each space having its own corresponding one or more air conditioning devices. The spaces under the IoT account can be living rooms, bedrooms, kitchens, balconies, etc. This embodiment does not specifically limit this. Device function refers to the functions possessed by the air conditioning device. For example, for an air conditioner, the device functions include cooling, heating, temperature regulation, and air blowing. For an air purifier, the device functions include sterilization and dust removal. This embodiment does not specifically limit this.

[0094] The device query component can be used to find the space corresponding to an IoT account, the air conditioning device corresponding to each space, and the device function of each air conditioning device. The device query component can store information such as the space corresponding to an IoT account, the air conditioning device, and its corresponding device function. This facilitates the search for scene environment data by calling the device query component through a preset large language model.

[0095] For example, the IoT account of the air conditioning device that collects voice control commands is obtained, and the device query component is called through a preset large language model to find all spaces corresponding to the IoT account, all air conditioning devices corresponding to each space, and the device functions of each air conditioning device.

[0096] Step S213: If there is no environmental control value in the control requirement keywords, call the environmental query component through the preset large language model to find the regional environmental parameters of the area where the air conditioning equipment is located, and determine that the scene environmental data includes the regional environmental parameters.

[0097] It should be noted that environmental control values ​​represent specific parameters of the controlled environment. For example, environmental control values ​​include controlling the temperature at N℃ and / or controlling the humidity at M. Environmental control values ​​are control parameters with definite numerical values, and the operation of air conditioning equipment can be directly controlled through environmental control values. Here, N and M represent actual values. For example, N can be a value of 26, 28, etc., and M can be a value of 55, 51, etc. This embodiment does not make specific limitations on this. For example, the voice control command given by the user might be "Set the temperature of the bedroom air conditioner to 26℃ at 9 pm today." Then, the corresponding control requirement keywords would include "set the temperature to 26℃." In this case, the control requirement keywords include environmental control values.

[0098] When the keywords for a control request do not include environmental control values, it indicates that the user's voice control command does not explicitly specify control parameters, but rather describes the control effect. For example, the user's command might be "Create a cool environment in the bedroom at 9 PM today," where the keywords for the control request do not include environmental control values. Since no specific control parameters are provided, it is necessary to query the regional environmental parameters for the area where the air conditioning unit is located. This allows for the subsequent determination of specific control parameters using a pre-set large-scale language model and regional environmental parameters, thus facilitating the fulfillment of the user's control needs.

[0099] The environmental query component can be used to find regional environmental parameters, which can be characterized by the environment of the area where the air conditioning equipment is located, such as temperature, humidity, and control quality. This embodiment does not make specific limitations on these parameters.

[0100] For example, if the environmental control value exists in the requirement keywords of the control requirement, it is not necessary to search for the regional environmental parameters. The scene environmental data may not include the regional environmental parameters. If the environmental control value does not exist in the requirement keywords of the control requirement, it is necessary to call the environmental query component through the preset large language model to find the regional environmental parameters of the area where the air conditioning device is located, and to determine that the scene environmental data includes the regional environmental parameters.

[0101] This embodiment uses a preset large language model to call the corresponding component to find scene environment data, which facilitates the subsequent determination of the corresponding control task based on the scene environment data, and enables the generation of control tasks in combination with the actual situation.

[0102] Step S22: Based on the requirement keywords, identify the target control device in the scene environment data and determine the control sub-task of the target control device.

[0103] It should be noted that the target control device is the air conditioning equipment that needs to operate to meet control requirements, while the control subtask is the task that the target control device needs to perform to meet those requirements. The target control device is the air conditioning equipment that exists in the scene environment data.

[0104] In this embodiment, the target control device and its control sub-tasks can be determined from the scene environment data by using the requirement keywords of the control requirements.

[0105] This embodiment generates control tasks step by step by pre-setting a large language model and corresponding thought chains, thereby enabling the pre-setting large language model to gradually think and generate control tasks, which facilitates the improvement of the accuracy of control task generation.

[0106] In a feasible embodiment, the control subtask includes an advance subtask and an agreed-upon subtask; step S22 further includes steps S221 to S223:

[0107] Step S221: Using a preset large language model, determine the target control space corresponding to the demand keywords in each space of the IoT account, and determine the target control device in each air conditioning device in the target control space.

[0108] It should be noted that the target control space is characterized as the space that needs to be controlled to meet the user's control requirements, and the target control device is the air conditioning equipment within the target control space. There can be multiple target control devices.

[0109] The pre-defined large language model can determine the target control space in the IoT account and the target control device within the target control space by controlling the corresponding demand keywords.

[0110] For example, a pre-set large language model can be used to determine the target control space corresponding to the required keywords and the target control device in the target control space.

[0111] For example, if the control requirements keywords are "8 pm tonight, living room dinner, seven or eight people, comfortable environment, special requirements - one person has a cold and one person has their period", the living room and dinner are mentioned in the requirements keywords. The living room is a location, and dinners are usually held in the living room. Therefore, the corresponding living room in the IoT account can be identified as the target control space. The requirements keywords mention the comfortable environment, one person has a cold, dinner, etc., so the air conditioner in the living room and the air purifier in the living room can be identified as target control devices.

[0112] Step S222: For each target control device, if the scene environment data does not include regional environment parameters, then based on the preset large language model, the device function of the target control device and the requirement keywords, generate the advance sub-tasks and agreed sub-tasks of the target control device.

[0113] Step S223: For each target control device, if the scene environment data includes regional environment parameters, then based on the preset large language model, the device function of the target control device, the requirement keywords, and the regional environment parameters, generate the advance sub-tasks and agreed sub-tasks of the target control device.

[0114] Among them, the advance subtask includes the advance start time and advance control parameters, and the agreed subtask includes the agreed start time and agreed control parameters, with the advance start time being earlier than the agreed start time.

[0115] It should be noted that each target control device has its own corresponding pre-defined subtasks and pre-defined subtasks. Pre-defined subtasks are the tasks that the target control device needs to perform within the user-specified time frame, while pre-defined subtasks are the tasks that the target control device needs to perform within the user-specified time frame. Pre-defined subtasks allow the environment that meets control requirements to be created in advance, while pre-defined subtasks ensure that the environment meets control requirements within the user-specified time frame. The user-specified time frame can also be determined from the requirement keywords of the control requirements.

[0116] The advance start time is the start time when the target controlled device executes an advance sub-task, while the agreed start time is the start time when the target controlled device executes an agreed sub-task. The agreed start time can be a user-specified time; for example, if the user indicates via voice control that it needs to be turned on at 8 PM tonight, then the agreed start time could be 8 PM tonight. Advance control parameters are the control parameters of the target controlled device before the agreed start time arrives; agreed control parameters are the control parameters of the target controlled device after the agreed start time. Advance control parameters and agreed control parameters can be the same or different. The advance start time is earlier than the agreed start time, thus facilitating the creation of a spatial environment that meets the user's control requirements in advance.

[0117] For example, for each target control device, when the scene environment data does not include regional environmental parameters, it indicates that environmental control parameters exist in the requirement keywords. Using a pre-set large language model, the device functions of the target control device, and the requirement keywords, the pre-control parameters, pre-control timing, agreed-upon control parameters, and agreed-upon control timing of the target control device are generated. The agreed-upon control parameters can include environmental control parameters. For instance, if the environmental control parameters include a temperature setting of 28°C for room temperature heating, and the target control device has a heating function, then the agreed-upon control parameters of the target control device can also include a heating mode and a temperature setting of 28°C. The pre-control parameters and pre-control timing can be predicted using a pre-set large language model, combined with the device functions and requirement keywords.

[0118] For each target control device, if the scene environment data includes regional environmental parameters, it indicates that environmental control parameters are not present in the control requirement keywords. Therefore, it is necessary to generate the target control device's advance control parameters, advance control timing, agreed-upon control parameters, and agreed-upon control timing based on a pre-defined large language model, regional environmental parameters, requirement keywords, and the device's functions. The pre-defined large language model can utilize regional environmental parameters and requirement keywords to determine control parameters that meet the control requirements, and combine this with the target control device's functions to determine the agreed-upon control parameters and / or advance control parameters. Furthermore, the pre-defined large language model can also be used to determine the corresponding advance start time and agreed-upon start time based on the requirement keywords.

[0119] If the target control device does not have a timer function, a subtask corresponding to the time when the target control device needs to be turned on can be issued by the cloud server to achieve the timed activation of the target control device.

[0120] This embodiment determines pre-defined and agreed-upon sub-tasks by pre-setting a large language model, requirement keywords, and the equipment functions of the target control device. This allows for the determination of the target control device's control sub-tasks based on its actual equipment functions and control requirements, thereby enabling flexible adjustment of the target control device's control sub-tasks. This improves the intelligence level of air conditioning equipment control.

[0121] Furthermore, refer to Figure 3 In another feasible embodiment, the air conditioning method further includes steps A10 to A20:

[0122] Step A10: Identify supplementary requirements for generating control tasks that are not included in the control requirements using a preset large language model;

[0123] Step A20: By calling the context component through the preset large language model, based on the collection time of the voice control command, the associated voice data is obtained in the first preset duration segment before the collection time and the second preset duration segment after the collection time. The keywords to be supplemented are obtained from the associated voice data and then supplemented into the required keywords.

[0124] It should be noted that the supplementary requirements are characterized as the requirements needed to generate the control task. For example, the supplementary requirements may be requirements for control space and / or control effects, etc. This embodiment does not specifically limit them.

[0125] For example, in the process of generating control tasks using a pre-set large language model, if there are control sub-tasks of the target control space, target control equipment, and / or target control equipment that cannot be determined, the pre-set large language model can be used to identify supplementary requirements needed to generate control tasks that are not included in the control requirements.

[0126] The context component can be used to store voice data collected by the air conditioning device. The collection time is the moment when the air conditioning device collects the voice control command. The first preset duration segment and the second preset duration segment can be set based on actual conditions, and this embodiment does not impose specific limitations on them. The durations of the first preset duration segment and the second preset duration segment can be the same or different. The first preset duration segment is the period before the collection time, and the second preset duration segment is the period after the collection time. The first preset duration segment can be the duration within one hour before the collection time, and the second preset duration segment can be the duration between the collection time and the current time.

[0127] The associated voice data consists of voice data collected by the air conditioning device within a first preset duration period and a second preset duration period. The keywords to be supplemented are the keywords corresponding to the desired supplementation requirements within the associated voice data. These keywords can be extracted from the associated voice data using a preset large language model.

[0128] There may be additional requirements in the associated speech data. Therefore, by searching the associated speech data, we can obtain the keywords for the additional requirements and add them to the requirement keywords of the control requirements. This allows the requirement keywords to include the additional keywords, facilitating the generation of control tasks by the pre-set large language model based on the requirement keywords that include the additional keywords.

[0129] For example, consider a control requirement: "Please create a comfortable environment at 8 PM tonight." This requirement doesn't specify the location and / or purpose of creating this comfortable environment. Therefore, the preset large language model might not be able to determine the target control space. In this case, the preset large language model can invoke a context component to search for associated speech data within the first and second preset durations of the data collection period. The preset large language model can then extract keywords to be added to the requirement keywords from this associated speech data, allowing it to further determine the corresponding target control space. For instance, the keyword to be added might be "dinner party," which can be added to the requirement keywords. Based on this keyword, the preset large language model can determine the target control space as the living room. The above example is just one illustration for ease of understanding; this embodiment does not specifically limit the content of the keywords to be added, the requirements to be added, etc.

[0130] This embodiment identifies the supplementary requirements by using a preset large language model and supplements them, thereby improving the success rate of generating control tasks through the preset large language model.

[0131] In a feasible embodiment, step S30 further includes steps S31 to S32:

[0132] Step S31: Determine the expected execution effect of the control task by using a pre-set large language model;

[0133] It should be noted that the expected execution effect is the effect of the control task execution evaluated by the preset large language model. The expected execution effect can be used to determine whether the control task meets the user's control needs.

[0134] For example, by using a pre-set large language model, the execution effect of the pre-defined sub-tasks and agreed sub-tasks of each target control device can be evaluated to obtain the expected execution effect of the control task.

[0135] Step S32: Verify the expected execution effect. If the verification passes, send the control sub-task of the target control device to the target control device via the Internet of Things protocol so that the target control device can execute the control sub-task.

[0136] It should be noted that verifying the expected execution effect is to determine whether the control task can meet the control requirements and whether the control task can be executed in this IoT scenario. This IoT scenario refers to the IoT scenario in which the air conditioning device that collects voice control commands is located.

[0137] The successful verification of the expected execution effect indicates that the control task can meet the control requirements and can be executed in this IoT scenario. The IoT protocol can be the communication protocol of the Internet of Things (IoT), through which control subtasks of the target control device can be sent to the target control device.

[0138] For example, the cloud server can use a pre-set large language model to verify the expected execution effect. If the verification is successful, the cloud server can send the control sub-task of the target control device to the target control device through the Internet of Things protocol. For example, it can send it to the controller of the target control device so that the controller of the target control device can execute the control sub-task.

[0139] This embodiment evaluates the expected execution effect of the control task through a preset large language model. If the expected execution effect verification passes, for each target control device, the communication protocol of the target control device is sent to the target control device via the Internet of Things (IoT) communication protocol. This ensures the executability of the control task and guarantees that the control task can meet the control requirements.

[0140] In a feasible embodiment, step S32 further includes steps S321 to S324:

[0141] Step S321: Determine the effect requirements for expressing control needs through a preset large language model, and evaluate whether the expected execution effect meets the effect requirements;

[0142] Step S322: By using a preset large language model, determine the target device functions required for the target control device to perform control sub-tasks, and determine whether the device functions of the target control device include the target device functions.

[0143] It should be noted that the desired effect is represented by the control effect the user expects to achieve. Because the pre-defined large language model possesses powerful semantic understanding capabilities, it can be used to determine the desired effect expressed by the control requirement, and also to evaluate whether the expected execution effect meets the requirement. If the expected execution effect does not meet the requirement, it means that the expected execution effect does not achieve the user's desired control effect; if the expected execution effect meets the requirement, it means that the expected execution effect can achieve the user's desired control effect. The pre-defined large language model can be used to evaluate whether the expected execution effect meets the requirement.

[0144] The target device function refers to the device functions required for the target control device to execute the control sub-task. If the target control device does not have the target device function, it means that the target control device may not be able to execute the corresponding control sub-task. If the target control device has the target device function, it means that the target control device can execute the corresponding control sub-task.

[0145] For example, if there are multiple target control devices in the control task, for each target control device, the target device function required for the target control device to perform the control sub-task is determined by a preset large language model, and it is determined whether the device function of the target control device includes the target device function.

[0146] Step S323: If the expected execution effect meets the effect requirements and the device function of the target control device includes the target device function, then the verification of the expected execution effect is determined to be successful.

[0147] Step S324: If the expected execution effect does not meet the effect requirements, and / or the device function of the target control device does not include the target device function, then it is determined that the verification of the expected execution effect fails, and the execution step is returned: Determine the control requirements from the text data through the preset large language model until the verification of the expected execution effect passes.

[0148] It should be noted that if the expected execution effect meets the effect requirements, and for each target control device, the device function of the target control device includes the target device function, then it means that the control task can achieve the control effect expected by the user, and each target control device can execute the corresponding control sub-task. It can be determined that the verification of the expected execution effect has passed.

[0149] If the expected execution effect does not meet the effect requirements, it means that the control task does not achieve the control effect expected by the user. For any target control device, if there is a target control device whose device function does not include the target device function, it means that if the control task is issued, there will be a situation where the target control device cannot execute the corresponding control sub-task. Even if the preset large language model evaluates that the expected execution effect meets the effect requirements, if the target control device cannot execute the corresponding control sub-task, it will not actually meet the user's control requirements and will not achieve the control effect expected by the user.

[0150] For example, if the expected execution effect does not meet the effect requirements, and / or the device functions of the target control device do not include the target device functions, then the verification of the expected execution effect is determined to have failed, and the execution steps are returned: control requirements are determined from the text data using a preset large language model until the verification of the expected execution effect passes. That is, the control requirements are redefined using the preset large language model, and the control task is regenerated until the verification of the expected execution effect passes.

[0151] This embodiment verifies the expected execution effect, thereby ensuring the executability of the control task while flexibly adjusting the control task, and ensuring that the control task can meet the user's control needs.

[0152] To better understand this embodiment, please refer to the following example to illustrate the process of generating the control task in this embodiment:

[0153] Using the voice data from the collected voice control commands as an example, "Seven or eight friends are coming over for dinner tonight at 8 pm. One of them has a cold, and another is on their period and needs special care. Please prepare a comfortable living room environment for the gathering," the generation process of the control task is explained. To better understand the generation process, the thought process of the pre-set large language model is first explained.

[0154] In this embodiment, the thought process can include, in sequence: keyword focusing, execution capability retrieval, environment retrieval, execution strategy thinking, execution strategy planning, expectation verification, and result output.

[0155] The process is as follows: Keyword Focusing extracts keywords related to control requirements; Execution Capability Retrieval retrieves the space of the IoT scenario where the air conditioning device is located, the air conditioning device within that space, and its functions; Environmental Retrieval retrieves environmental parameters for the area where the air conditioning device is located; Execution Strategy Thinking represents the thought process for generating control tasks, such as determining the target control space, target control device, and control requirements of the target control device; Execution Strategy Planning determines the control task and evaluates its expected execution effect; Expected Validation validates the expected execution effect of the control task; and Result Output outputs the control task through a pre-defined large language model. For pre-planned voice control commands, corresponding control tasks can be generated based on this thought process. For voice control commands where environmental control parameters are present in the control requirements, the environmental retrieval step in the thought process can be omitted.

[0156] The following example illustrates the process of generating control tasks using a pre-set large language model, based on the thought chain of "Seven or eight friends are coming over for dinner tonight at 8 pm, one of whom has a cold and another is on her period and needs special care. Please help me prepare a comfortable living room environment for the gathering" and the aforementioned pre-set large language model:

[0157] I. Key words to focus on: 8 pm tonight, living room dinner, seven or eight people, special needs: one person has a cold and one person is on their period;

[0158] II. Execution Capability Retrieval: Invoke the device query component to query the corresponding scene environment data; the device query component may include an account information query component, a home device list component, and a device function information component;

[0159] Call the account information query component: retrieve the space corresponding to the IoT account, such as the living room, bedroom, balcony, etc.

[0160] Call the home device list component: retrieve the air conditioning devices for each space, such as living room air conditioner 1, living room air purifier 1, bedroom air conditioner 1, etc.

[0161] Device function information component: Retrieves the device functions of each air conditioning unit, such as the living room air conditioner: {on / off, mode, temperature, fan speed, no-wind function, fresh air intake, timed tasks}. …… Living room air purifier {on / off switch, dust removal, sterilization, timed tasks} …… Bedroom air conditioner {on / off, mode} ……}wait.

[0162] III. Environmental Information Retrieval: Call the environmental query component to query regional environmental parameters. The environmental query component can include a time query component and a weather query component.

[0163] Call the time query component: Get the current time, for example, November 13, 2024, 14:10:00;

[0164] Call the weather query component: Get the current temperature of 29℃, air humidity of comfortable, and air quality of excellent.

[0165] IV. Implementation Strategy Considerations:

[0166] (1) Determine the target control space:

[0167] Reasoning: The key requirement mentions "living room", and the main dining activities take place in the living room.

[0168] Conclusion: Target space to be controlled: living room.

[0169] (2) Determine the target control equipment:

[0170] Reasoning: The keywords in the requirement mention that there are seven or eight friends having dinner together, and it is necessary to create a comfortable environment, especially taking care of friends who have colds or are on their periods.

[0171] Result of consideration: Target control devices: living room air conditioner, living room air purifier.

[0172] (3) Operation method:

[0173] Reasoning: All target control devices support scheduled tasks. The current time is 14:10 on November 13, 2024. The dinner party is scheduled for 8 pm tonight, so preparations need to be made in advance.

[0174] Conclusion: Operation method: Scheduled tasks, including pre-arranged subtasks and pre-defined subtasks.

[0175] (4) Operational requirements for the target control device being the living room air conditioner:

[0176] Reasoning: The keywords in the requirements mention a large number of people who need to maintain air circulation and a comfortable temperature; the weather temperature is high, so cooling is needed; the outdoor air quality is good, so the fresh air function can be used; one person has a cold and needs to avoid direct airflow; one person is menstruating and needs to avoid direct cold airflow.

[0177] Conclusion: The operation includes cooling and temperature control, fresh air supply, and draft-free system; among them, fresh air supply can replace the air for gatherings and when you have a cold; draft-free system can prevent cold air from blowing directly on you.

[0178] (5) The target control device is an operating requirement for living room air purification:

[0179] Reasoning: There are a large number of people, so it's necessary to keep the air fresh. Some people have colds, so air disinfection is needed.

[0180] Conclusion: Dust removal in advance, and disinfection during gatherings.

[0181] V. Implementation Strategy Planning:

[0182] Target control device: living room air conditioner.

[0183] Pre-programmed sub-task: Schedule to turn on the unit at 19:30 on November 13, 2024, in cooling mode, at 22°C, with maximum fan speed. Expected effect: To create a comfortable temperature upon entering the room.

[0184] Sub-task agreed upon: Start the machine at 20:00 on November 13, 2024, in cooling mode, at 24°C, with maximum fan speed, no wind, and fresh air intake; Expected effect: Create a comfortable dining environment, while also taking into account the needs of those on their period to avoid direct cold airflow, and to accommodate those with colds by providing fresh air to the room.

[0185] Target control device: Living room air purifier.

[0186] Sub-tasks to be scheduled ahead of time: Start-up and dust removal on November 13, 2024 at 19:30. Expected result: To create a clean air environment ahead of schedule.

[0187] Sub-task: Scheduled task to start up and disinfect on November 13, 2024 at 20:00. Expected effect: To disinfect the air, taking into account the possibility of someone having a cold.

[0188] VI. Expected Validation: Using a pre-set large language model, the expected performance of each target control device is validated. This verifies whether the expected performance of each target control device can achieve the user's desired control effect, and whether the target control device can execute the corresponding pre-defined subtasks and agreed-upon subtasks. Once the expected performance validation of each target control device passes, the execution result is output.

[0189] VII. Output Results: Output control tasks through the preset large language model: advance subtasks and agreed subtasks for the living room air conditioner, and advance subtasks and agreed subtasks for the living room air purifier.

[0190] After the preset large language model outputs control tasks, the cloud server can use the Internet of Things communication protocol to send the control sub-tasks of the target control devices to each target control device for execution.

[0191] This invention also provides an air conditioning device 40, see reference. Figure 4 The air conditioning unit 40 includes:

[0192] The requirement determination module 10 is used to receive the user's voice control command and determine the control requirements of the voice control command through a preset large language model.

[0193] The task generation module 20 is used to generate control tasks corresponding to control requirements based on control requirements and scene environment data of IoT scenarios corresponding to the acquired voice control commands. The control tasks include target control devices and control sub-tasks of the target control devices.

[0194] The execution module 30 is used to control the target control device in the Internet of Things scenario according to the control task.

[0195] The air conditioning device provided by this invention, employing the air conditioning method described in the above embodiments, can solve the technical problem of low intelligence level in voice control. Compared with the prior art, the beneficial effects of the air conditioning device provided by the embodiments of this invention are the same as those of the air conditioning method provided in the above embodiments, and other technical features in the air conditioning device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0196] This invention provides an air conditioning device, including a main body, an operating module disposed within the main body, and an air conditioning unit; the air conditioning unit includes a memory, a processor, and an air conditioning program stored in the memory and executable on the processor. When the air conditioning program is executed by the processor, it enables at least one processor to execute the air conditioning method described in the above embodiments.

[0197] The following is for reference. Figure 5 It shows a schematic diagram of an air conditioning device suitable for implementing embodiments of the present disclosure. Figure 5 The structure of the air conditioning device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments disclosed herein.

[0198] like Figure 5As shown, the air conditioning device may include a processor 101, such as a CPU, a communication bus 102, a user interface 103, a network interface 104, and a memory 105. The communication bus 102 is used to enable communication between these components. The user interface 103 may include a display screen and an input unit such as a keyboard; optionally, the user interface 103 may also include a standard wired interface or a wireless interface. The network interface 104 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 105 may be high-speed RAM or non-volatile memory, such as a disk drive. Optionally, the memory 105 may also be a storage device independent of the aforementioned processor 101.

[0199] Those skilled in the art will understand that Figure 5 The air conditioning unit structure shown does not constitute a limitation on the air conditioning unit and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0200] like Figure 5 As shown, the memory 105, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and an air conditioning program.

[0201] exist Figure 5 In the air conditioning device shown, the network interface 104 is mainly used to connect to the backend server and communicate data with the backend server; the user interface 103 is mainly used to connect to the client and communicate data with the client; and the processor 101 can be used to call the air conditioning program stored in the memory 105 to execute the steps of the air conditioning method.

[0202] The air conditioning device provided by this invention, employing the air conditioning method described in the above embodiments, can solve the technical problem of low intelligence level in voice control. Compared with the prior art, the beneficial effects of the air conditioning device provided by this invention are the same as those of the air conditioning method provided in the above embodiments, and other technical features of this air conditioning device are the same as those disclosed in the method of the previous embodiment, and will not be repeated here.

[0203] It should be understood that various parts of this disclosure can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0204] The above are merely specific embodiments 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.

[0205] This invention provides a computer-readable storage medium including computer-readable program instructions stored thereon, which are used to execute the air conditioning method in Embodiment 1 above.

[0206] The computer-readable storage medium provided in this embodiment of the invention may be, for example, a USB flash drive, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to, electrical connections including one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0207] The aforementioned computer-readable storage medium may be included in the air conditioning unit; or it may exist independently and not assembled into the air conditioning unit.

[0208] The aforementioned computer-readable storage medium carries one or more programs. When the one or more programs are executed by the air conditioning device, the air conditioning device: receives a user's voice control command and determines the control requirements of the voice control command through a preset large language model; generates a control task corresponding to the control requirements based on the control requirements and the scene environment data of the IoT scene corresponding to the acquired voice control command, wherein the control task includes a target control device and a control sub-task of the target control device; and controls the target control device of the IoT scene according to the control task.

[0209] Computer program code for performing the operations of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0210] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0211] The modules described in the embodiments of this disclosure can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0212] The readable storage medium provided by this invention is a computer-readable storage medium that stores computer-readable program instructions for executing the above-described air conditioning method, thereby solving the technical problem of low intelligence level in voice control. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in the embodiments of this invention are the same as the beneficial effects of the air conditioning method provided in the above-described embodiments, and will not be repeated here.

[0213] This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the air conditioning method described above.

[0214] The computer program product provided by this invention can solve the technical problem of low intelligence level in voice control. Compared with the prior art, the beneficial effects of the computer program product provided in the embodiments of this invention are the same as the beneficial effects of the air conditioning method provided in the above embodiments, and will not be repeated here.

[0215] The above are merely preferred embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent scope of the present invention.

Claims

1. An air conditioning method, characterized in that, The air conditioning method includes: Receive user's voice control commands and determine the control requirements of the voice control commands through a preset large language model; Based on the control requirements and the scene environment data of the IoT scenario corresponding to the acquired voice control command, a control task corresponding to the control requirements is generated, wherein the control task includes a target control device and a control sub-task of the target control device. Control the target control device in the IoT scenario according to the control task.

2. The air conditioning method as described in claim 1, characterized in that, The step of determining the control requirements of the voice control command through a preset large language model includes: Convert the voice data in the voice control commands into text data; Based on the text data, determine the instruction type of the voice control command; When the instruction type is a preset pre-planned type, the text data is input into a preset large language model, and the text data is semantically analyzed by the preset large language model to obtain the control requirements.

3. The air conditioning method as described in claim 2, characterized in that, The step of determining the instruction type of the voice control instruction based on the text data includes: The text data is input into a preset instruction classification model, and the text data is classified by the instruction classification model to obtain the instruction type of the voice control instruction.

4. The air conditioning method as described in claim 1, characterized in that, The step of generating the control task corresponding to the control requirement based on the control requirement and the scene environment data of the IoT scene corresponding to the acquired voice control command includes: By using a pre-defined large language model and the corresponding thought chain, the requirement keywords are extracted from the control requirements, and the scene environment data of the Internet of Things scenario is obtained through the pre-defined large language model. Based on the aforementioned requirement keywords, the target control device is identified in the scenario environment data, and the control sub-task of the target control device is determined.

5. The air conditioning method as described in claim 4, characterized in that, The scene environment data includes the space, the air conditioning equipment within the space, and the equipment functions of the air conditioning equipment. The step of obtaining the scene environment data of the IoT scenario through the preset large language model includes: Obtain the IoT account of the voice acquisition device that collects the voice control commands; Using the preset large language model, the device query component is invoked to find the space corresponding to the IoT account, the air conditioning device corresponding to each space, and the device function of each air conditioning device. If no environmental control value is found in the required keywords of the control requirements, the environmental query component is invoked through a preset large language model to find the regional environmental parameters of the area where the air conditioning device is located, and it is determined that the scene environmental data includes the regional environmental parameters.

6. The air conditioning method as described in claim 4, characterized in that, The control subtasks include advance subtasks and agreed-upon subtasks; The steps of determining the target control device from the scene environment data based on the required keywords, and determining the control sub-task of the target control device, include: By using a pre-set large language model, the target control space corresponding to the required keywords is determined in each space of the IoT account, and the target control device is determined in each air conditioning device in the target control space. For each target control device, if the scene environment data does not include regional environment parameters, then based on the preset large language model, the device function of the target control device, and the requirement keywords, the advance sub-tasks and agreed sub-tasks of the target control device are generated. For each target control device, if the scene environment data includes regional environment parameters, then based on the preset large language model, the device function of the target control device, the requirement keywords, and the regional environment parameters, the advance sub-tasks and agreed sub-tasks of the target control device are generated. The advance subtask includes an advance start time and advance control parameters, while the agreed subtask includes an agreed start time and agreed control parameters, with the advance start time being earlier than the agreed start time.

7. The air conditioning method according to any one of claims 1-6, characterized in that, The method includes: The system identifies additional requirements for generating control tasks that are not included in the control requirements by using a pre-defined large language model. By calling the context component through the preset large language model, based on the acquisition time of the voice control command, related voice data is obtained in the first preset duration segment before the acquisition time and the second preset duration segment after the acquisition time. The keywords to be supplemented are obtained from the related voice data and the keywords to be supplemented are added to the required keywords.

8. The air conditioning method as described in claim 1, characterized in that, The step of controlling the target control device in the IoT scenario according to the control task includes: The expected execution effect of the control task is determined by pre-setting a large language model; The expected execution effect is verified. If the verification passes, the control subtask of the target control device is sent to the target control device via the Internet of Things protocol so that the target control device can execute the control subtask.

9. The air conditioning method as described in claim 8, characterized in that, The steps for verifying the expected execution effect include: By using a pre-defined large language model, the desired effect of the control requirement expression is determined, and the expected execution effect is evaluated to determine whether the desired effect is met. By using a pre-defined large language model, the target device functions required for the target control device to execute the control sub-task are determined, and it is determined whether the device functions of the target control device include the target device functions. If the expected execution effect meets the effect requirements, and the device function of the target control device includes the target device function, then the verification of the expected execution effect is determined to be successful. If the expected execution effect does not meet the effect requirements, and / or the device function of the target control device does not include the target device function, then it is determined that the verification of the expected execution effect fails, and the execution step is returned: determine the control requirements from the text data through the preset large language model until the verification of the expected execution effect passes.

10. An air conditioning device, characterized in that, The system includes a main body and an air conditioning device disposed within the main body; the air conditioning device includes a memory, a processor, and an air conditioning program stored in the memory and executable on the processor, wherein when the air conditioning program is executed by the processor, it performs the steps of the air conditioning method as described in any one of claims 1-9.

11. A storage medium, characterized in that, The storage medium is a computer-readable storage medium storing an air conditioning program that can run on a processor, the air conditioning program being invoked by the processor to implement the steps of the air conditioning method according to any one of claims 1-9.

12. A program product, characterized in that, The program product is a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the air conditioning method as described in any one of claims 1-9.

Images