Air conditioner load prediction method and device, computer equipment and medium

Abstract

The invention provides an air conditioner load prediction method and device, computer equipment and a medium, and the method comprises the steps that environmental factor data of an air conditioner area in a preset time period are obtained, and the environmental factor data comprise the current radiation intensity, the current temperature and the current relative humidity; based on the environmental factor data and an initial load prediction model, the type of the load of the air conditioner area in the preset time period is determined, and the initial load prediction model is used for conducting initial prediction on the air conditioner load through the environmental factor data; and determining a load prediction result of the air conditioner area in the preset time period based on the load category of the air conditioner area in the preset time period. Therefore, by introducing the machine learning method, a new technical path is provided for air conditioner load prediction, and accurate prediction of the air conditioner load is facilitated.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the technical field of air conditioning load prediction, and in particular, to an air conditioning load prediction method, apparatus, computer equipment, and medium. Background Art

[0002] Air conditioning load refers to the amount of heat that needs to be removed from the room per unit time (cold load) while keeping the indoor air thermal and humidity parameters within a certain required range, or the amount of heat (heat load) supplied to the building by the heating system per unit time to achieve the required indoor temperature at a certain outdoor temperature in winter.

[0003] Air conditioning load forecasting is of great significance in building energy management, data center operations, and intelligent building control. With the continuous increase in energy consumption and the improvement of energy conservation and emission reduction requirements, how to accurately predict air conditioning load to optimize energy use and reduce operating costs has become a problem that needs to be solved urgently. Summary of the invention

[0004] The embodiments described herein provide an air conditioning load prediction method, apparatus, device, and medium that overcome the above-mentioned problems.

[0005] In a first aspect, according to the present disclosure, there is provided an air conditioning load prediction method, comprising: Obtaining environmental factor data of the air-conditioned area within a preset period of time, the environmental factor data including: current radiation intensity, current temperature and current relative humidity; Based on the environmental factor data and the initial load prediction model, determine the load category of the air-conditioning area within the preset period, and the initial load prediction model is used to make an initial prediction of the air-conditioning load based on the environmental factor data; Determine the load forecast result of the air-conditioned area within the preset period based on the load category of the air-conditioned area within the preset period; Among them, the load categories of the air-conditioned area within the preset time period include: a first load category and a second load category. The first load category is used to describe that the load prediction value of the air-conditioned area within the preset time period is greater than the preset load threshold, and the second load category is used to describe that the load prediction value of the air-conditioned area within the preset time period is less than or equal to the preset load threshold.

[0006] In a second aspect, according to the content of the present disclosure, an air conditioning load prediction device is provided, comprising: An acquisition module is used to acquire environmental factor data of the air-conditioned area within a preset period of time, the environmental factor data including: current radiation intensity, current temperature and current relative humidity; A first determination module is used to determine the load category of the air-conditioning area within a preset period based on the environmental factor data and the initial load prediction model, wherein the initial load prediction model is used to make an initial prediction of the air-conditioning load based on the environmental factor data; A second determination module, configured to determine a load prediction result of the air-conditioned area within a preset period of time based on the category to which the load of the air-conditioned area belongs within the preset period of time; Among them, the load categories of the air-conditioned area within the preset time period include: a first load category and a second load category. The first load category is used to describe that the load prediction value of the air-conditioned area within the preset time period is greater than the preset load threshold, and the second load category is used to describe that the load prediction value of the air-conditioned area within the preset time period is less than or equal to the preset load threshold.

[0007] In a third aspect, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps of the air conditioning load prediction method in any of the above embodiments are implemented.

[0008] In a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the air conditioning load prediction method in any of the above embodiments are implemented.

[0009] The air conditioning load prediction method provided in the embodiment of the present application obtains the environmental factor data of the air-conditioned area within a preset time period, and the environmental factor data includes: current radiation intensity, current temperature and current relative humidity; based on the environmental factor data and the initial load prediction model, the load category of the air-conditioned area within the preset time period is determined, and the initial load prediction model is used to make an initial prediction of the air-conditioning load through the environmental factor data; based on the load category of the air-conditioned area within the preset time period, the load prediction result of the air-conditioned area within the preset time period is determined; wherein, the load category of the air-conditioned area within the preset time period includes: a first load category and a second load category, the first load category is used to describe that the load prediction value of the air-conditioned area within the preset time period is greater than a preset load threshold, and the second load category is used to describe that the load prediction value of the air-conditioned area within the preset time period is less than or equal to the preset load threshold. Thus, by introducing the machine learning method, a new technical path is provided for air-conditioning load prediction, which facilitates the accurate prediction of air-conditioning load.

[0010] The above description is only an overview of the technical solution of the embodiment of the present application. In order to more clearly understand the technical means of the embodiment of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the embodiment of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings of the embodiments will be briefly described below. It should be noted that the drawings described below only relate to some embodiments of the present disclosure, but are not intended to limit the present disclosure, wherein: Figure 1 It is a flow chart of an air conditioning load prediction method provided by an embodiment of the present disclosure.

[0012] Figure 2A It is a heat map of correlation coefficients between data provided by an embodiment of the present disclosure.

[0013] Figure 2B It is an effect diagram of a timing model provided by an embodiment of the present disclosure.

[0014] Figure 2C It is a rendering of a random forest model provided by an embodiment of the present disclosure.

[0015] Figure 2D It is an effect diagram of the initial load prediction model + target load prediction model provided by the embodiment of the present disclosure.

[0016] Figure 3 It is a structural schematic diagram of an air conditioning load prediction device provided in an embodiment of the present disclosure.

[0017] Figure 4 It is a structural diagram of a computer device provided in an embodiment of the present disclosure.

[0018] It should be noted that the elements in the drawings are schematic and not drawn to scale. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the embodiments of the present disclosure clearer, the technical solution of the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present disclosure, rather than all of the embodiments. Based on the described embodiments of the present disclosure, all other embodiments obtained by those skilled in the art without creative work also fall within the scope of protection of the present disclosure.

[0020] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by a person skilled in the art to which the subject matter of the present disclosure belongs. It will be further understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the specification and the relevant art, and will not be interpreted in an idealized or overly formal form unless otherwise explicitly defined herein. As used herein, a statement that two or more parts are "connected" or "coupled" together shall mean that the parts are joined together directly or through one or more intermediate components.

[0021] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase "embodiments" in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0022] The term "and / or" in this article is only a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists, A and B exist at the same time, and B exists. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship. Terms such as "first" and "second" are only used to distinguish one component (or a part of a component) from another component (or another part of a component).

[0023] In the description of the present application, unless otherwise specified, “plurality” means more than two (including two), and similarly, “plurality groups” means more than two (including two).

[0024] In terms of energy management and operating costs, air conditioning systems account for a significant proportion of energy consumption in data centers and large buildings. Air conditioning load forecasting can help managers take effective measures to optimize energy distribution during high-load periods, thereby reducing overall energy consumption and operating costs. Accurate load forecasting can also avoid overcooling and unnecessary energy waste, thereby achieving more efficient energy management.

[0025] In terms of environmental comfort and user experience, the main task of the air conditioning system is to maintain the comfort of the indoor environment. By accurately predicting the air conditioning load, the air conditioning operating parameters can be adjusted in advance to ensure the best comfort under different environmental conditions and improve the user experience. Especially in smart buildings, air conditioning load prediction can be linked with other smart systems to provide more personalized and humane services.

[0026] In terms of intelligent buildings and automated control, with the development of Internet of Things technology, intelligent buildings are becoming more and more popular. As part of intelligent buildings, air conditioning load forecasting can be integrated with the automated control system to adjust the operating status of air conditioning equipment in real time and improve the intelligence level and response speed of the system. This not only improves the energy efficiency of the building, but also enhances the adaptive ability and intelligence of the system.

[0027] In terms of sustainable development and environmental protection, reducing energy consumption and carbon emissions is a common goal of all countries in the context of global climate change and environmental protection. By optimizing the operation of air-conditioning systems, energy consumption can be significantly reduced, thereby reducing carbon emissions and achieving the goal of sustainable development. As an important means to achieve this goal, air-conditioning load forecasting has received widespread attention.

[0028] Traditional air conditioning load forecasting methods mostly rely on empirical formulas and simple statistical models, mainly based on historical data and meteorological parameters. Although these methods can provide load forecasting to a certain extent, they have obvious deficiencies in dealing with complex nonlinear relationships, real-time dynamic changes, and multi-dimensional data fusion, resulting in low prediction accuracy and reliability.

[0029] With the improvement of computing power and the development of big data technology, machine learning methods have been widely used in air conditioning load forecasting. Machine learning can handle complex nonlinear relationships, analyze multidimensional data, and provide high-precision forecasting results. In particular, the combination of ensemble learning (such as random forest) and deep learning (such as multilayer perceptron) methods provides strong technical support for air conditioning load forecasting.

[0030] Random forest is an ensemble learning method based on decision trees. It improves the accuracy and stability of prediction by building multiple decision trees. It can process high-dimensional feature data, has strong noise resistance and robustness, and is suitable for multi-factor impact analysis and data noise processing in air conditioning load forecasting.

[0031] Multi-Layer Perceptron (MLP) is a feedforward neural network with strong nonlinear fitting capabilities, which can capture the complex relationship between input features and output results. Through multiple layers of neurons and nonlinear activation functions, MLP performs well in air conditioning load forecasting, especially in processing time series data and dynamic changes.

[0032] Air conditioning load forecasting plays an important role in modern energy management and intelligent building control. With the increase in data volume and the development of computing technology, the introduction of machine learning methods (such as random forests and MLP) provides a new technical path to solve the complex problems in air conditioning load forecasting. By accurately predicting air conditioning load, multiple goals can be achieved, such as optimizing energy use, improving system efficiency, reducing operating costs, and enhancing user experience, contributing to sustainable development and environmental protection.

[0033] This embodiment uses the random forest + MLP method to predict air conditioning load, which can effectively solve key technical problems such as complex data relationship modeling, multi-factor impact analysis, data noise and outlier processing, dynamic load change prediction, efficient calculation and real-time prediction, insufficient data and model generalization. These advantages make random forest + MLP a powerful and effective air conditioning load prediction tool, providing reliable technical support for energy management of data centers and buildings.

[0034] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0035] Figure 1 is a flow chart of an air conditioning load prediction method provided by an embodiment of the present disclosure, such as Figure 1 As shown in FIG. 1 , the specific process of the air conditioning load forecasting method includes: S110, obtaining environmental factor data of the air-conditioned area within a preset period of time.

[0036] The environmental factor data include: current radiation intensity, current temperature and current relative humidity. The current radiation intensity may include: solar diffuse radiation intensity, solar direct radiation intensity and solar total radiation intensity; the current temperature may include: current dry bulb temperature and current wet bulb temperature.

[0037] In addition, environmental factor data may also include: current wind speed, current air pressure value, current rainfall, week factor, hour factor, etc.

[0038] It should be noted that the air-conditioning area is the range area where the air-conditioning equipment is installed, that is, the air-conditioning equipment is installed in the range area; in this case, the air-conditioning equipment in the range area can be in the on state or in the off state, and this embodiment does not make any specific limitations on this.

[0039] S120: Determine the load category of the air-conditioning area within a preset period based on the environmental factor data and the initial load prediction model.

[0040] The initial load prediction model is used to make an initial prediction of the air conditioning load through environmental factor data, that is, the initial load prediction model makes a rough prediction of the air conditioning load.

[0041] The load categories of the air-conditioned area within the preset time period include: a first load category and a second load category. The first load category is used to describe that the load prediction value of the air-conditioned area within the preset time period is greater than the preset load threshold, and the second load category is used to describe that the load prediction value of the air-conditioned area within the preset time period is less than or equal to the preset load threshold.

[0042] It should be noted that the load prediction value of the air-conditioned area within the preset time period referred to in this embodiment is greater than the preset load threshold, that is, the load prediction value of the air-conditioned area within the preset time period is considered to be not 0; the load prediction value of the air-conditioned area within the preset time period is less than or equal to the preset load threshold, that is, the load prediction value of the air-conditioned area within the preset time period is considered to be 0 or almost 0.

[0043] Optionally, based on the environmental factor data and the initial load forecasting model, the load category of the air-conditioning area within the preset period is determined, including: The environmental factor data is input into the initial load prediction model, and the initial load prediction value corresponding to the air-conditioned area within the preset time period is determined based on the output result of the initial load prediction model; if the initial load prediction value corresponding to the air-conditioned area within the preset time period is greater than the preset load threshold, the load category of the air-conditioned area within the preset time period is determined to be the first load category; if the initial load prediction value corresponding to the air-conditioned area within the preset time period is less than or equal to the preset load threshold, the load category of the air-conditioned area within the preset time period is determined to be the second load category.

[0044] The initial load prediction model may be an MLP model. By using the MLP model as the initial prediction of the air-conditioning load, it is convenient to effectively determine the load range in the air-conditioning prediction process.

[0045] It should be noted that if the initial load prediction value corresponding to the air-conditioned area in the preset period is less than or equal to the preset load threshold, it is determined that the air-conditioned area is not turned on in the preset period; if the initial load prediction value corresponding to the air-conditioned area in the preset period is greater than the preset load threshold, it is determined that the air-conditioned area is turned on in the preset period. Thus, the MLP model is used to determine the load category of the air-conditioned area in the preset period, which is convenient for the subsequent processing of the air-conditioning load prediction.

[0046] S130. Determine a load prediction result of the air-conditioned area within the preset period based on the category to which the load of the air-conditioned area within the preset period belongs.

[0047] The load prediction result of the air-conditioning area within the preset time period includes: the air-conditioning cooling load value or the air-conditioning heating load value.

[0048] Optionally, based on the category to which the load of the air-conditioned area belongs during the preset period, determining the load forecast result of the air-conditioned area during the preset period includes: If the load category of the air-conditioned area within the preset time period belongs to the first load category, obtain the target load prediction model; input the environmental factor data into the target load prediction model, and determine the target load prediction value of the air-conditioned area within the preset time period based on the output result of the target load prediction model; determine the target load prediction value as the load prediction result of the air-conditioned area within the preset time period.

[0049] The target load prediction model is used to make a final prediction of the air conditioning load through the environmental factor data. That is, the target load prediction model is a detailed prediction of the air conditioning load. The target load prediction model can be a random forest model.

[0050] Therefore, when the MLP model is used to determine whether to turn on the air conditioner in the air-conditioning area within the preset time period, it is convenient to accurately zone the air-conditioning load through the target load prediction model.

[0051] Optionally, based on the load category of the air-conditioned area in the preset period, the load prediction result of the air-conditioned area in the preset period is determined, including: if the load category of the air-conditioned area in the preset period is the second load category, then the initial load prediction value corresponding to the air-conditioned area in the preset period is determined as the load prediction result of the air-conditioned area in the preset period. Therefore, when it is determined through the MLP model that the air-conditioned area in the preset period is not turned on, there is no need to predict the air-conditioning load again, and the initial load prediction value corresponding to the air-conditioned area in the preset period can be directly used as the load prediction result of the air-conditioned area in the preset period to avoid resource loss.

[0052] Optionally, obtain a target load forecasting model, including: Acquire sample training data; perform data preprocessing on the sample training data to obtain sample factor data, wherein the preprocessing includes: missing data processing and / or abnormal data processing; train a random forest network model based on the sample factor data to obtain a target load forecasting model.

[0053] The sample training data includes: sample radiation intensity, sample temperature and sample relative humidity. The sample radiation intensity may include: the solar scattered radiation intensity at the previous moment, the solar direct radiation intensity at the previous moment and the solar total radiation intensity at the previous moment; the sample temperature may include: the dry bulb temperature at the previous moment (such as the dry bulb temperature three hours ago) and the wet bulb temperature at the previous moment (such as the wet bulb temperature three hours ago).

[0054] In addition, environmental factor data may also include: wind speed at the previous moment, air pressure value at the previous moment, rainfall at the previous moment, week factor, hour factor, etc.

[0055] After obtaining the sample training data, it can be stored in the database to facilitate subsequent operations such as data cleaning.

[0056] Therefore, the random forest network model is trained through historical environmental data to effectively obtain the target load forecasting model.

[0057] In addition, the training process of the initial load forecasting model may include: training the MLP network model based on sample factor data to obtain the initial load forecasting model.

[0058] Optionally, data preprocessing is performed on the sample training data to obtain sample factor data, including: Check whether there is missing data in the sample training data; if it is determined that there is missing data at the target position in the sample training data, fill the missing data at the target position based on the adjacent position data corresponding to the target position; check whether there is abnormal data in the sample training data; if it is determined that there is abnormal data in the sample training data, smooth the abnormal data to obtain sample factor data.

[0059] Among them, when processing missing data, if there are many missing values ​​at the same time, the row of data can be directly deleted. If the moment (that is, the collection moment corresponding to the target location collection) is a missing value, the average value of the two moments before and after it is used to fill the missing data.

[0060] When smoothing abnormal data, it is necessary to determine the load sequence x ( d , t ) (i.e., sample training data, t =0,1,2,...,23, indicating the time of day t a moment, d The abnormal data in the air conditioning load value of a certain day is detected and optimized. The specific steps are as follows.

[0061] Step 1: Read the load sequence x ( d , t ),like t =0, then calculate N Daily load point t The mean E ( t ) and standard deviation σ ( t ) ; Calculate the deviation rate of each load point η ( d , t ), and obtain the maximum deviation rate of the load point η max ( t );judge η max ( t ) is greater than C (a preset value). If not, go to step 2; if so, go to step 3.

[0062] Step 2: Judgement t Is it greater than 23? If so, go to step 4; if not, end.

[0063] Step 3: d =1.

[0064] Step 4: Judgement d Is it greater than N? If so, go to step five; if not, go to step six.

[0065] Step 5: t = t +1, continue to perform calculations N Daily load point t The mean E ( t ) and standard deviation σ ( t ) and subsequent steps.

[0066] Step 6: Judgement η ( d , t ) is greater than C; if so, execute step seven; if not, execute step eight.

[0067] Step 7: Calculate and replace x ( d , t ), and update d , d = d +1; proceed to step 4.

[0068] In addition, after obtaining the sample factor data, normalization processing can be performed. The normalization processing can refer to formula (1) or formula (2).

[0069] ; (1) ; (2) x min is the minimum value in the sample factor data, x max is the maximum value in the sample factor data.

[0070] Formula (1) can normalize the sample factor data to the data range [0,1], and formula (2) can normalize the sample factor data to the data range [-1,1].

[0071] Optionally, before performing data preprocessing on the sample training data to obtain the sample factor data, the method further includes: obtaining the data correlation between every two data in the sample training data; and updating the sample training data based on the data correlation between every two data in the sample training data.

[0072] Among them, the data correlation between each two data in the sample training data can effectively reflect the statistically significant relationship between the two data. For example, if the data correlation between two data is less than 0.05, it is determined that the two data are statistically insignificant; if the data correlation between two data is greater than or equal to 0.05, it is determined that the two data are statistically significant. 2A is a heat map of the correlation coefficient between data.

[0073] Based on the data correlation between every two data in the sample training data, the sample training data is updated, which may include: deleting data with a data correlation less than 0.05 from the sample training data, so as to improve the data compactness in the sample training data.

[0074] In addition, the analysis of the correlation between hourly load and each feature and the subsequent work based on the correlation analysis. The analysis shows that among the real-time collected features of radiation intensity, temperature, and relative humidity, the radiation intensity has the strongest correlation; the embedding dimension of this feature can be increased to enable it to obtain more adequate training. In the mapping of time-related features of date factor and week factor, the correlation of date factor is stronger, and the week factor shows a weak negative correlation; the change of hourly load is indeed based on a 24h cycle, but its load value fluctuates violently between different weeks. The existing week mapping is relatively rigid, and it will be converted into two features of day_of_week and is_holiday in the future.

[0075] This embodiment obtains environmental factor data of the air-conditioned area within a preset period of time, and the environmental factor data includes: current radiation intensity, current temperature and current relative humidity; based on the environmental factor data and the initial load prediction model, the load category of the air-conditioned area within the preset period of time is determined, and the initial load prediction model is used to make an initial prediction of the air-conditioning load through the environmental factor data; based on the load category of the air-conditioned area within the preset period of time, the load prediction result of the air-conditioned area within the preset period of time is determined; wherein, the load category of the air-conditioned area within the preset period of time includes: a first load category and a second load category, the first load category is used to describe that the load prediction value of the air-conditioned area within the preset period of time is greater than the preset load threshold, and the second load category is used to describe that the load prediction value of the air-conditioned area within the preset period of time is less than or equal to the preset load threshold. Thus, by introducing machine learning methods, a new technical path is provided for air-conditioning load prediction, which facilitates accurate prediction of air-conditioning load.

[0076] In addition, the method of this embodiment also provides the comparison results of the initial load forecasting model + target load forecasting model, the time series model, and the random forest model. Figure 2B This is the effect diagram of the time series model. As the model depth increases and the embedding dimension increases, the model tends to underfit and the performance decreases. To fully utilize the performance of the time series model, more data and feature processing are needed. Figure 2C This is the effect diagram of the random forest model. The evaluation index data are: R_square=0.9730055645581513, MAE=1333.390221454934, MSE=4816749.249375275, RMSE=2194.7093769734697.

[0077] Figure 2D This is the effect diagram of the initial load forecasting model + target load forecasting model. The evaluation index data are: R_square=0.9911627436154883, MAE=675.2799131578946, MSE=1576874.9136588215, RMSE=1255.736801108744.

[0078] In this embodiment, two machine learning methods, random forest and multi-layer perceptron, are innovatively combined for air conditioning load prediction; random forest is used for preliminary feature selection and importance analysis, and MLP is used for deep learning and refined prediction. Before the prediction model is built, data preprocessing and feature engineering are performed to ensure the quality of input data and the prediction accuracy of the model, including steps such as data cleaning, normalization, feature extraction, and feature selection. The data is preliminarily screened for features and ranked for importance through random forest, and then MLP is used for model training and optimization, and cross-validation, regularization and other methods are used to improve the generalization ability and stability of the model.

[0079] In this embodiment, random forest can capture complex nonlinear relationships and effectively improve the accuracy of prediction by integrating multiple decision trees; MLP further enhances the model's ability to fit complex relationships through a multi-layer neural network structure; the combination of the two significantly improves the accuracy of air-conditioning load prediction and can more accurately reflect the actual load changes. Random forest has strong robustness to noise and outliers in the data, and can maintain high prediction performance in the face of imperfect data; MLP further improves the stability of the model through regularization techniques (such as L2 regularization and Dropout). The combination of the two methods enables the prediction model to operate stably in various complex and dynamic environments. It can comprehensively consider various factors that affect air-conditioning load, such as ambient temperature, humidity, building characteristics, usage, etc. Through the feature importance analysis of random forest and the deep learning ability of MLP, the model can accurately identify and quantify the impact of various factors on air-conditioning load, thereby providing more comprehensive prediction results. The air conditioning load has significant temporal dynamic characteristics. The present invention combines the methods of random forest and MLP, which can effectively capture and predict the changes in load over time; in particular, the advantage of MLP in processing time series data enables the model to better cope with the dynamic changes in load and improve the real-time and accuracy of prediction. Through accurate load prediction, the present invention can provide data support for the intelligent control of the air-conditioning system, realize automatic adjustment of air-conditioning equipment, and optimize energy use; this not only improves the intelligence level and response speed of the system, but also enhances the adaptive ability of the air-conditioning system and provides a better user experience. It can accurately predict the air-conditioning load, optimize the operation strategy of the air-conditioning system, reduce energy consumption, and reduce operating costs; by improving energy efficiency and reducing unnecessary energy waste, the present invention helps to achieve the goal of energy conservation and emission reduction, and meets the requirements of sustainable development.

[0080] Compared with existing technologies, air conditioning load is affected by multiple factors, and the relationships between these factors are complex. Traditional methods find it difficult to capture these nonlinear relationships. Random forests can capture complex nonlinear relationships by integrating multiple decision trees. MLP further enhances the modeling ability of complex relationships through a multi-layer neural network structure. Air conditioning load is affected by multiple factors such as environmental factors (such as temperature and humidity), building characteristics, and usage. Traditional methods find it difficult to comprehensively consider these factors. Random forests can process high-dimensional feature data and identify the impact of various factors on the load through feature importance analysis. MLP uses deep learning to comprehensively consider multiple input features and improve prediction accuracy. Actual data often contains noise and outliers, which affect the performance and stability of the model. Random forests naturally have noise resistance by integrating multiple trees. MLP can process noisy data through regularization techniques (such as L2 regularization and Dropout) to improve the robustness of the model. Air conditioning load has significant temporal dynamic characteristics, and the prediction model needs to accurately capture the load that changes over time; MLP can use recurrent neural network (RNN) structure or long short-term memory network (LSTM) to process temporal dynamic characteristics in time series prediction; random forest can capture time series characteristics through sliding window method. Air conditioning load prediction requires efficient computing power to achieve real-time prediction, and traditional methods may not perform well under large data volumes; random forest can improve the efficiency of training and prediction through parallel computing; MLP can use GPU to accelerate training and reasoning to meet real-time prediction needs. In the case of insufficient data or uneven data distribution, the model is prone to overfitting and insufficient generalization ability; random forest enhances generalization ability through Bootstrap sampling and integration methods; MLP improves performance in the case of insufficient data through data enhancement technology and regularization methods. A single model may not perform well in certain specific scenarios, affecting the overall prediction performance; this embodiment combines a hybrid model of random forest and MLP to make full use of the high interpretability and noise resistance of random forest and the powerful nonlinear fitting ability of MLP to achieve higher prediction accuracy and robustness.

[0081] Figure 3 This is a schematic diagram of the structure of an air conditioning load prediction device provided in this embodiment, wherein the air conditioning load prediction device may include: an acquisition module 310 , a first determination module 320 and a second determination module 330 .

[0082] The acquisition module 310 is used to acquire environmental factor data of the air-conditioned area within a preset period of time, and the environmental factor data includes: current radiation intensity, current temperature and current relative humidity; A first determination module 320 is used to determine the load category of the air-conditioning area within a preset period based on the environmental factor data and the initial load prediction model, wherein the initial load prediction model is used to make an initial prediction of the air-conditioning load based on the environmental factor data; The second determination module 330 is used to determine the load prediction result of the air-conditioning area within the preset period of time based on the category to which the load of the air-conditioning area belongs within the preset period of time; Among them, the load categories of the air-conditioned area within the preset time period include: a first load category and a second load category. The first load category is used to describe that the load prediction value of the air-conditioned area within the preset time period is greater than the preset load threshold, and the second load category is used to describe that the load prediction value of the air-conditioned area within the preset time period is less than or equal to the preset load threshold.

[0083] Optionally, the first determining module 320 is specifically configured to: The environmental factor data is input into the initial load prediction model, and the initial load prediction value corresponding to the air-conditioned area within the preset time period is determined based on the output result of the initial load prediction model; if the initial load prediction value corresponding to the air-conditioned area within the preset time period is greater than the preset load threshold, the load category of the air-conditioned area within the preset time period is determined to be the first load category; if the initial load prediction value corresponding to the air-conditioned area within the preset time period is less than or equal to the preset load threshold, the load category of the air-conditioned area within the preset time period is determined to be the second load category.

[0084] Optionally, the second determination module 330 includes: an acquisition unit and a determination unit; An acquisition unit, configured to acquire a target load prediction model if the load category of the air-conditioning area within a preset period belongs to the first load category, wherein the target load prediction model is used to make a final prediction of the air-conditioning load through environmental factor data; The determination unit is used to input the environmental factor data into the target load prediction model, and determine the target load prediction value of the air-conditioned area within a preset period based on the output result of the target load prediction model; the target load prediction value is determined as the load prediction result of the air-conditioned area within the preset period.

[0085] Optionally, the second determining module 330 is specifically configured to: If the load category of the air-conditioned area in the preset period belongs to the second load category, an initial load prediction value corresponding to the air-conditioned area in the preset period is determined as a load prediction result of the air-conditioned area in the preset period.

[0086] Optionally, obtain the unit, specifically used to: Obtain sample training data, the sample training data includes: sample radiation intensity, sample temperature and sample relative humidity; perform data preprocessing on the sample training data to obtain sample factor data, the preprocessing includes: missing data processing and / or abnormal data processing; train the random forest network model based on the sample factor data to obtain a target load prediction model.

[0087] Optionally, obtain the unit, specifically used to: Check whether there is missing data in the sample training data; if it is determined that there is missing data at the target position in the sample training data, fill the missing data at the target position based on the adjacent position data corresponding to the target position; check whether there is abnormal data in the sample training data; if it is determined that there is abnormal data in the sample training data, smooth the abnormal data to obtain sample factor data.

[0088] Optionally, it also includes: an updating unit; An acquisition unit, used for acquiring the data correlation between every two data in the sample training data; The updating unit is used to update the sample training data based on the data correlation between every two data in the sample training data.

[0089] The air conditioning load prediction device provided in the present disclosure can execute the above method embodiments. Its specific implementation principles and technical effects can be found in the above method embodiments, and the present disclosure will not repeat them here.

[0090] The present application also provides a computer device. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0091] The computer device includes a memory 410 and a processor 420 that are connected to each other through a system bus. It should be noted that the figure only shows a computer device with a memory 410 and a processor 420, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable gate arrays (Field-Programmable Gate Array, FPGA), digital processors (Digital Signal Processor, DSP), embedded devices, etc.

[0092] Computer devices can be computing devices such as desktop computers, notebooks, PDAs, and cloud servers. Computer devices can interact with users through keyboards, mice, remote controls, touch pads, or voice control devices.

[0093] The memory 410 includes at least one type of readable storage medium, and the readable storage medium includes a non-volatile memory or a volatile memory, such as a flash memory, a hard disk, a multimedia card, a card-type memory (such as an SD or DX memory, etc.), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc., and the RAM may include a static RAM or a dynamic RAM. In some embodiments, the memory 410 may be an internal storage unit of a computer device, such as a hard disk or a memory of the computer device. In other embodiments, the memory 410 may also be an external storage device of a computer device, such as a plug-in hard disk, a smart memory card (SMC), a secure digital (SD) card, or a flash card (Flash Card), etc., equipped on the computer device. Of course, the memory 410 may also include both an internal storage unit of the computer device and an external storage device thereof. In this embodiment, the memory 410 is generally used to store an operating system and various application software installed on the computer device, such as the program code of the above method, etc. In addition, the memory 410 may also be used to temporarily store various data that have been output or are to be output.

[0094] The processor 420 is generally used to perform the overall operation of the computer device. In this embodiment, the memory 410 is used to store program codes or instructions, the program code includes computer operation instructions, and the processor 420 is used to execute the program codes or instructions stored in the memory 410 or process data, such as running the program code of the above method.

[0095] In this article, the bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus system can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0096] Another embodiment of the present application also provides a computer-readable medium, which may be a computer-readable signal medium or a computer-readable medium. A processor in a computer reads a computer-readable program code stored in the computer-readable medium, so that the processor can execute the functional actions specified in each step or a combination of steps in the above method; and generate a device for implementing the functional actions specified in each block or a combination of blocks in the block diagram.

[0097] Computer-readable media include but are not limited to electronic, magnetic, optical, electromagnetic, infrared memory or semiconductor systems, devices or apparatuses, or any appropriate combination of the foregoing, the memory is used to store program codes or instructions, the program codes include computer operating instructions, and the processor is used to execute the program codes or instructions of the above methods stored in the memory.

[0098] For the definitions of memory and processor, please refer to the description of the aforementioned computer device embodiment and will not be repeated here.

[0099] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of modules or units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0100] Each functional unit or module in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0101] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) or a processor (processor) to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., and other media that can store program codes.

[0102] In the claims, any reference symbols placed between brackets shall not be construed as limiting the claims. The word "comprising" described in the present application does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present application can be implemented with the aid of hardware comprising several different elements and with the aid of a suitably programmed computer. In a unit claim that lists a number of devices, several units of these devices may be embodied by the same hardware item. The use of first, second, and third, etc. does not indicate any order, and these words may be interpreted as names. The steps in the above embodiments, unless otherwise specified, should not be understood as limitations on the order of execution.

[0103] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for predicting air conditioning load, characterized in that: include: Acquire environmental factor data of the air-conditioned area within a preset period of time, wherein the environmental factor data includes: current radiation intensity, current temperature and current relative humidity; Determine the category to which the load of the air-conditioning area in the preset time period belongs based on the environmental factor data and an initial load prediction model, wherein the initial load prediction model is used to make an initial prediction of the air-conditioning load based on the environmental factor data; Determining a load prediction result of the air-conditioned area within the preset period of time based on the category to which the load of the air-conditioned area within the preset period of time belongs; Among them, the categories to which the load of the air-conditioned area within the preset time period belongs include: a first load category and a second load category, the first load category is used to describe that the load prediction value of the air-conditioned area within the preset time period is greater than a preset load threshold, and the second load category is used to describe that the load prediction value of the air-conditioned area within the preset time period is less than or equal to the preset load threshold.

2. The method according to claim 1, characterized in that: The step of determining the load category of the air-conditioning area within the preset period based on the environmental factor data and the initial load prediction model includes: Inputting the environmental factor data into the initial load prediction model, and determining the initial load prediction value corresponding to the air-conditioning area within the preset time period based on the output result of the initial load prediction model; If the initial load prediction value corresponding to the air-conditioned area within the preset time period is greater than the preset load threshold, determining that the load category of the air-conditioned area within the preset time period belongs to the first load category; If the initial load prediction value corresponding to the air-conditioned area within the preset time period is less than or equal to the preset load threshold, it is determined that the load category of the air-conditioned area within the preset time period belongs to the second load category.

3. The method according to claim 1, characterized in that The step of determining the load prediction result of the air-conditioned area within the preset period of time based on the category to which the load of the air-conditioned area within the preset period of time belongs comprises: If the load category of the air-conditioning area within the preset time period belongs to the first load category, obtaining a target load prediction model, wherein the target load prediction model is used to make a final prediction of the air-conditioning load through the environmental factor data; Inputting the environmental factor data into the target load prediction model, and determining the target load prediction value of the air-conditioning area within the preset time period based on the output result of the target load prediction model; The target load forecast value is determined as a load forecast result of the air-conditioning area within the preset time period.

4. The method according to claim 1, characterized in that: The step of determining the load prediction result of the air-conditioned area within the preset period of time based on the category to which the load of the air-conditioned area within the preset period of time belongs comprises: If the load category of the air-conditioned area within the preset period belongs to the second load category, an initial load forecast value corresponding to the air-conditioned area within the preset period is determined as a load forecast result of the air-conditioned area within the preset period.

5. The method according to claim 3, characterized in that: The obtaining of the target load prediction model comprises: Acquire sample training data, wherein the sample training data includes: sample radiation intensity, sample temperature and sample relative humidity; Performing data preprocessing on the sample training data to obtain sample factor data, wherein the preprocessing includes: missing data processing and / or abnormal data processing; The random forest network model is trained based on the sample factor data to obtain the target load prediction model.

6. The method according to claim 5, characterized in that The step of performing data preprocessing on the sample training data to obtain sample factor data includes: Detecting whether there is missing data in the sample training data; If it is determined that there is missing data at the target position in the sample training data, the missing data at the target position is filled based on the adjacent position data corresponding to the target position; Detecting whether there is abnormal data in the sample training data; If it is determined that abnormal data exists in the sample training data, the abnormal data is smoothed to obtain the sample factor data.

7. The method according to claim 5, characterized in that Before the sample training data is preprocessed to obtain sample factor data, the method further includes: Obtaining the data correlation between every two data in the sample training data; Based on the data correlation between every two data in the sample training data, the sample training data is updated.

8. An air conditioning load prediction device, characterized in that: include: An acquisition module is used to acquire environmental factor data of the air-conditioned area within a preset period of time, wherein the environmental factor data includes: current radiation intensity, current temperature and current relative humidity; A first determination module, configured to determine the load category of the air-conditioning area within the preset period based on the environmental factor data and an initial load prediction model, wherein the initial load prediction model is configured to perform an initial prediction of the air-conditioning load using the environmental factor data; A second determination module, configured to determine a load prediction result of the air-conditioned area within the preset period of time based on the category to which the load of the air-conditioned area within the preset period of time belongs; Among them, the categories to which the load of the air-conditioned area within the preset time period belongs include: a first load category and a second load category, the first load category is used to describe that the load prediction value of the air-conditioned area within the preset time period is greater than a preset load threshold, and the second load category is used to describe that the load prediction value of the air-conditioned area within the preset time period is less than or equal to the preset load threshold.

9. A computer device, characterized in that: The method comprises a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the air conditioning load prediction method as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the air conditioning load prediction method as described in any one of claims 1 to 7 is implemented.

Images