Multi-mode temperature control method and device, electronic equipment and storage medium

Abstract

The invention discloses a multi-mode temperature control method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring first multi-modal data influencing the regulation and control temperature of to-be-controlled equipment from at least two dimensions, and preprocessing the multi-modal data to obtain second multi-modal data; performing feature extraction on the second multi-modal data, endowing the second multi-modal data with a weight, and performing superposition processing on the second multi-modal data with the weight and a basic set temperature of the to-be-controlled equipment to obtain a target temperature of the to-be-controlled equipment; and inputting the second multi-modal data and the target temperature of the to-be-controlled equipment into a preset temperature prediction model for model training and iterative updating until the loss function converges, thereby obtaining a target temperature prediction model. By adopting the technical scheme of the embodiment of the invention, the temperature change trend of the to-be-controlled equipment can be sensed in advance, so that the temperature of the to-be-controlled equipment is adjusted in advance, and unnecessary energy waste is avoided.

Description

Technical Field

[0001] The present invention relates to the field of building control technology, and in particular, to a multi-modal temperature control method, device, electronic device and storage medium. Background Art

[0002] In modern building construction, the temperature control of building equipment usually adopts manual mode and automatic mode to control the building temperature.

[0003] Among them, the manual mode is usually adjusted manually by the user through the temperature controller panel, such as turning on / off the air conditioner, setting the air conditioner wind speed, air conditioner working mode, etc. After the air conditioner equipment starts to work, it usually continues to work in a specified state. The automatic mode is an intelligent operation mode of the air conditioner equipment, which can automatically adjust the operation state according to environmental changes and switch the working mode. For example, when the temperature difference is large, it will run at a high wind speed to quickly adjust the temperature; when the temperature difference is small, it will reduce the wind speed to maintain the temperature in a more gentle way, while reducing noise and energy consumption; when the temperature is higher than the set comfortable temperature range, the air conditioner automatically switches to the cooling mode to lower the indoor temperature; when the indoor temperature is lower than the set range, the air conditioner switches to the heating mode to raise the indoor temperature.

[0004] Among them, each temperature adjustment in the manual mode requires manual intervention. In the automatic mode, the personalized needs of users may not be met. For example, users may prefer a cooler or warmer environment; in addition, the temperature usually fluctuates within the range of the system default temperature (such as 26 degrees), which is set at the factory and users cannot change this default temperature through regular operations. Summary of the Invention

[0005] The present invention provides a multi-modal temperature control method, device, electronic device and storage medium to solve the problems that the personalized needs cannot be met and the temperature control cannot be adaptively realized in the automatic mode.

[0006] According to one aspect of the present invention, a multi-modal temperature control method is provided. The method includes: Obtaining first multi-modal data affecting the regulated temperature of the device to be controlled from at least two dimensions, and preprocessing the multi-modal data to obtain second multi-modal data; Extracting features from the second multi-modal data and assigning weights, and superimposing the second multi-modal data with assigned weights on the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled; Inputting the second multi-modal data and the target temperature of the device to be controlled into a preset temperature prediction model for model training and iterative update until the loss function converges to obtain a target temperature prediction model.

[0007] According to another aspect of the present invention, there is provided a multimodal temperature control device, the device comprising: A multimodal data acquisition module, configured to acquire first multimodal data affecting the regulated temperature of a device to be controlled from at least two dimensions, and preprocess the multimodal data to obtain second multimodal data; A multimodal data weight assignment module, configured to extract features from the second multimodal data and assign weights, and superimpose the second multimodal data with assigned weights on the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled; A target temperature prediction model training module, configured to input the second multimodal data and the target temperature of the device to be controlled into a preset temperature prediction model for model training and iterative update until the loss function converges, to obtain a target temperature prediction model.

[0008] According to another aspect of the present invention, there is provided an electronic device, the electronic device comprising: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the multimodal temperature control method according to any embodiment of the present invention.

[0009] According to another aspect of the present invention, there is provided a computer-readable storage medium storing computer instructions for causing a processor to implement the multimodal temperature control method according to any embodiment of the present invention when executed.

[0010] According to another aspect of the present invention, there is provided a computer program product comprising a computer program which, when executed by a processor, implements the multimodal temperature control method according to any embodiment of the present invention.

[0011] In the technical solution of the embodiment of the present invention, first multi-modal data affecting the regulated temperature of the device to be controlled is obtained from at least two dimensions, and the multi-modal data is preprocessed to obtain second multi-modal data; feature extraction is performed on the second multi-modal data and weights are assigned, and the second multi-modal data with weights assigned is superimposed on the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled; the second multi-modal data and the target temperature of the device to be controlled are input into a preset temperature prediction model for model training and iterative update until the loss function converges, and a target temperature prediction model is obtained. This solves the problems that personalized needs cannot be met in the automatic mode and temperature regulation cannot be adaptively achieved, and achieves the beneficial effect of pre-sensing the temperature change trend of the device to be controlled, adjusting the temperature of the device to be controlled in advance, and avoiding unnecessary energy waste.

[0012] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present invention, nor is it used to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0014] Figure 1 FIG. is a flowchart of a multi-modal temperature control method according to Embodiment 1 of the present invention.

[0015] Figure 2 FIG. is a flowchart of a multi-modal temperature control method according to Embodiment 2 of the present invention.

[0016] Figure 3 FIG. is a schematic structural diagram of a multi-modal temperature control device according to Embodiment 3 of the present invention.

[0017] Figure 4 FIG. is a schematic structural diagram of an electronic device according to Embodiment 4 of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0018] To enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0019] Among them, the acquisition, storage, use, and processing of data in the technical solution of this application all comply with the relevant provisions of laws and regulations. It should be noted that the terms "first", "second", "target", "original", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances, so that the embodiments of the present invention described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including", "etc.", and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0020] Embodiment 1: Figure 1 A flowchart of a multimodal temperature control method is provided for Embodiment 1 of the present invention. This embodiment is applicable to the situation of automatically adjusting the temperature of a device to be controlled according to multimodal indoor and outdoor environmental conditions. This method can be executed by a multimodal temperature control device, which can be implemented in the form of hardware and / or software, and can be configured in any electronic device with network communication functions. As Figure 1 shown, the method includes: S110. Obtain first multimodal data affecting the regulated temperature of the device to be controlled from at least two dimensions, and preprocess the multimodal data to obtain second multimodal data.

[0021] Among them, the device to be controlled may refer to a device that needs to be temperature-controlled in building equipment. For example, it may refer to an air-conditioning device.

[0022] In an embodiment of the present invention, first multi-modal data affecting the regulated temperature of a device to be controlled is obtained from at least two dimensions. The first multi-modal data affecting the regulated temperature of the device to be controlled may refer to that due to the change of the first multi-modal data, the demand for the indoor temperature changes, so the temperature of the device to be controlled needs to be regulated. The multi-modal data includes, but is not limited to, indoor and outdoor temperatures, indoor and outdoor humidities, the number of people in the room, the indoor space size, user preferences, and the outdoor temperature situation on the current day.

[0023] The preprocessing may refer to validating the effectiveness of the first multi-modal data and data augmentation; for example, determining whether the first multi-modal data is empty, determining whether the first multi-modal data exceeds the preset numerical range, converting the first multi-modal data type, and augmenting the first multi-modal data samples.

[0024] In an embodiment of the present invention, first multi-modal data affecting the regulated temperature of a device to be controlled is obtained from at least two dimensions, and the multi-modal data is preprocessed to obtain second multi-modal data; and the second multi-modal data is saved in a CSV format file for future use.

[0025] S120. Extract features from the second multi-modal data and assign weights, and superimpose the second multi-modal data with assigned weights on the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled.

[0026] Among them, the feature extraction may refer to obtaining features from the second multi-modal data and performing normalization processing for subsequent weighted calculation. For example, extract features from the indoor and outdoor temperatures obtained over a period of time, obtain the change trend of the indoor and outdoor temperatures, and perform normalization processing to perform weighted processing on the indoor and outdoor temperatures. The weighted processing may refer to assigning weights to the second multi-modal data, and the weights can be assigned according to the importance of the second multi-modal data; for example, in summer, users prefer cold, so the weight of user preferences can be assigned relatively large to regulate the temperature according to user preferences during temperature regulation.

[0027] Among them, the basic set temperature may refer to the original set temperature of the device to be controlled. For example, the basic set temperature of an air-conditioning device is usually 26°C. In an embodiment of the present invention, on the basis of the basic set temperature, the second multi-modal data with assigned weights is superimposed on it to obtain the target temperature of the device to be controlled. For example, in winter, the outdoor temperature is lower than the basic set temperature, and the heating mode should be turned on to raise the indoor temperature to the set temperature; assign weights to the outdoor temperature, and add the outdoor temperature with assigned weights to the basic set temperature to obtain the target temperature of the device to be controlled.

[0028] The target temperature may refer to the temperature that needs to be achieved indoors under the influence of the second multi-modal data. In the embodiments of the present invention, weights are assigned to the second multi-media data, and the weighted second multi-modal data is superimposed on the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled.

[0029] S130: Input the second multi-modal data and the target temperature of the device to be controlled into a preset temperature prediction model for model training and iterative update until the loss function converges, and obtain the target temperature prediction model.

[0030] Among them, the preset temperature prediction model may refer to a temperature prediction model based on a long short-term memory network. By inputting the second multi-modal data, the weight value of the second multi-modal data, and the target temperature of the device to be controlled into the preset temperature prediction model for model training, a converged target temperature prediction model can be obtained. In the embodiments of the present invention, according to the multi-modal data affecting the indoor temperature, the temperature of the device to be controlled at the next moment is predicted through the target temperature prediction model to realize the automatic temperature control of the device to be controlled.

[0031] The embodiments of the present invention provide a multi-modal temperature control method. First, the first multi-modal data affecting the temperature regulation of the device to be controlled is obtained from at least two dimensions, and the multi-modal data is preprocessed to obtain the second multi-modal data; then, feature extraction is performed on the second multi-modal data and weights are assigned, and the weighted second multi-modal data is superimposed on the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled; finally, the second multi-modal data and the target temperature of the device to be controlled are input into a preset temperature prediction model for model training and iterative update until the loss function converges, and the target temperature prediction model is obtained. By adopting the technical solution of the embodiments of the present invention, the temperature change trend of the device to be controlled can be sensed in advance, so that the temperature of the device to be controlled can be adjusted in advance, avoiding unnecessary energy waste.

[0032] Embodiment 2: Figure 2 The figure is a flowchart of a multi-modal temperature control method provided by Embodiment 2 of the present invention. The embodiments of the present invention further optimize the foregoing embodiments on the basis of the above embodiments, and the embodiments of the present invention can be combined with various alternative solutions in one or more of the above embodiments. As Figure 2 shown, the method includes: S210: Obtain the first multi-modal data affecting the temperature regulation of the device to be controlled from at least two dimensions, and preprocess the multi-modal data to obtain the second multi-modal data.

[0033] Among them, in the embodiments of the present invention, first multi-modal data affecting the regulated temperature of the device to be controlled is obtained from at least two dimensions, and the multi-modal data is preprocessed to obtain second multi-modal data.

[0034] As an optional but non-limiting implementation, the obtaining of the first multi-modal data affecting the regulated temperature of the device to be controlled from at least two dimensions and the preprocessing of the multi-modal data to obtain second multi-modal data include, but are not limited to, steps A1 - A2: Step A1: Obtain the first multi-modal data affecting the regulated temperature of the device to be controlled from at least two dimensions; wherein, the first multi-modal data includes indoor and outdoor temperatures, indoor and outdoor humidities, the number of people indoors, the indoor space size, user preferences, and the outdoor temperature situation on the current day.

[0035] Step A2: Preprocess the first multi-modal data to obtain second multi-modal data, and save the second multi-modal data in a target file in a preset format; wherein, the preprocessing includes determining whether the multi-modal data is empty, determining whether the multi-modal data exceeds a preset numerical range, converting the multi-modal data type, and augmenting the multi-modal data samples.

[0036] Among them, the first multi-modal data affecting the regulated temperature of the device to be controlled is obtained from at least two dimensions; for example, a temperature sensor is used to collect the real-time indoor and outdoor temperatures, a humidity sensor is used to collect the indoor and outdoor humidity values, a people counting sensor is used to collect the number of people in the room, the room size, user preferences (cold, normal, hot), the lowest and highest outdoor temperatures on the current day, and the date are determined. Among them, the real-time indoor and outdoor temperatures and humidities are actively collected at 1-minute intervals, and the people counting sensor automatically collects data when someone passes by. The lowest and highest temperatures are sourced from online weather data.

[0037] After obtaining the first multi-modal data, the first multi-modal data is preprocessed to obtain second multi-modal data. For example, validity verification is performed on the first multi-modal data, and the validity verification includes determining whether the multi-modal data is empty, determining whether the multi-modal data exceeds a preset numerical range, and converting the multi-modal data type; wherein, if the multi-modal data is empty, it needs to be collected again; if the multi-modal data exceeds the preset numerical range, it is necessary to determine whether the device is damaged; the multi-modal data is converted to floating-point type or integer type. Additionally, in order to enrich the data samples, the obtained first multi-modal data is augmented using the linear interpolation method. After that, the obtained second multi-modal data is saved in a local CSV-format file for future use.

[0038] S220. Extract features from the second multimodal data and assign weights thereto, and superimpose the second multimodal data with assigned weights on the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled.

[0039] Among them, before extracting features from the second multimodal data, parameter initialization is also included; specifically, it includes: initializing the weights of the indoor space size, the number of people indoors, user preferences, outdoor temperature, and humidity influence factors, initializing the local historical data record buffer (for example, recording data for 24 hours), and setting a learning rate for periodically updating and adjusting the weights.

[0040] Extracting features from the second multimodal data and assigning weights thereto may refer to obtaining the second multimodal data from a CSV file and extracting features, and performing normalization processing on each obtained feature for subsequent weighted calculation. The calculation of the target temperature may refer to performing proportional weighting based on the current input feature values and weights, and superimposing it on the basic set temperature to obtain the target temperature.

[0041] As an optional but non-limiting implementation manner, extracting features from the second multimodal data and assigning weights thereto, and superimposing the second multimodal data with assigned weights on the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled includes, but is not limited to, steps B1 - B3: Step B1: Extract features from the outdoor temperature included in the second multimodal data and assign corresponding weights to the outdoor temperature.

[0042] Step B2: Determine the difference between the outdoor temperature and the basic set temperature of the device to be controlled.

[0043] Step B3: Determine the first target temperature of the device to be controlled based on the difference, the weight assigned to the outdoor temperature, the mode coefficient, and the basic device temperature.

[0044] Among them, taking the second multimodal data including the outdoor temperature as an example, weights are assigned to the outdoor temperature, and the target temperature is determined. Read the current outdoor temperature, the lowest temperature, the highest temperature, and the date of the day, and fit the working mode that the current air conditioner should turn on, heating, ventilation, or cooling. For example, in spring, when the outdoor temperature is lower than the basic set temperature, the heating mode should be turned on to raise the indoor temperature to the set temperature; when the outdoor temperature is higher than the basic set temperature, the ventilation mode should be turned on to introduce external heat to raise the indoor temperature to the set temperature. The lowest and highest temperatures are used to assist in judging the working mode of the air conditioner. For example, if the highest temperature is lower than the set temperature, the heating mode must be turned on.

[0045] Based on the differences between the air conditioner working mode, the outdoor real-time temperature and the current temperature, the proportional weighting of the outdoor temperature characteristics is carried out, that is, the target temperature is equal to the basic set temperature plus the characteristic weighting value of the difference between the outdoor real-time temperature and the current indoor real-time temperature, dynamically adjusting the target temperature, intelligently maintaining the indoor body feeling comfort, and improving the energy efficiency at the same time.

[0046] Optionally, the higher the outdoor temperature, the lower the target temperature; the lower the outdoor temperature, the higher the target temperature. Different air conditioner working modes, the temperature difference in the cooling mode is equal to the outdoor temperature minus the indoor temperature; the temperature difference in the heating mode is equal to the indoor temperature minus the outdoor temperature. By compensating the basic set temperature with the outdoor temperature, the temperature control system can more reasonably cope with different environmental conditions. The logic in the cooling / heating mode is symmetric but in the opposite direction, specifically expressed as: In the cooling mode, the calculation formula of the target temperature can be expressed as: Among them, the represents the target temperature after dynamic adjustment of the device to be controlled; represents the basic set temperature of the device to be controlled; represents the outdoor temperature; represents the weight value assigned to the outdoor temperature. When , the outdoor is hotter, and the target temperature needs to be further reduced to offset heat conduction. When , the outdoor is cooler, and the target temperature needs to be appropriately increased to avoid excessive cooling.

[0047] In the heating mode, the calculation formula of the target temperature can be expressed as: Among them, when , the outdoor is colder, and the target temperature needs to be further increased to compensate for heat loss. When , the outdoor is warmer, and the target temperature needs to be appropriately reduced to avoid excessive heating.

[0048] In summary, it can be determined that in the cooling mode, is negatively correlated with the outdoor temperature difference, and should resist the outdoor high temperature to enhance the cooling effect; in the heating mode, is positively correlated with the outdoor temperature difference, and should resist the outdoor low temperature to enhance the heating effect. Among them, the weight value determines the intensity of the influence of the outdoor temperature and needs to be calibrated through experiments or experience. The cooling and heating formulas are essentially the same logic with opposite signs, reflecting the duality of the mode. Therefore, the two formulas can be combined into a unified first target temperature calculation formula, and the heating mode or the cooling mode is distinguished by the mode coefficient. Among them, the calculation formula of the first target temperature is expressed as: Among them, the represents the first target temperature of the device to be controlled; represents the basic set temperature of the device to be controlled; represents the outdoor temperature; represents a mode coefficient. In the cooling mode, ; in the heating mode, .

[0049] For example, taking the basic set temperature as 24 °C and the outdoor temperature weight as 0.1 as an example, in the cooling mode, when the outdoor temperature is 32 °C, the obtained first target temperature is 23.2 °C; in the heating mode, when the outdoor temperature is 10 °C, the obtained first target temperature is 25.4 °C.

[0050] In the embodiments of the present invention, the outdoor temperature is used as an influencing factor affecting the temperature regulation of the device to be controlled, and the outdoor temperature is weighted to achieve the temperature regulation of the device to be controlled.

[0051] As an optional but non-limiting implementation manner, the feature extraction and weighting of the second multimodal data, and the superposition processing of the weighted second multimodal data and the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled further include, but are not limited to, steps C1 - C3: Step C1: Extract features from the number of people in the room and the indoor space size included in the second multimodal data, and assign corresponding weights to the number of people in the room and the indoor space size.

[0052] Step C2: Determine the second target temperature of the device to be controlled under different indoor number of people conditions according to the number of people in the room, the weight assigned to the number of people in the room, the current target temperature, and the first adjustment range; wherein, the current target temperature refers to the target temperature of the device to be controlled obtained in the previous stage, and the first adjustment range refers to the maximum temperature adjustment range under different indoor number of people conditions.

[0053] Step C3: Alternatively, determine the third target temperature of the device to be controlled under different indoor space size conditions according to the indoor space size, the indoor space size threshold for triggering temperature adjustment, the weight assigned to the indoor space size, the current target temperature, and the second adjustment range; wherein, the second adjustment range refers to the maximum temperature adjustment range under different indoor space size conditions.

[0054] Among them, after the temperature of the device to be controlled is regulated according to the outdoor temperature, the temperature can also be regulated according to the number of people in the room. Assign weights to the number of people in the room, the current target temperature, and the first adjustment range, and determine the second target temperature of the device to be controlled under different indoor number of people conditions; the second target temperature can be expressed as: Among them, It may refer to the second target temperature of the device to be controlled under different indoor occupancy conditions; It may refer to the target temperature of the device to be controlled obtained in the previous stage, and here it may refer to the first target temperature; Represents the indoor occupancy; Represents the weight assigned to the indoor occupancy; Represents the maximum adjustment range, through which overcooling or overheating can be avoided.

[0055] Optionally, the heat of human respiration has a certain impact on the temperature, but it is assumed here that when the indoor occupancy is less than 2 people, the impact on the temperature can be ignored; as the indoor occupancy begins to gradually increase, the impact of the indoor occupancy on the temperature will not exceed a threshold value, and it is assumed that the maximum adjustment range is 0.5 degrees. When there are more people, the metabolic heat production increases, and more cooling or weaker heating is required.

[0056] Among them, based on the indoor occupancy, the impact of the indoor space size on the temperature is further calculated. A weight is assigned to the indoor space size, and according to the indoor space size threshold for triggering temperature adjustment, the current target temperature, and the second adjustment range, the third target temperature of the device to be controlled under different indoor space size conditions is determined. The third target temperature can be expressed as: Among them, Represents the third target temperature of the device to be controlled under different indoor space size conditions; It may refer to the target temperature of the device to be controlled obtained in the previous stage, and here it may refer to the second target temperature; Represents the current indoor space size, Represents the indoor space size threshold for triggering temperature adjustment, Represents the space size adjustment interval, Represents the weight of the space size for each interval. When , the indoor temperature needs to be regulated; every time it exceeds , the temperature drops °C; by calculating the number of temperature adjustments to determine the final temperature value to be adjusted, and the maximum temperature adjustment value does not exceed .

[0057] For example, taking , , , as an example, the third target temperature is determined. If the current indoor space size , at this time , there is no need to adjust the temperature of the device to be controlled. At this time, the third target temperature is the target temperature obtained in the previous stage, that is, the second target temperature. If , , , that is, it is necessary to adjust by 0.15 . At this time, the third target temperature is . If , , , that is, it is necessary to adjust according to the maximum adjustment range. At this time, the third target temperature is .

[0058] In the embodiments of the present invention, the target temperature of the device to be controlled is adjusted according to the indoor space size, so that the indoor temperature is suitable for the corresponding space size, and the indoor is prevented from being too cold or too hot due to the excessive space.

[0059] As an optional but non-limiting implementation manner, the method for extracting features from the second multimodal data, assigning weights to the second multimodal data, and superimposing the weighted second multimodal data on the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled further includes steps D1-D2, including but not limited to: Step D1: Extract features from the user preferences included in the second multimodal data, and assign corresponding weights to the user preferences; wherein, the user preferences include preferring cold, preferring hot, and preferring moderate.

[0060] Step D2: Enumerate the user preferences, and determine the fourth target temperature of the device to be controlled under different user preference conditions according to the enumeration value, the weight assigned to the user preferences, and the current target temperature; wherein, the enumeration value for preferring hot is 1, the enumeration value for preferring cold is -1, and the enumeration value for preferring moderate is 0.

[0061] Among them, calculate the proportional weighted value of the user preferences. Here, the default maximum adjustment range for the preferring hot and preferring cold parameters is plus or minus 1 degree of the set temperature, and the specific values will be dynamically adjusted according to the weights. Assign weights to the user preferences, enumerate the user preferences, and determine the fourth target temperature of the device to be controlled under different user preference conditions. The fourth target temperature can be expressed as: Among them, represents the fourth target temperature of the device to be controlled under different user preference conditions, here refers to the third target temperature; represents the weight assigned to the user preferences; represents the user preference coefficient, .

[0062] Optionally, taking , For example, when the user prefers cold, the ; when the user prefers moderate, the ; when the user prefers cold, the .

[0063] In the embodiments of the present invention, by obtaining multi-modal data that affects the regulated temperature of the device to be controlled and performing a weighting process on the multi-modal data, when the multi-modal data changes, the temperature of the device to be controlled will trigger real-time calculation again to dynamically give the latest appropriate temperature for subsequent temperature adjustment.

[0064] Optionally, after each appropriate temperature is calculated, an error calculation is performed between it and the real-time temperature, which is used to evaluate the performance of the model and update the weights, and is recorded in the historical data buffer.

[0065] In an alternative solution of the embodiments of the present invention, the method further includes updating the basic set temperature and the weights of the multi-modal data. By obtaining the average value of the real-time temperature data of the device to be controlled within a recorded period of time (such as 24 hours) as the latest basic set temperature. According to the error value between the target temperature and the real-time temperature cached in the historical data buffer and the feature weights, the weights of the multi-modal data are updated; the weight update can be expressed as: Where The weight of the i-th multi-modal data after update, represents the weight of the i-th multi-modal data before update; represents the learning rate; represents the number of historical records participating in the extreme, such as the data recorded in 24 hours; represents the error of the k-th record data; represents the normalized eigenvalue of the k-th record, such as the indoor space size and the number of people in the room need to be normalized within a reasonable range.

[0066] In the embodiments of the present invention, the weights of the multi-modal data and the basic set temperature of the device to be controlled are updated irregularly to make the determined target temperature more suitable for the current indoor environmental conditions.

[0067] S230. Input the second multi-modal data and the target temperature of the device to be controlled into a preset temperature prediction model for model training and iterative update until the loss function converges to obtain a target temperature prediction model.

[0068] Wherein, after determining the target temperature of the device to be controlled, the target temperature and the second multi-modal data are input into a preset temperature prediction model for model training to obtain a converged target temperature prediction model.

[0069] As an optional but non-limiting implementation, inputting the second multimodal data and the target temperature of the device to be controlled into a preset temperature prediction model for model training and iterative update until the loss function converges to obtain the target temperature prediction model includes, but is not limited to, steps E1 - E3: Step E1: Construct a preset temperature prediction model based on a long short-term memory network; wherein, modify the output value of the fully connected layer of the preset temperature prediction model so that the number of output values is consistent with the prediction length of the sequence sample.

[0070] Step E2: Construct a time series sample based on the second multimodal data and the target temperature of the device to be controlled.

[0071] Step E3: Input the time series sample into the preset temperature prediction model for model training and iterative update until the loss function converges to obtain the target temperature prediction model.

[0072] Among them, read the CSV file, standardize the features of the collected multimodal data column by column to obtain the mean and standard deviation of the feature data, convert the data to data with a mean of 0 and a standard deviation of 1, eliminate the differences in dimension and numerical range between different features, make different features have the same scale, and improve the training effect and accuracy of the model. Secondly, read the time information (such as 2025-01-20 14:30:00) of each group (row) of data, decompose and extract it into hours and specific dates to help the model understand the time features and capture the time rules. Then splice the parsed time features with other numerical values as inputs to construct a time series sample for network model training.

[0073] Construct a preset temperature prediction model based on a long short-term memory network. The preset temperature prediction model is a multi-step prediction model. By modifying the output value of the last output layer of the network, the length of the predicted temperature of the device to be controlled is determined. For example, if the output value of the output layer is modified to 5, the predicted temperatures of the device to be controlled at the next 5 time points can be predicted. Optionally, the number of output values of the output layer needs to be consistent with the prediction length of the sequence sample, and its formula can be expressed as: Among them, represents the predicted temperature at the i-th future time point, represents the prediction length of the sequence sample.

[0074] Input the processed time series sample into the preset temperature prediction model for model training, and use the MSE loss function (average the errors of 5 prediction steps) for iterative training. The loss function can be expressed as: Among them, n represents the number of samples, and k represents the number of predicted time steps. represents the true value at the j-th time point of the i-th sample. represents the predicted value at the j-th time point of the i-th sample.

[0075] In an alternative solution of the embodiment of the present invention, after determining the target temperature prediction model, it is also necessary to verify the trained target temperature prediction model. After obtaining the trained target temperature prediction model, use the test set for verification. After multiple repeated verifications, a better network model is obtained. If the prediction effect is not good, modify the trained model parameters and perform iterative training again. After that, only need to input the current set of input features into the model, and the model will directly output the temperature prediction values for the next 5 time steps.

[0076] As shown in Table 1 and Table 2, assume a simple time series dataset, record the temperature once per minute, and use it to predict the temperature for the next 5 minutes. If the time series length T = 10, when the model predicts the temperature at the next time point, it will use the data of the previous 10 minutes. The model uses the data from time point 1 to 10 to predict the temperature at time points 11 - 15.

[0077] Table 1 Input time series samples Time point 1 2 3 4 5 6 7 8 9 10 Temperature 20 20.1 20.1 20.2 20.1 20.4 20.5 20.4 20.6 20.7 Table 2 Output time series samples Time point 11 12 13 14 15 Predicted temperature 20.8 20.7 20.9 20.9 21.0

[0078] After verifying the model, it is necessary to convert and deploy the model. For example, convert the trained network model to the ONNX format for deployment on the edge server platform of AIBox.

[0079] The embodiment of the present invention trains, iteratively updates, and verifies the preset temperature prediction model to determine the converged target temperature prediction model.

[0080] S240, Energy consumption prediction.

[0081] Among them, after determining the target temperature prediction model, use the target temperature prediction model for temperature prediction, and use the PID algorithm to adjust the temperature of the device to be controlled to the predicted temperature to achieve energy consumption savings. The PID algorithm operates according to the input deviation value according to the functional relationship of proportional P, integral I, and differential D, and the operation result is used to control the output. In the embodiment of the present invention, a preset energy consumption prediction model is constructed to adaptively adjust the temperature of the device to be controlled using the PID algorithm to save energy.

[0082] As an optional but non-limiting implementation, after obtaining the target temperature prediction model, the method further includes energy consumption prediction, specifically including but not limited to steps F1 - F3: Step F1: Use the target temperature prediction model to predict the temperature of the device to be controlled at the next moment, and obtain the predicted temperature.

[0083] Step F2: Input the predicted temperature and the real-time temperature at the current moment into a preset energy consumption prediction model to determine the minimum energy consumption and the predicted working state of the device to be controlled at the next moment.

[0084] Step F3: Perform a temperature control operation on the device to be controlled according to the predicted working state.

[0085] Taking the predicted temperature for the next 5 minutes as an example, use the target temperature prediction model to determine the predicted temperature within the next 5 minutes, and input the predicted temperature and the real-time temperature at the current moment into a preset energy consumption prediction model; where the predicted temperature is the target temperature of the device to be controlled. On the premise of fixing the air conditioner power and cooling capacity, use the shortest path algorithm to calculate the minimum energy consumption required to adjust the real-time temperature to the target temperature within 5 minutes and the working state (0 - 100) of the air conditioner equipment. After obtaining the results of the minimum energy consumption and the working state, assuming that the air conditioner can increase or decrease the temperature by 2 degrees within 5 minutes in the full-speed working state, then if the difference between the current real-time temperature and the target temperature is within 1 degree, the device to be controlled can first work at full speed for 2 minutes to keep the real-time temperature within the target temperature, and the rest of the time is in standby sleep; it can also work at 70% of the full speed (minimum energy consumption / total energy consumption of 5-minute work) for 5 minutes to achieve the purpose of energy conservation.

[0086] In the embodiment of the present invention, a target temperature prediction model is introduced to predict the temperature, enabling the system to perceive the temperature change trend in advance, so as to make adjustments in advance and avoid unnecessary energy waste; based on the preset energy consumption prediction model, intelligent control of the heating or cooling system is realized based on the temperature prediction result, optimizing energy use and achieving energy conservation and consumption reduction.

[0087] In the embodiment of the present invention, a multi-modal temperature control method is provided. Under the machine learning framework, the collected multi-modal data is integrated. By adaptively allocating proportional weights, updating the weights and performing weighted calculations, an appropriate temperature setting value is given; secondly, the collected multi-modal data is sent into a long short-term memory neural network model for temperature prediction, and the temperature value in the future time period can be obtained to perceive whether there is a deviation and make corresponding adjustment strategies in advance; finally, an energy consumption model is established based on the real-time temperature value and the temperature prediction value to adjust the real-time temperature within the predicted temperature range within the specified time, so that the required energy consumption is minimized to achieve the purpose of energy conservation.

[0088] Embodiment 3: Figure 3 The following is a schematic structural diagram of a multi-modal temperature control device provided in Embodiment 3 of the present invention. As Figure 3 shown, the device includes: A multi-modal data acquisition module 310, configured to acquire first multi-modal data affecting the temperature regulation of the device to be controlled from at least two dimensions, and preprocess the multi-modal data to obtain second multi-modal data; A multi-modal data weight assignment module 320, configured to extract features from the second multi-modal data and assign weights, and superimpose the second multi-modal data with the weights assigned thereon on the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled; A target temperature prediction model training module 330, configured to input the second multi-modal data and the target temperature of the device to be controlled into a preset temperature prediction model for model training and iterative update until the loss function converges, to obtain a target temperature prediction model.

[0089] Optionally, the multi-modal data acquisition module is specifically configured to: Acquire first multi-modal data affecting the temperature regulation of the device to be controlled from at least two dimensions; wherein, the first multi-modal data includes indoor and outdoor temperatures, indoor and outdoor humidities, the number of people indoors, the indoor space size, user preferences, and the outdoor temperature condition on the current day; Preprocess the first multi-modal data to obtain second multi-modal data, and save the second multi-modal data in a target file in a preset format; wherein, the preprocessing includes determining whether the multi-modal data is empty, determining whether the multi-modal data exceeds a preset numerical range, multi-modal data type conversion, and multi-modal data sample expansion.

[0090] Optionally, the multi-modal data weight assignment module is specifically configured to: Extract features from the outdoor temperature included in the second multi-modal data, and assign corresponding weights to the outdoor temperature; Determine the difference between the outdoor temperature and the basic set temperature of the device to be controlled; Determine the first target temperature of the device to be controlled according to the difference, the weight assigned to the outdoor temperature, the mode coefficient, and the basic device temperature; Wherein, the calculation formula of the first target temperature is expressed as: Wherein, the represents the first target temperature of the device to be controlled; represents the basic set temperature of the device to be controlled; represents the outdoor temperature; Represents the weight value assigned to the outdoor temperature; Represents the mode coefficient. In the cooling mode, ; in the heating mode, .

[0091] Optionally, the multi-modal data weight assignment module is further specifically configured to: Extract features from the indoor number of people and the indoor space size included in the second multi-modal data, and assign corresponding weight values to the indoor number of people and the indoor space size; Determine the second target temperature of the device to be controlled under different indoor number of people conditions based on the indoor number of people, the weight value assigned to the indoor number of people, the current target temperature, and the first adjustment range; wherein, the current target temperature refers to the target temperature of the device to be controlled obtained in the previous stage, and the first adjustment range refers to the maximum temperature adjustment range under different indoor number of people conditions; Or, determine the third target temperature of the device to be controlled under different indoor space size conditions based on the indoor space size, the indoor space size threshold for triggering temperature adjustment, the weight value assigned to the indoor space size, the current target temperature, and the second adjustment range; wherein, the second adjustment range refers to the maximum temperature adjustment range under different indoor space size conditions.

[0092] Optionally, the multi-modal data weight assignment module is further specifically configured to: Extract features from the user preferences included in the second multi-modal data, and assign corresponding weight values to the user preferences; wherein, the user preferences include preferring cold, preferring heat, and preferring moderate; Enumerate the user preferences, and determine the fourth target temperature of the device to be controlled under different user preference conditions based on the enumeration value, the weight value assigned to the user preferences, and the current target temperature; wherein, the enumeration value for preferring heat is 1, the enumeration value for preferring cold is -1, and the enumeration value for preferring moderate is 0.

[0093] Optionally, the target temperature prediction model training module is specifically configured to: Construct a preset temperature prediction model based on a long short-term memory network; wherein, modify the output value of the fully connected layer of the preset temperature prediction model so that the number of output values is consistent with the prediction length of the sequence sample; Construct a time series sample based on the second multi-modal data and the target temperature of the device to be controlled; Input the time series sample into the preset temperature prediction model for model training and iterative update until the loss function converges to obtain the target temperature prediction model.

[0094] Optionally, after obtaining the target temperature prediction model, the device further includes an energy consumption prediction module, specifically configured to: Use the target temperature prediction model to predict the temperature of the device to be controlled at the next moment, and obtain the predicted temperature; Input the predicted temperature and the real-time temperature at the current moment into a preset energy consumption prediction model to determine the minimum energy consumption and the predicted working state of the device to be controlled at the next moment; Execute a temperature control operation on the device to be controlled according to the predicted working state.

[0095] The multi-modal temperature control device provided in the embodiments of the present invention can execute the multi-modal temperature control method provided in any of the above embodiments of the present invention, and has the corresponding functions and beneficial effects of executing the multi-modal temperature control method. For the detailed process, refer to the relevant operations of the multi-modal temperature control method in the foregoing embodiments.

[0096] Embodiment 4: Figure 4 The structural schematic diagram of an electronic device 10 that can be used to implement the embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, personal digital processors, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0097] As Figure 4 shown, the electronic device 10 includes at least one processor 11, and a memory communicatively connected to at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc. Among them, the memory stores a computer program executable by at least one processor. The processor 11 can execute various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 into the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. The input / output (I / O) interface 15 is also connected to the bus 14.

[0098] Multiple components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disc, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0099] The processor 11 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the multi-modal temperature control method.

[0100] In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains program codes for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication unit 19, or installed from the storage unit 18, or installed from the ROM 12. When the computer program is executed by the processor 11, the above functions defined in the method of the embodiment of the present invention are executed.

[0101] In some embodiments, the multi-modal temperature control method can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the multi-modal temperature control method described above can be executed. Alternatively, in other embodiments, the processor 11 can be configured to execute the multi-modal temperature control method by any other suitable means (e.g., by means of firmware).

[0102] The various embodiments of the systems and techniques described above in this specification can be implemented in digital electronic circuitry, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on a chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be a special-purpose or a general-purpose programmable processor that receives data and instructions from, and transmits data and instructions to, a storage system, at least one input device, and at least one output device.

[0103] The computer programs for implementing the methods of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, a special purpose computer, or other programmable data processing apparatus, such that the computer programs, when executed by the processor, cause the functions / operations specified in the flowchart and / or block diagram to be implemented. The computer programs can be executed entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine or entirely on the remote machine or server.

[0104] In the context of the present invention, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer-readable storage medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, the computer-readable storage medium can be a machine-readable signal medium. More specific examples of the machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0105] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and a pointing device (e.g., a mouse or a trackball) through which the user can provide input to the electronic device. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, speech input, or tactile input).

[0106] The systems and techniques described herein can be implemented in a computing system including backend components (e.g., as a data server), or a computing system including middleware components (e.g., an application server), or a computing system including frontend components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system including any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected to each other by digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), blockchain network, and the Internet.

[0107] The computing system can include a client and a server. The client and the server are generally far from each other and usually interact through a communication network. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system and solves the defects of difficult management and weak business scalability existing in traditional physical hosts and VPS services.

[0108] It should be understood that various forms of the processes shown above can be used, with steps reordered, added, or deleted. For example, the steps recited in the present invention can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved, and no limitation is made herein.

[0109] The above specific embodiments do not constitute a limitation on the protection scope of the present invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A multimodal temperature control method, characterized in that, The method includes: Obtaining first multimodal data that affects the regulated temperature of the device to be controlled from at least two dimensions, and preprocessing the multimodal data to obtain second multimodal data; Performing feature extraction on the second multimodal data and assigning weights, and superimposing the weighted second multimodal data with the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled; Inputting the second multimodal data and the target temperature of the device to be controlled into a preset temperature prediction model for model training and iterative update until the loss function converges, to obtain a target temperature prediction model; Among them, the step of inputting the second multimodal data and the target temperature of the device to be controlled into a preset temperature prediction model for model training and iterative update until the loss function converges to obtain a target temperature prediction model includes: Constructing a preset temperature prediction model based on a long short-term memory network; among them, modifying the output value of the fully connected layer of the preset temperature prediction model so that the number of output values is consistent with the prediction length of the time series sample; constructing a time series sample based on the second multimodal data and the target temperature of the device to be controlled; inputting the time series sample into the preset temperature prediction model for model training and iterative update until the loss function converges to obtain a target temperature prediction model.

2. The method according to claim 1, wherein The step of obtaining first multimodal data that affects the regulated temperature of the device to be controlled from at least two dimensions and preprocessing the multimodal data to obtain second multimodal data includes: Obtaining first multimodal data that affects the regulated temperature of the device to be controlled from at least two dimensions; among them, the first multimodal data includes indoor and outdoor temperature, indoor and outdoor humidity, the number of people indoors, the indoor space size, user preferences, and the outdoor temperature situation on the current day; Preprocessing the first multimodal data to obtain second multimodal data, and saving the second multimodal data in a target file in a preset format; among them, the preprocessing includes determining whether the multimodal data is empty, determining whether the multimodal data exceeds a preset numerical range, multimodal data type conversion, and multimodal data sample expansion.

3. The method according to claim 1, characterized in that, The step of performing feature extraction on the second multimodal data and assigning weights, and superimposing the weighted second multimodal data with the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled includes: Performing feature extraction on the outdoor temperature included in the second multimodal data and assigning corresponding weights to the outdoor temperature; Determining the difference between the outdoor temperature and the basic set temperature of the device to be controlled; Determining the first target temperature of the device to be controlled based on the difference, the weight assigned to the outdoor temperature, the mode coefficient, and the basic device temperature; Among them, the calculation formula of the first target temperature is expressed as: Among them, the represents the first target temperature of the device to be controlled; represents the basic set temperature of the device to be controlled; represents the outdoor temperature; represents the weight value assigned to the outdoor temperature; represents the mode coefficient. In the cooling mode, ; in the heating mode, .

4. The method according to claim 1, wherein The step of performing feature extraction on the second multimodal data and assigning weights, and superimposing the weighted second multimodal data with the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled further includes: Extract features from the number of people indoors and the indoor space size included in the second multimodal data, and assign corresponding weights to the number of people indoors and the indoor space size; Determine the second target temperature of the device to be controlled under different indoor population conditions based on the number of people indoors, the weight assigned to the number of people indoors, the current target temperature, and the first adjustment range; wherein, the current target temperature refers to the target temperature of the device to be controlled obtained in the previous stage, and the first adjustment range refers to the maximum temperature adjustment range under different indoor population conditions; Alternatively, determine the third target temperature of the device to be controlled under different indoor space size conditions based on the indoor space size, the indoor space size threshold for triggering temperature adjustment, the weight assigned to the indoor space size, the current target temperature, and the second adjustment range; wherein, the second adjustment range refers to the maximum temperature adjustment range under different indoor space size conditions.

5. The method according to claim 1, wherein The feature extraction and weight assignment to the second multimodal data, and the superposition processing of the second multimodal data with the assigned weights and the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled further include: Extract features from the user preferences included in the second multimodal data, and assign corresponding weights to the user preferences; wherein, the user preferences include preferring cold, preferring hot, and preferring moderate. Enumerate the user preferences, and determine the fourth target temperature of the device to be controlled under different user preference conditions based on the enumeration value, the weight assigned to the user preferences, and the current target temperature; wherein, the enumeration value for preferring hot is 1, the enumeration value for preferring cold is -1, and the enumeration value for preferring moderate is 0.

6. The method according to claim 1, wherein After obtaining the target temperature prediction model, the method further includes energy consumption prediction, specifically including: Use the target temperature prediction model to predict the temperature of the device to be controlled at the next moment, and obtain the predicted temperature; Input the predicted temperature and the real-time temperature at the current moment into a preset energy consumption prediction model to determine the minimum energy consumption and the predicted working state of the device to be controlled at the next moment; Perform temperature control operations on the device to be controlled according to the predicted working state.

7. A multimodal temperature control device, characterized in that, The device includes: A multimodal data acquisition module, configured to acquire first multimodal data affecting the temperature regulation of the device to be controlled from at least two dimensions, and preprocess the multimodal data to obtain second multimodal data; A multimodal data weight assignment module, configured to extract features from the second multimodal data and assign weights, and perform superposition processing on the second multimodal data with the assigned weights and the basic set temperature of the device to be controlled to obtain the target temperature of the device to be controlled; A target temperature prediction model training module, configured to input the second multimodal data and the target temperature of the device to be controlled into a preset temperature prediction model for model training and iterative update until the loss function converges to obtain a target temperature prediction model; Wherein, the target temperature prediction model training module is specifically configured to: Construct a preset temperature prediction model based on a long short-term memory network; among them, modify the output value of the fully connected layer of the preset temperature prediction model so that the number of output values is consistent with the prediction length of the time series sample; construct a time series sample based on the second multimodal data and the target temperature of the device to be controlled; input the time series sample into the preset temperature prediction model for model training and iterative update until the loss function converges to obtain a target temperature prediction model.

8. An electronic device, characterized in that, The electronic device includes: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the multimodal temperature control method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing a processor to implement the multimodal temperature control method according to any one of claims 1-6 when executed.

Images