Control method and device of temperature adjusting device and electronic equipment

Abstract

The invention discloses a control method and device of a temperature adjusting device and electronic equipment. The method comprises the steps that at the initial working temperature of the temperature adjusting device, the state index of a target object is obtained, the target object is located in the working environment of the temperature adjusting device, and the state index is used for representing the comfortable state of the target object in the working environment; the state index is analyzed through a load prediction model, a load prediction result is obtained, the load prediction result is used for representing the load state of the temperature adjusting device at the initial working temperature in the future time period, and the load prediction model is obtained by training a deep learning model; based on the load prediction result, a control strategy of the temperature adjusting device in the future time period is determined; and according to the control strategy, the temperature of the temperature adjusting device is adjusted to the target working temperature from the initial working temperature. The technical problem that the control accuracy of the temperature adjusting device is low is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy and energy conservation, and in particular, to a control method, device and electronic device for a temperature regulation device. Background Art

[0002] Currently, in the design and operation of temperature regulation devices (such as air conditioners), they are often designed and controlled for a single scenario. For example, heating is the main function in winter and cooling is the main function in summer, ignoring the control of air conditioners in the case of large temperature fluctuations within a single day. Moreover, when designing and constructing air conditioners, they tend to be static and lack consideration of the dynamic change characteristics of air conditioner loads. This results in the inability to accurately adjust the control strategy, leading to problems such as excessive air conditioner operation costs and insufficient indoor environmental comfort. Therefore, there is still the technical problem of low accuracy in the control of temperature regulation devices.

[0003] For the above problems, no effective solution has been proposed yet. Summary of the Invention

[0004] Embodiments of the present invention provide a control method, device and electronic device for a temperature regulation device to at least solve the technical problem of low accuracy in the control of temperature regulation devices.

[0005] According to one aspect of the embodiments of the present invention, there is provided a control method for a temperature regulation device, including: obtaining a state index of a target object at an initial operating temperature of the temperature regulation device, where the target object is in the operating environment of the temperature regulation device, and the state index is used to represent the comfort state of the target object in the operating environment; analyzing the state index by using a load prediction model to obtain a load prediction result, where the load prediction result is used to represent the load state of the temperature regulation device at the initial operating temperature in a future time period, and the load prediction model is obtained by training a deep learning model; determining a control strategy for the temperature regulation device in the future time period based on the load prediction result, where the control strategy is used to represent the rule for adjusting the initial operating temperature; and adjusting the temperature of the temperature regulation device from the initial operating temperature to a target operating temperature according to the control strategy, where the load state of the temperature regulation device at the target operating temperature meets the load state requirement of the temperature regulation device.

[0006] Optionally, the method further includes: obtaining historical environmental data of the working environment, as well as historical power consumption load data and historical time data corresponding to the working environment, where the historical environmental data is used to represent the meteorological conditions of the temperature regulation device during a historical period, the historical power consumption load data is used to represent the power consumption load of the temperature regulation device during the historical period, and the historical time data is used to represent the time of the temperature regulation device during the historical period; analyzing the state index using a load prediction model to obtain a load prediction result, including: analyzing the state index, historical environmental data, historical power consumption load data, and historical time data using the load prediction model to obtain a load prediction result.

[0007] Optionally, the load prediction model includes an influencing factor attention module, a time series neural network module, and a time attention module. Among them, analyzing the state index, historical environmental data, historical power consumption load data, and historical time data using the load prediction model to obtain a load prediction result includes: extracting features of the state index, historical environmental data, historical power consumption load data, and historical time data using the influencing factor attention module to obtain target features; determining the attention weights corresponding to the target features; performing a fusion operation on the attention weights and the target features to obtain a fusion result; using the fusion result to adjust the attention weights to obtain an attention weight feature vector; analyzing the attention weight feature vector using the time series neural network module to obtain a key feature vector, where the key feature vector is used to represent the feature vector corresponding to the attention weight greater than the attention weight threshold in the attention weight feature vector; using the time attention module to analyze the key feature vector according to the time series to obtain a time attention weight feature vector, where the time attention weight feature vector is used to represent the importance degree of the time attention weight feature vector in the time series; performing a fusion operation on the key feature vector and the time attention weight feature vector, and obtaining a load prediction result through the load prediction model.

[0008] Optionally, obtaining the state index of the target object at the initial working temperature of the temperature regulation device includes: obtaining the state information of the target object and the attribute information in the working environment where the temperature regulation device is located, where the state information of the target object is used to represent the state related to the behavior of the target object in the working environment, and the attribute information is used to represent the environmental factors corresponding to the working environment; determining the state index of the target object based on the state information and the attribute information.

[0009] Optionally, based on the load prediction result, determine the control strategy of the temperature adjustment device in the future time period, including: in response to the load prediction result being less than or equal to the load prediction threshold, determine the initial working temperature as the target working temperature, and determine the control strategy as the first control strategy; in response to the load prediction result being greater than the load prediction threshold, adjust the initial working temperature to the target working temperature, and determine the control strategy as the second control strategy.

[0010] Optionally, according to the control strategy, adjust the temperature of the temperature adjustment device from the initial working temperature to the target working temperature, including: generate control prompt information according to the control strategy; based on the control prompt information, adjust the temperature of the temperature adjustment device to the target working temperature.

[0011] According to another aspect of the embodiments of the present invention, there is also provided a control device for a temperature adjustment device, including: an acquisition unit, configured to acquire the state index of the target object at the initial working temperature of the temperature adjustment device, where the target object is in the working environment of the temperature adjustment device, and the state index is used to represent the comfort state of the target object in the working environment; an analysis unit, configured to analyze the state index by using a load prediction model to obtain a load prediction result, where the load prediction result is used to represent the load state of the temperature adjustment device at the initial working temperature in the future time period, and the load prediction model is obtained by training a deep learning model; a determination unit, configured to determine the control strategy of the temperature adjustment device in the future time period based on the load prediction result, where the control strategy is used to represent the rule for adjusting the initial working temperature; an adjustment unit, configured to adjust the temperature of the temperature adjustment device from the initial working temperature to the target working temperature according to the control strategy, where the load state of the temperature adjustment device at the target working temperature meets the load state requirement of the temperature adjustment device.

[0012] According to another aspect of the embodiments of the present invention, there is also provided a computer-readable storage medium, where the computer-readable storage medium stores multiple instructions, and the instructions are suitable for being loaded and executed by a processor to perform any one of the above control methods of the temperature adjustment device.

[0013] According to another aspect of the embodiments of the present invention, there is also provided an electronic device, including a memory and a processor, where a computer program is stored in the memory, and the processor is configured to run the computer program to execute any one of the above control methods of the temperature adjustment device.

[0014] According to another aspect of the embodiments of the present invention, there is also provided a computer program product, including a computer program, where the computer program implements any one of the above control methods of the temperature adjustment device when being executed by a processor.

[0015] In an embodiment of the present invention, if it is necessary to adjust the temperature of the temperature adjustment device, the state index of the target object can be obtained at the initial operating temperature of the temperature adjustment device; the load prediction model can be used to analyze the state index to obtain a load prediction result; based on the load prediction result, a control strategy for the temperature adjustment device in the future time period can be determined; according to the control strategy, the temperature of the temperature adjustment device can be adjusted from the initial operating temperature to the target operating temperature. In this embodiment, by obtaining the state index of the target object, the comfort state of the target object can be determined; by analyzing the state index through the load prediction model, the accuracy of predicting the load state of the temperature adjustment device (such as an air conditioner) in the future time period can be improved. Without reducing the comfort state of the target object, the air conditioner temperature setting is dynamically adjusted according to the air conditioner load state, reducing unnecessary energy consumption, thereby achieving the technical effect of improving the accuracy of the control of the temperature adjustment device and solving the technical problem of low accuracy of the control of the temperature adjustment device. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings described herein are used to provide a further understanding of the present invention, and constitute a part of this application. The illustrative embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0017] Figure 1 is a flowchart of a control method for a temperature adjustment device according to an embodiment of the present invention;

[0018] Figure 2 is a flowchart of a method for predicting the dynamic adjustable potential of an air conditioner load in a commercial building considering human comfort according to an embodiment of the present invention;

[0019] Figure 3 is a schematic structural diagram of a converter-long short-term memory network model according to an embodiment of the present invention;

[0020] Figure 4 is a schematic structural diagram of a control device for a temperature adjustment device according to an embodiment of the present invention;

[0021] Figure 5 is a schematic diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] To enable those skilled in the art to better understand the solution of the present invention, the following will clearly and completely describe the technical solution in the embodiments of the present invention 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 the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0023] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order different from those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0024] Embodiment 1

[0025] According to an embodiment of the present invention, a control embodiment of a temperature adjustment device is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. And although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that here.

[0026] Figure 1 is a flowchart of a control method of a temperature adjustment device according to an embodiment of the present invention. As Figure 1 shown, the method includes the following steps:

[0027] Step S102, obtain the state index of the target object at the initial working temperature of the temperature adjustment device.

[0028] In the technical solution provided in step S102 of the embodiment of the present invention, the temperature adjustment device may be an air conditioner. The initial working temperature is the temperature set by the air conditioner before ensuring the load state. For example, in a commercial building, the initial working temperature may be a preset temperature. For example, in summer, it may be 24°C, and in winter, it may be 20°C. The target object may be a person in the working environment of the temperature adjustment device. For example, in a commercial building, it may be a person inside the commercial building. The state index can be used to represent the comfort state of the target object in the working environment, and may include factors such as the activities of people, the clothing of people, and the corresponding temperature and humidity of the working environment in the working environment of the temperature adjustment device, and can be determined by the Predicted Mean Vote (PMV) value.

[0029] It should be noted that the setting of the above initial working temperature value is only for illustrative purposes and is not specifically limited here.

[0030] In this embodiment, obtaining the state index of the target object at the initial working temperature of the temperature adjustment device can ensure that before temperature regulation, the comfort state of the target object (such as a person in a commercial building) in the current working environment is accurately evaluated and understood, which is crucial for achieving intelligent regulation that is both energy-saving and can maintain the comfort of the target object.

[0031] Optionally, the activities, clothing, etc. of the people in the working environment of the temperature adjustment device are monitored through a human activity detection sensor, a thermal imager, etc. at the initial working temperature of the temperature adjustment device. The temperature, humidity, etc. corresponding to the working environment are obtained through a temperature sensor, a humidity sensor. Thus, the state index can be comprehensively obtained.

[0032] It should be noted that the above process of determining the state index is only for illustrative purposes and is not specifically limited here. As long as it can be used to obtain parameter data related to the comfort state of the target object in the working environment, it is within the protection scope of the embodiment of the present invention.

[0033] Step S104, analyze the state index using a load prediction model to obtain a load prediction result.

[0034] In the technical solution provided in step S104 of the embodiment of the present invention, after obtaining the state index of the target object, the load prediction model can be used to analyze the state index to obtain the load prediction result. Among them, the load prediction result can be used to represent the load state of the temperature adjustment device at the initial working temperature in the future time period; the load prediction model can be obtained by training a deep learning model. The load prediction model can be a Transformer-Long Short-Term Memory Network (abbreviated as Transformer-LSTM) model. The load prediction result output by the load prediction model can also be called the predicted air-conditioning load result, which can be the load prediction value for a specific time period or the load prediction sequence for each time point within the entire prediction cycle. The load prediction result can reflect the load demand of the air conditioner in the commercial building at each prediction time period based on the current (initial) working temperature.

[0035] Optionally, the load prediction model can analyze the relationship between the state index and the air-conditioning load through Transformer, and analyze the time series characteristics in the state index through LSTM. The time series characteristics in the state index can be determined by the amount of human activity corresponding to human activities and the clothing coefficient corresponding to human clothing changing over time (such as different clothing on weekdays and rest days, in summer and winter). Thus, the air-conditioning load state can be dynamically predicted, and the predicted load state of the air conditioner in the future time period can be output.

[0036] Step S106, determine the control strategy of the temperature adjustment device in the future time period based on the load prediction result.

[0037] In the technical solution provided in step S106 of the embodiment of the present invention, after using the load prediction model to analyze the state index to obtain the load prediction result, the control strategy of the temperature adjustment device in the future time period can be determined based on the load prediction result. Among them, the control strategy can be used to represent the rule for adjusting the initial working temperature.

[0038] In this embodiment, the control strategy determined based on the load prediction result can dynamically adjust the temperature of the temperature adjustment device to adapt to the load state demand in the future time period, save energy and reduce consumption, and ensure the comfort of personnel at the same time.

[0039] Optionally, according to the load prediction result, the time points when the load state is at the peak and trough can be identified, and the change in the load state demand of the air conditioner in the future time period can be predicted. For example, in a commercial building, if the load prediction result shows that the air-conditioning load rises significantly in the afternoon of a certain working day, it is the result of the combined action of high temperature weather and the peak of the number of people in the commercial building.

[0040] Optionally, in a commercial building, during peak load periods, the peak load can be mitigated and the air-conditioning load reduced by pre-cooling or pre-heating the building in advance, adjusting the air-conditioning temperature, or utilizing the thermal storage capacity within the commercial building (such as materials or water with a large heat capacity). During off-peak load periods, the air-conditioning temperature can be appropriately increased to reduce energy consumption.

[0041] Step S108: According to the control strategy, adjust the temperature of the temperature adjustment device from the initial operating temperature to the target operating temperature.

[0042] In the technical solution provided in step S108 of the embodiment of the present invention, after determining the control strategy of the temperature adjustment device in the future period based on the load prediction result, the temperature of the temperature adjustment device can be adjusted from the initial operating temperature to the target operating temperature according to the control strategy. Among them, the load state of the temperature adjustment device at the target operating temperature can meet the load state requirements of the temperature adjustment device.

[0043] In this embodiment, according to the control strategy, the temperature of the temperature adjustment device is dynamically adjusted, and the temperature of the temperature adjustment device is adjusted from the initial operating temperature to the target operating temperature, thereby ensuring the comfort of maintenance personnel while meeting the load change requirements, and realizing the energy-saving, efficient and user-friendly management of the commercial building air-conditioning system.

[0044] Optionally, adjusting the temperature of the temperature adjustment device according to the air-conditioning response time and the adaptability of personnel to temperature changes can avoid discomfort caused by sudden temperature changes. For example, the air-conditioning temperature can be adjusted a few minutes before the peak load period, or a larger adjustment of the air-conditioning temperature can be implemented during off-peak periods (such as at night).

[0045] In steps S102 to S108 of the embodiment of the present invention, if the temperature of the temperature adjustment device needs to be adjusted, the state index of the target object can be obtained at the initial operating temperature of the temperature adjustment device; the load prediction model can be used to analyze the state index to obtain the load prediction result; based on the load prediction result, the control strategy of the temperature adjustment device in the future period can be determined; and the temperature of the temperature adjustment device can be adjusted from the initial operating temperature to the target operating temperature according to the control strategy. In this embodiment, by obtaining the state index of the target object, the comfort state of the target object can be determined; by analyzing the state index through the load prediction model, the accuracy of predicting the load state of the temperature adjustment device (such as an air conditioner) in the future period can be improved. Without reducing the comfort state of the target object, the air-conditioning temperature setting is dynamically adjusted according to the air-conditioning load state, reducing unnecessary energy consumption, thereby achieving the technical effect of improving the accuracy of the control of the temperature adjustment device and solving the technical problem of low accuracy of the control of the temperature adjustment device.

[0046] The embodiments of the present invention will be described in detail below in combination with the above steps.

[0047] As an optional embodiment, historical environmental data of the working environment, historical power consumption load data corresponding to the working environment, and historical time data are obtained, where the historical environmental data is used to represent the meteorological conditions of the temperature regulation device during the historical period, the historical power consumption load data is used to represent the power consumption load of the temperature regulation device during the historical period, and the historical time data is used to represent the time of the temperature regulation device during the historical period; in step S104, the load prediction model is used to analyze the state index to obtain a load prediction result, including: using the load prediction model to analyze the state index, historical environmental data, historical power consumption load data, and historical time data to obtain a load prediction result.

[0048] In this embodiment, the historical environmental data can be used to represent the meteorological conditions of the temperature regulation device during the historical period, such as indoor temperature, relative humidity, air velocity, and outdoor temperature. The historical power consumption load data can be used to represent the power consumption load of the temperature regulation device during the historical period. The historical time data can be used to represent the time of the temperature regulation device during the historical period, such as date characteristics (such as weekdays, weekends, holidays). Obtaining the historical environmental data of the working environment, the historical power consumption load data corresponding to the working environment, and the historical time data is an important basis for air-conditioning load prediction and formulating an effective temperature regulation strategy. In the process of using the load prediction model to analyze the state index to obtain a load prediction result, the load prediction model can be used to analyze the state index, historical environmental data, historical power consumption load data, and historical time data to obtain a load prediction result.

[0049] Optionally, the historical environmental data can be obtained through a sensor network, such as temperature sensors, humidity sensors, wind speed sensors, etc. The historical power consumption load data can be read through a power metering device (such as a smart meter) or extracted from a data platform. The historical time data can be obtained from the logs of a building management system or a power metering device. The logs of the building management system or the power metering device record historical timestamps, which can be associated with the historical environmental data and the historical power consumption load data for analyzing the air-conditioning load characteristics at specific times (such as weekdays, weekends, holidays).

[0050] It should be noted that the above-mentioned acquisition methods of the historical environmental data, historical power consumption load data, and historical time data are only for illustrative purposes and are not specifically limited here. As long as the data can be acquired, it is within the protection scope of the embodiments of the present invention.

[0051] Optionally, after obtaining the historical environmental data of the working environment, as well as the historical power consumption load data and historical time data corresponding to the working environment, the historical power consumption load data can be processed for missing values, outliers, and normalized processing. The obtained historical time data can be processed for feature encoding. For example, the date data in the historical time data can be encoded to mark date features such as holidays, weekends, and working days. Thus, ensuring that all data input into the load prediction model is within a suitable range.

[0052] As an alternative embodiment, the load prediction model includes an influencing factor attention module, a time series neural network module, and a time attention module. Among them, the load prediction model is used to analyze the state index, historical environmental data, historical power consumption load data, and historical time data to obtain a load prediction result, including: using the influencing factor attention module to extract features from the state index, historical environmental data, historical power consumption load data, and historical time data to obtain target features; determining the attention weights corresponding to the target features; performing a fusion operation on the attention weights and the target features to obtain a fusion result; using the fusion result to adjust the attention weights to obtain an attention weight feature vector; using the time series neural network module to analyze the attention weight feature vector to obtain a key feature vector, where the key feature vector is used to represent the feature vector corresponding to the attention weight greater than the attention weight threshold in the attention weight feature vector; using the time attention module to analyze the key feature vector according to the time series to obtain a time attention weight feature vector, where the time attention weight feature vector is used to represent the importance of the time attention weight feature vector in the time series; performing a fusion operation on the key feature vector and the time attention weight feature vector, and obtaining a load prediction result through the load prediction model.

[0053] In this embodiment, the load prediction model may include an influencing factor attention module, a time series neural network module, and a time attention module. If it is necessary to analyze the state index, historical environmental data, historical electricity load data, and historical time data using the load prediction model to obtain the load prediction result, the influencing factor attention module can be used to extract features from the state index, historical environmental data, historical electricity load data, and historical time data to obtain target features; the attention weights corresponding to the target features can be determined; a fusion operation can be performed on the attention weights and the target features to obtain a fusion result; the attention weights can be adjusted using the fusion result to obtain an attention weight feature vector; the time series neural network module can be used to analyze the attention weight feature vector to obtain a key feature vector, where the key feature vector is used to represent the feature vector corresponding to the attention weight greater than the attention weight threshold in the attention weight feature vector; the time attention module can be used to analyze the key feature vector according to the time series to obtain a time attention weight feature vector, where the time attention weight feature vector is used to represent the importance degree of the time attention weight feature vector in the time series; the key feature vector and the time attention weight feature vector can be fused, and through the load prediction model, the load prediction result can be obtained.

[0054] Optionally, the influencing factor attention module can use the attention mechanism to distinguish the difference in the influence degree of different influencing factors on the building air-conditioning load under various conditions. For commercial buildings, during the night period, the influence degree of the "outdoor temperature" dimension on the air-conditioning load is greater than that of the "whether it is a working day" dimension. During the day, especially at noon, the influence degree of the "whether it is a working day" dimension on the building load is greater.

[0055] Optionally, after the state index, historical environmental data, historical electricity load data, and historical time data are used as the initial influencing factor matrix I and input into the influencing factor attention module, they can be split into 7 influencing factor row vectors i by row r (r = 1, 2, 3, …, 7). The convolutional layer structure can be composed of k single-channel 1×1 convolutional kernels, an activation function, and 1 k-channel 1×1 convolutional kernel connected in series. Among them, the activation function can adopt the form of a rectified linear unit (ReLU for short). In the convolutional layer structure, k single-channel 1×1 convolutional kernels can capture the non-monotonic relationship between the influencing factors and the building air-conditioning load. 1 k-channel 1×1 convolutional kernel can re-fuse the k T-dimensional row vectors into 1 T-dimensional feature row vector. After the calculation of the convolutional layer, each i r can obtain the corresponding feature row vector c r .

[0056] Optionally, introduce a T-dimensional influencing factor attention vector ui , and calculate the c of each feature row vector r and the inner product u i , and use the normalized exponential function (softmax function) for normalization to obtain a 7-dimensional influence factor attention weight vector a1. Among them, the value a1(r) of the r-th dimension of the vector a1 can be calculated by the following formula:

[0057]

[0058] Optionally, multiply each feature row vector c r by a1(r) to obtain 7 row vectors, and concatenate them by row to obtain a feature matrix C containing the influence factor attention mechanism.

[0059] Optionally, the convolution kernel parameters of the convolutional layer and the value of the influence factor attention vector u1 can be automatically learned during the training of the Transformer-LSTM model.

[0060] Optionally, the time series neural network module can be an LSTM. The feature matrix C containing the influence factor attention can be split into feature matrix column vectors by column as the input of the LSTM model. The feature information of each time step can be extracted to obtain the hidden layer vector h corresponding to each column vector t .

[0061] Optionally, the time attention module can distinguish the difference in the influence degree of influence factors at different times on the building air conditioning load through the attention mechanism.

[0062] Optionally, after inputting T l-dimensional hidden layer vectors into the time attention module, an l-dimensional influence factor attention vector u2 can be introduced, and the inner product of each hidden layer vector can be calculated. The softmax function can be used for normalization to obtain a T-dimensional influence factor attention weight vector a2. Among them, the value a2(t) of the t-th dimension of the vector a2 can be calculated by the following formula:

[0063]

[0064] Optionally, multiply each hidden layer vector h t by a2(t) and sum them to obtain an output vector h0 containing the time attention mechanism. Among them, the value of the time attention vector u2 can be automatically learned during the training of Transformer-LSTM.

[0065] Optionally, after the calculation of the completion time attention module, the output vector h0 with time attention can be input into a Support Vector Regression (SVR) model, and the final predicted value of the building air-conditioning load can be calculated by the support vector regression method. The parameters of the SVR model can be automatically learned during the model training process.

[0066] As an alternative embodiment, in step S102, at the initial operating temperature of the temperature regulating device, obtaining the state index of the target object includes: obtaining the state information of the target object and the attribute information in the working environment where the temperature regulating device is located, where the state information of the target object is used to represent the state related to the behavior of the target object in the working environment, and the attribute information is used to represent the environmental factors corresponding to the working environment; based on the state information and the attribute information, determining the state index of the target object.

[0067] In this embodiment, if it is necessary to obtain the state index of the target object at the initial operating temperature of the temperature regulating device, the state information of the target object and the attribute information in the working environment where the temperature regulating device is located can be obtained; the state index of the target object, such as the PMV value, can be determined based on the state information and the attribute information. Among them, the state information of the target object can be used to represent the state related to the behavior of the target object in the working environment, and the attribute information can be used to represent the environmental factors corresponding to the working environment.

[0068] Optionally, the state information may include the human metabolic rate. The attribute information may include the indoor temperature, outdoor temperature, relative humidity, air velocity, etc. Combining the state information and the attribute information can reflect the comfort state of the target object under specific environmental conditions.

[0069] Optionally, the human metabolic rate refers to the rate at which a person consumes energy during various activities, reflecting the level of activity of the person in the working environment. For example, in the shopping mall environment of a commercial building, the activity levels of customers and employees are different, and the human metabolic rate will also vary; the number of people is usually higher on weekends and holidays than on weekdays, and the activity levels during shopping and entertainment are also higher, so the overall human metabolic rate will increase, and at this time the load demand for the air conditioner will also increase. The indoor temperature can be the ambient temperature in the working environment where the temperature regulating device is located. For example, in the shopping mall environment, it can be the indoor temperature in the shopping mall. The outdoor temperature has a direct impact on the indoor temperature and the air conditioner load status. For example, in the shopping mall environment, when the doors and windows of the shopping mall are opened or people enter and exit frequently, it will affect the indoor temperature and the air conditioner load status. The relative humidity is the ratio of the actual water vapor content in the air to the maximum water vapor content that the air can hold at that temperature, usually expressed as a percentage. In the shopping mall environment, a higher relative humidity will slow down the evaporation of sweat from the body, thus affecting the natural cooling process of the human body and making people feel more stuffy. Even at the same temperature, the comfort state of people is worse. The air velocity is the speed of air flow in the working environment where the temperature regulating device is located, affecting the air conditioner load status and the comfort state of people.

[0070] Optionally, the PMV value can be determined by the following formula:

[0071]

[0072] where M can represent the human metabolic rate, 1 met = 58 W / m 2 ; W can represent the mechanical power generated by human activities, which can be ignored for most activities and can be 0 W / m when sitting still 2 ; p a can represent the water vapor pressure, which is related to the relative humidity of the air; T in can represent the indoor temperature; f cl can represent the clothing coefficient; T cl can represent the temperature of the outer surface of the clothes; can represent the mean radiant temperature; h c can represent the convective heat transfer coefficient.

[0073] The water vapor pressure p a can be determined by the following formula:

[0074]

[0075] where r h can represent the relative humidity.

[0076] The clothing coefficient f cl can be determined by the following formula:

[0077]

[0078] Among them, I cl can represent the clothing thermal resistance, which is determined by the amount of clothing a person wears, with the unit of col. The insulation coefficient of wearing an inner shirt and an ordinary outer coat can be 1 col.

[0079] The clothing thermal resistance I cl can be determined by the following formula:

[0080]

[0081] Among them, T out,6 can represent the outdoor temperature at 6 am.

[0082] The temperature on the outer surface of the clothes T cl can be determined by the following formula:

[0083]

[0084] The convective heat transfer coefficient h c can be determined by the following formula:

[0085]

[0086] Among them, v ar can represent the air velocity.

[0087] As an alternative embodiment, in step S106, based on the load prediction result, determine the control strategy of the temperature regulation device in the future time period, including: in response to the load prediction result being less than or equal to the load prediction threshold, determine the initial working temperature as the target working temperature, and determine the control strategy as the first control strategy; in response to the load prediction result being greater than the load prediction threshold, adjust the initial working temperature to the target working temperature, and determine the control strategy as the second control strategy.

[0088] In this embodiment, in the process of determining the control strategy of the temperature regulation device in the future time period based on the load prediction result, when the load prediction result is less than or equal to the load prediction threshold, the initial working temperature can be determined as the target working temperature, and the control strategy can be determined as the first control strategy; when the load prediction result is greater than the load prediction threshold, the initial working temperature can be adjusted to the target working temperature, and the control strategy can be determined as the second control strategy.

[0089] Optionally, in order to avoid unnecessary energy waste when the load demand is low, and take measures to reduce the load when the demand is high to avoid load overload of the temperature regulation device, a load prediction threshold can be set.

[0090] Optionally, when the load prediction result is less than or equal to the load prediction threshold, it indicates that the current temperature (i.e., the initial operating temperature) can meet the future load demand, while also considering occupant comfort. In this case, there is no need to adjust the temperature setting of the temperature control device, that is, to keep the temperature control device operating at the current initial operating temperature. When the load prediction result is greater than the load prediction threshold, it means that if no measures are taken, the future load demand will put pressure on the temperature control device, resulting in energy waste. In this case, the initial operating temperature of the temperature control device can be dynamically adjusted to a more energy-efficient target operating temperature. For example, slightly increase the air conditioner set temperature in summer, or slightly lower the air conditioner set temperature in winter, to reduce the future load demand while maintaining occupant comfort.

[0091] As an alternative embodiment, in step S108, adjusting the temperature of the temperature control device from the initial operating temperature to the target operating temperature according to the control strategy includes: generating control prompt information according to the control strategy; and adjusting the temperature of the temperature control device to the target operating temperature based on the control prompt information.

[0092] In this embodiment, during the process of adjusting the temperature of the temperature control device from the initial operating temperature to the target operating temperature according to the control strategy, control prompt information can be generated according to the control strategy; and the temperature of the temperature control device can be adjusted to the target operating temperature based on the control prompt information.

[0093] Optionally, the generated control prompt information can be sent to the temperature control device through a network or an internal communication system. After receiving the control prompt information, the temperature control device can adjust the output power of the refrigeration or heating device according to the adjustment strategy in the prompt information, so as to change the initial operating temperature to reach the target operating temperature.

[0094] In the embodiment of the present invention, if the temperature of the temperature control device needs to be adjusted, the status index of the target object can be obtained at the initial operating temperature of the temperature control device; the load prediction model can be used to analyze the status index to obtain the load prediction result; the control strategy of the temperature control device in the future time period can be determined based on the load prediction result; and the temperature of the temperature control device can be adjusted from the initial operating temperature to the target operating temperature according to the control strategy. In this embodiment, by obtaining the status index of the target object, the comfort state of the target object can be determined; by analyzing the status index through the load prediction model, the accuracy of predicting the load state of the temperature control device (such as an air conditioner) in the future time period can be improved. Without reducing the comfort state of the target object, the air conditioner temperature setting is dynamically adjusted according to the air conditioner load state, reducing unnecessary energy consumption, thereby achieving the technical effect of improving the accuracy of the control of the temperature control device and solving the technical problem of low accuracy of the control of the temperature control device.

[0095] Embodiment 2

[0096] The following is a detailed description in combination with another optional specific implementation manner.

[0097] At present, with the rapid growth of power load and the continuous growth of seasonal peak load, the contradiction between the instability of new energy supply and the high seasonal electricity demand has intensified the test of the safe operation of the power grid and the reliability of power supply. In order to ensure the matching of renewable energy power supply and power demand and improve its own regulation ability, it is difficult to ensure the stable operation of the system only by focusing on the supply side. There is a large regulation potential on the demand side and the cost is relatively low. Among them, the electricity consumption of commercial buildings shows significant peak characteristics. The temperature is high and the passenger flow is large during the day in summer, and there are significant differences in the passenger flow on weekdays and weekends, with great adjustable potential.

[0098] In the related art, regarding the regulation potential of air-conditioning load on the demand side, the regulation strategy is mainly analyzed from the power grid level, or the analysis is carried out on the performance and start-stop behavior of the air conditioner. Under the constraint of human comfort, the refined research on the regulation potential of air-conditioning load in commercial buildings remains to be supplemented. Therefore, there is still the technical problem of low control accuracy of the temperature regulation device.

[0099] The present invention proposes a method for predicting the dynamic adjustable potential of air-conditioning load in commercial buildings considering human comfort. By fully considering the human comfort in commercial building groups and giving the influence on air-conditioning load under different regulation strategies, the prediction of the dynamic adjustable potential of air-conditioning load is realized. On the premise of ensuring indoor human comfort, the load level of the day to be adjusted is accurately predicted, the load levels that can be released under different adjustment strategies are provided, and the air-conditioning load of the building is dynamically adjusted to promote the supply-demand balance of the power grid. Furthermore, the technical effect of improving the control accuracy of the temperature regulation device is achieved, and the technical problem of low control accuracy of the temperature regulation device is solved.

[0100] The following is a further introduction to this method.

[0101] In this embodiment, Figure 2 is a flowchart of a method for predicting the dynamic adjustable potential of air-conditioning load in commercial buildings considering human comfort according to an embodiment of the present invention. As Figure 2 shown, the method may include:

[0102] Step S201, obtaining a human perception index.

[0103] In this embodiment, the human perception index can reflect human comfort, and can be the PMV value. The PMV value is an index used to describe human thermal comfort and can be determined by the following formula:

[0104]

[0105] Among them, M can represent the human metabolic rate, and 1 met = 58 W / m 2 ; W can represent the mechanical power generated by human activities, which can be ignored for most activities and can be 0 W / m during sitting still 2 ; p a can represent the water vapor pressure, which is related to the relative humidity of the air; T in can represent the indoor temperature; f cl can represent the clothing coefficient; T cl can represent the temperature of the outer surface of the clothes; can represent the mean radiant temperature; h c can represent the convective heat transfer coefficient.

[0106] The water vapor pressure p a can be determined by the following formula:

[0107]

[0108] Among them, r h can represent the relative humidity.

[0109] The clothing coefficient f cl can be determined by the following formula:

[0110]

[0111] Among them, I cl can represent the clothing thermal resistance, which is determined by the amount of clothing a person wears, with the unit of col, The insulation coefficient of wearing an inner shirt and an ordinary outer coat can be 1 col.

[0112] The clothing thermal resistance I cl can be determined by the following formula:

[0113]

[0114] Among them, T out,6 can represent the outdoor temperature at 6:00 in the morning.

[0115] The temperature of the outer surface of the clothes T cl can be determined by the following formula:

[0116]

[0117] The convective heat transfer coefficient h c can be determined by the following formula:

[0118]

[0119] Among them, var It can represent the air flow velocity.

[0120] Optionally, collect indicators such as outdoor temperature, humidity, and air flow velocity in different periods in the Beijing area. The indoor temperature is determined according to the air conditioner set temperature. The human energy metabolism rate M can be set to 1 met, the mechanical work W done by the human body can be set to 0.5, the clothing insulation coefficient can be taken as 0.5 col in summer, 1 col in spring and autumn, and 1.5 col in winter. The indoor temperature set value can be gradually adjusted to obtain different PMV index values in the corresponding environment. According to the ISO-7730 standard, the recommended value of the PMV index is between -0.5 and 0.5. If the change in the indoor temperature keeps the PMV value within this range, the adjustment strategy is considered reasonable. The ISO 7730 standard is formulated by the International Organization for Standardization and is used to evaluate and determine the thermal comfort of the human body in the indoor thermal environment.

[0121] Optionally, based on the air conditioner temperature with 24°C in summer and 20°C in winter as the benchmark, and based on the above-determined human metabolism rate index, combined with the meteorological data at the main times of each day in history, calculate the corresponding human perception index.

[0122] Step S202, obtain the historical electricity load data of the construction unit.

[0123] In this embodiment, the historical electricity load data of typical shopping mall buildings from 9 to 6 o'clock in the past three years can be extracted based on the data middle platform.

[0124] Step S203, calculate the air conditioner load data.

[0125] In this embodiment, after extracting the historical electricity load data of typical shopping mall buildings from 9 to 6 o'clock in the past three years based on the data middle platform, the maximum load average value in April of each year with the sensible temperature between 15 - 25°C can be used as the basic load, and subtracting the basic load from the total load at each moment in summer can obtain the air conditioner load data.

[0126] Step S204, obtain the date data.

[0127] In this embodiment, the date data can be obtained to judge whether it is a working day or a holiday.

[0128] Step S205, perform data preprocessing.

[0129] In this embodiment, perform feature encoding on the date data to mark date features such as holidays, weekends, and working days. The air conditioner load data can be viewed and missing values, outliers, and error values can be processed. Combining meteorological data such as outdoor temperature, humidity, and air circulation speed, the data can be normalized.

[0130] Assume an index x, and the maximum value is x max, the minimum value is x min , then the normalized x at time t i is as follows:

[0131]

[0132] Step S206: Divide the input data into a training set and a test set.

[0133] In this embodiment, the data from May to October in 2021 - 2023 can be used as the training set of the model, and the data from May to July in 2024 can be used as the test set. Metrics such as Mean Absolute Percentage Error (MAPE for short) and Root Mean Square Error (RMSE for short) can be used to evaluate the prediction effect of the model.

[0134]

[0135]

[0136] Among them, there are D load values to be predicted. The actual value of the load can be expressed as y0(d), and the predicted value can be denoted as y(d); d = 1, 2,... D.

[0137] Step S207: Input the data into the Transformer - LSTM prediction model.

[0138] In this embodiment, the human perception index, air - conditioning load data, date data, etc. can be input into the Transformer - LSTM prediction model to predict the load result.

[0139] Step S208: Adjust the air - conditioning set temperature.

[0140] In this embodiment, different air - conditioning set temperatures will affect the air - conditioning load. The air - conditioning set temperature can be adjusted, and the human perception index (such as the PMV value) under different air - conditioning set temperatures can be calculated to evaluate to what extent the air - conditioning load can be adjusted while ensuring the comfort of indoor personnel.

[0141] Step S209: Obtain a new human perception index.

[0142] In this embodiment, the air - conditioning set temperature can be adjusted to obtain different human perception indices for the day to be predicted respectively.

[0143] Step S210: Output the predicted benchmark air - conditioning load and the adjusted load prediction result.

[0144] In this embodiment, the human perception index, air-conditioning load data, date data, etc. are input into the Transformer-LSTM prediction model for load result prediction to obtain the predicted benchmark air-conditioning load. By adjusting the air-conditioning set temperature, different human perception indices for the day to be predicted are obtained and substituted into the Transformer-LSTM prediction model in sequence, and the adjusted load prediction results can be obtained.

[0145] Step S211: Determine the adjustable potential evaluation result through comparison.

[0146] In this embodiment, by comparing the benchmark load prediction result with the air-conditioning load prediction values under the constraints of different human perception indices, the change of the load under different temperature settings can be analyzed, the adjustable potential of each prediction day can be evaluated, and thus the adjustable potential evaluation result of each prediction day can be obtained. Ultimately, without significantly reducing the human comfort level, the load can be reduced by adjusting the air-conditioning set temperature, and the adjustable space of the load can be evaluated. Based on these evaluation results, different air-conditioning temperature setting schemes are specifically given for the summer peak shaving and valley filling demands under different weather conditions.

[0147] Figure 3 It is a schematic structural diagram of a transformer-long short-term memory network model according to an embodiment of the present invention. As Figure 3 shown, it includes: an influencing factor attention module 301, an LSTM module 302, and a time attention module 303.

[0148] The influencing factor attention module 301 is used to distinguish the difference in the influence degree of different influencing factors on the building air-conditioning load under various conditions through the attention mechanism. For example, for commercial buildings, during the night period, the influence degree of the "outdoor temperature" dimension on the air-conditioning load is greater than that of the "whether it is a working day" dimension. During the day, especially at noon, the influence degree of the "whether it is a working day" dimension on the building load is greater.

[0149] Optionally, after the initial influencing factor matrix I is input into the influencing factor attention module 301, it can be split into 7 influencing factor row vectors i r (r = 1, 2, 3,..., 7) by rows. The convolutional layer structure can be concatenated by k single-channel 1×1 convolutional kernels, an activation function, and 1 k-channel 1×1 convolutional kernel. Among them, the activation function can adopt the ReLU form. In the convolutional layer structure, the k single-channel 1×1 convolutional kernels can capture the non-monotonic relationship between the influencing factors and the building air-conditioning load. The 1 k-channel 1×1 convolutional kernel can re-fuse the k T-dimensional row vectors into 1 T-dimensional feature row vector. After the calculation of the convolutional layer, each i r can obtain the corresponding feature row vector c r .

[0150] The T-dimensional influence factor attention vector u can be introduced i , and the c of each feature row vector is calculated r with the inner product u i . Using the softmax function for normalization, the 7-dimensional influence factor attention weight vector a1 is obtained. Among them, the value a1(r) of the r-th dimension of the vector a1 can be calculated by the following formula:

[0151]

[0152] Multiplying each feature row vector c r by a1(r), 7 row vectors can be obtained, and by concatenating them row by row, the feature matrix C with the influence factor attention mechanism can be obtained.

[0153] The convolution kernel parameters of the convolutional layer and the value of the influence factor attention vector u1 can be automatically learned during the training process of the Transformer-LSTM model.

[0154] The LSTM module 302, which consists of multiple LSTM cells, is used to receive the feature vector matrix from the influence factor attention module 301, and through its internal gating mechanism (input gate, forget gate, and output gate) to memorize and filter information, so as to capture the dynamic features of the data sequence. By generating a sequence of hidden state vectors, each vector corresponding to a time step, the above hidden state vectors contain the feature model of the time series and the potential load change trend.

[0155] Optionally, the feature matrix C containing the influence factor attention is split into feature matrix column vectors by column and used as the input of the LSTM model. The feature information of each time step can be extracted to obtain the hidden layer vector h corresponding to each column vector t .

[0156] The time attention module 303 is used to distinguish the difference in the influence degree of the influence factors at different moments on the building air-conditioning load through the attention mechanism, and receives the hidden layer vector h t .

[0157] Optionally, after inputting T l-dimensional hidden layer vectors into the time attention module, the l-dimensional influence factor attention vector u2 can be introduced, and the inner product of each hidden layer vector is calculated. The softmax function can be used for normalization to obtain the T-dimensional influence factor attention weight vector a2. Among them, the value a2(t) of the t-th dimension of the vector a2 can be calculated by the following formula:

[0158]

[0159] Multiplying each hidden layer vector h tMultiply it by a2(t) and sum it up to obtain the output vector h0 with a time attention mechanism. Among them, the value of the time attention vector u2 can be automatically learned during the training process of Transformer-LSTM.

[0160] After completing the calculation of the time attention module, the output vector h0 with time attention can be input into a Support Vector Regression (SVR) model, and the final predicted value of the building air-conditioning load can be calculated through the support vector regression method to obtain the air-conditioning load result. The parameters of the SVR model can be automatically learned during the model training process.

[0161] In the embodiment of the present invention, the Transformer-LSTM model can achieve multi-factor consideration and time series prediction of the commercial building air-conditioning load through the cascaded use of the above three modules. The Transformer-LSTM model can not only process various heterogeneous data affecting load prediction, but also dynamically adjust the weights of different factors and time points through the attention mechanism, so as to accurately predict the air-conditioning load under the constraint of human comfort, provide support for the peak shaving and valley filling strategy of the power grid, and realize the efficient utilization of building energy and load management.

[0162] Embodiment 3

[0163] The embodiment of the present invention provides a control device for a temperature regulation device. It should be noted that the control device for the temperature regulation device in the embodiment of the present invention can be used to execute Figure 1 the control for the temperature regulation device provided in the embodiment of the present invention. The following introduces the control device for the temperature regulation device provided in the embodiment of the present invention.

[0164] Figure 4 is a schematic structural diagram of a control device for a temperature regulation device according to an embodiment of the present invention. As Figure 4 shown, the control device 400 for the temperature regulation device may include: an acquisition unit 402, an analysis unit 404, a determination unit 406, and an adjustment unit 408.

[0165] The acquisition unit 402 is configured to acquire the state index of the target object at the initial working temperature of the temperature regulation device, where the target object is in the working environment of the temperature regulation device, and the state index is used to represent the comfort state of the target object in the working environment.

[0166] The analysis unit 404 is configured to analyze the state index by using a load prediction model to obtain a load prediction result, where the load prediction result is used to represent the load state of the temperature regulation device at the initial working temperature in a future time period, and the load prediction model is obtained by training a deep learning model.

[0167] A determination unit 406, configured to determine a control strategy of the temperature adjustment device in a future period based on a load prediction result, where the control strategy is used to represent a rule for adjusting an initial working temperature.

[0168] An adjustment unit 408, configured to adjust the temperature of the temperature adjustment device from the initial working temperature to a target working temperature according to the control strategy, where the load state of the temperature adjustment device at the target working temperature meets the load state requirement of the temperature adjustment device.

[0169] The control device of the temperature adjustment device provided by the embodiment of the present invention obtains, by an acquisition unit 402, a state index of a target object at an initial working temperature of the temperature adjustment device, where the target object is in the working environment of the temperature adjustment device, and the state index is used to represent the comfort state of the target object in the working environment; analyzes, by an analysis unit 404, the state index by using a load prediction model to obtain a load prediction result, where the load prediction result is used to represent the load state of the temperature adjustment device at the initial working temperature in a future period, and the load prediction model is obtained by training a deep learning model; determines, by a determination unit 406, a control strategy of the temperature adjustment device in a future period based on the load prediction result, where the control strategy is used to represent a rule for adjusting the initial working temperature; and adjusts, by an adjustment unit 408, the temperature of the temperature adjustment device from the initial working temperature to the target working temperature according to the control strategy, where the load state of the temperature adjustment device at the target working temperature meets the load state requirement of the temperature adjustment device. Therefore, the technical effect of improving the control accuracy of the temperature adjustment device is achieved, and the technical problem of low control accuracy of the temperature adjustment device is solved.

[0170] The control device of the above temperature adjustment device may further include a processor and a memory. The above units are stored in the memory as program units, and the corresponding functions are implemented by the processor executing the above program units stored in the memory.

[0171] The above processor includes a kernel, and the kernel retrieves the corresponding program unit from the memory. One or more kernels may be set, and by adjusting the kernel parameters, the devices to be shut down of the same device type are controlled to perform graceful shutdown.

[0172] The above memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of, for example, read-only memory (ROM) or flash RAM, and the memory includes at least one storage chip.

[0173] The processor contains cores, which retrieve corresponding program units from the memory. One or more cores can be set, and the work efficiency of traders can be improved by adjusting the core parameters.

[0174] The memory may include non-permanent memory in a computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM (f l ash RAM), and the memory includes at least one memory chip.

[0175] Embodiment 4

[0176] According to an embodiment of the present invention, there is also provided a computer-readable storage medium, on which a program is stored, and when the program is executed by a processor, it implements the control method of the temperature adjustment device.

[0177] Embodiment 5

[0178] According to an embodiment of the present invention, there is also provided a processor for running a program, and when the program runs, it implements the control method of the temperature adjustment device.

[0179] Embodiment 6

[0180] Figure 5 is a schematic diagram of an electronic device according to an embodiment of the present invention. As Figure 5 shown, an embodiment of the present invention also provides an electronic device, which includes a processor, a memory, and a program stored on the memory and executable on the processor. When the processor executes the program, the above steps are implemented.

[0181] The devices herein can be servers, PCs, PADs, mobile phones, etc.

[0182] The present invention also provides a computer program product, which is suitable for executing a program initialized with the above method steps when executed on a data processing device.

[0183] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0184] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, as well as the combination of flows and / or blocks in the flowchart and / or block diagram. These computer program instructions can be provided to the processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing device produce means for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or means for implementing the functions specified in multiple blocks.

[0185] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or means for implementing the functions specified in multiple blocks.

[0186] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or means for implementing the functions specified in multiple blocks.

[0187] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.

[0188] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM), and / or non-volatile memory, such as read-only memory (ROM) or flash memory (flash RAM). The memory is an example of computer-readable media.

[0189] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include transitory media such as modulated data signals and carrier waves.

[0190] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0191] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0192] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for controlling a temperature regulating device, characterized in that: include: At the initial working temperature of the temperature adjustment device, obtaining a state index of a target object, wherein the target object is in a working environment of the temperature adjustment device, and the state index is used to indicate a comfortable state of the target object in the working environment; Analyzing the state index using a load prediction model to obtain a load prediction result, wherein the load prediction result is used to represent the load state of the temperature adjustment device at the initial working temperature in a future period, and the load prediction model is obtained by training a deep learning model; Based on the load forecast result, determining a control strategy of the temperature adjustment device in the future period, wherein the control strategy is used to represent a rule for adjusting the initial operating temperature; According to the control strategy, the temperature of the temperature regulating device is adjusted from the initial operating temperature to the target operating temperature, wherein the load state of the temperature regulating device at the target operating temperature meets the load state requirement of the temperature regulating device.

2. The control method according to claim 1, characterized in that: The method further comprises: Acquire historical environment data of the working environment, and historical power load data and historical time data corresponding to the working environment, wherein the historical environment data is used to represent the meteorological conditions of the temperature control device in a historical period, the historical power load data is used to represent the power load of the temperature control device in the historical period, and the historical time data is used to represent the time of the temperature control device in the historical period; The state index is analyzed using a load forecasting model to obtain a load forecasting result, including: using the load forecasting model to analyze the state index, the historical environmental data, the historical power load data and the historical time data to obtain the load forecasting result.

3. The control method according to claim 2, characterized in that: The load forecasting model includes an influencing factor attention module, a temporal neural network module and a time attention module, wherein the load forecasting model is used to analyze the state index, the historical environmental data, the historical power load data and the historical time data to obtain the load forecasting result, including: Using the influencing factor attention module to extract features from the state index, the historical environment data, the historical power load data and the historical time data to obtain target features; Determining an attention weight corresponding to the target feature; Performing a fusion operation on the attention weight and the target feature to obtain a fusion result; Using the fusion result, adjusting the attention weight to obtain an attention weight feature vector; Analyzing the attention weight feature vector using the temporal neural network module to obtain a key feature vector, wherein the key feature vector is used to represent the feature vector corresponding to the attention weight in the attention weight feature vector being greater than an attention weight threshold; Using the time attention module, the key feature vector is analyzed according to the time series to obtain a time attention weight feature vector, wherein the time attention weight feature vector is used to indicate the importance of the time attention weight feature vector on the time series; The key feature vector is fused with the time attention weight feature vector, and the load forecasting result is obtained through the load forecasting model.

4. The control method according to claim 1, characterized in that: At the initial operating temperature of the temperature regulating device, the state index of the target object is obtained, including: Acquire state information of the target object and attribute information of the working environment where the temperature adjustment device is located, wherein the state information is used to represent a state related to the behavior of the target object in the working environment, and the attribute information is used to represent environmental factors corresponding to the working environment; Based on the state information and the attribute information, the state index of the target object is determined.

5. The control method according to claim 1, characterized in that: Determining a control strategy of the temperature regulating device in the future period based on the load forecast result includes: In response to the load prediction result being less than or equal to the load prediction threshold, determining the initial operating temperature as the target operating temperature, and determining the control strategy to be a first control strategy; In response to the load prediction result being greater than the load prediction threshold, the initial operating temperature is adjusted to the target operating temperature, and the control strategy is determined to be a second control strategy.

6. The control method according to claim 1, characterized in that: According to the control strategy, adjusting the temperature of the temperature adjustment device from the initial operating temperature to the target operating temperature includes: Generate control prompt information according to the control strategy; Based on the control prompt information, the temperature of the temperature adjustment device is adjusted to the target operating temperature.

7. A control device for a temperature regulating device, characterized in that: include: an acquisition unit, configured to acquire a state index of a target object at an initial working temperature of the temperature adjustment device, wherein the target object is in a working environment of the temperature adjustment device, and the state index is used to indicate a comfortable state of the target object in the working environment; an analysis unit, configured to analyze the state index using a load prediction model to obtain a load prediction result, wherein the load prediction result is used to represent the load state of the temperature adjustment device at the initial working temperature in a future period, and the load prediction model is obtained by training a deep learning model; a determining unit, configured to determine a control strategy of the temperature regulating device in the future period based on the load forecast result, wherein the control strategy is used to represent a rule for regulating the initial operating temperature; The regulating unit is used to regulate the temperature of the temperature regulating device from the initial operating temperature to the target operating temperature according to the control strategy, wherein the load state of the temperature regulating device at the target operating temperature meets the load state requirement of the temperature regulating device.

8. A processor, characterized in that: The processor is used to run a program, wherein the program, when run by the processor, executes the control method of the temperature adjustment device according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the control method of the temperature adjustment device according to any one of claims 1 to 6.

10. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to run the computer program to execute the control method of the temperature adjustment device according to any one of claims 1 to 6.

11. A computer program product, characterized in that The invention comprises a computer program, which, when executed by a processor, implements the control method of the temperature adjustment device according to any one of claims 1 to 6.

Images