Air conditioner cooling load prediction method and system

Abstract

The invention relates to the technical field of air conditioner cooling load prediction, and discloses an air conditioner cooling load prediction method and system. The method comprises the following steps: acquiring temperature, humidity and CO2 data through an environment sensor network and carrying out standardization processing; establishing a time slice matrix based on the environmental parameters and the personnel density, and obtaining user behavior characteristics; constructing a wall heat storage model by combining the environmental parameters, and calculating instantaneous heat flow; establishing a spatial downscaling model according to the heat flow and meteorological data; time sequence features are extracted based on local weather and heat flow data, and the cooling load is predicted; and a time-phased refrigeration control instruction is generated accordingly. By establishing a data processing and analysis process, factors such as building microclimate characteristics, user behavior modes and wall dynamic thermal characteristics are comprehensively considered, and more accurate prediction of the cooling load of the air conditioner is realized.

Description

Technical Field

[0001] This application relates to the technical field of air-conditioning cooling load prediction, and particularly to an air-conditioning cooling load prediction method and system. Background Art

[0002] With the continuous improvement of the requirements for building energy consumption management, the intelligent control of air-conditioning systems has become an important means of energy conservation and emission reduction. The current air-conditioning cooling load prediction methods mainly rely on historical building energy consumption data and meteorological forecast data, and use statistical analysis or machine learning algorithms for load prediction. These methods analyze the load change patterns in historical data, combine weather forecast information, and establish prediction models to estimate the cooling load demand in future time periods, providing a decision-making basis for the operation control of air-conditioning systems.

[0003] However, the existing air-conditioning cooling load prediction methods have problems with insufficient prediction accuracy. The main reason is that these methods do not fully consider the influence of building microclimate characteristics and user behavior patterns on the cooling load. Traditional prediction methods directly use regional meteorological data, ignoring the particularity of the local meteorological environment around the building; at the same time, there is a lack of in-depth analysis of the activity patterns of building users, resulting in a large deviation between the prediction results and the actual cooling load demand. In addition, the existing methods do not adequately consider the dynamic thermal characteristics of building walls, and cannot accurately reflect the influence of building heat storage and thermal inertia on the cooling load. Summary of the Invention

[0004] This application provides an air-conditioning cooling load prediction method and system, which are used to comprehensively consider factors such as building microclimate characteristics, user behavior patterns, and wall dynamic thermal characteristics by establishing a data processing and analysis process, so as to achieve a more accurate prediction of the air-conditioning cooling load, thereby providing a reliable control basis for the energy-saving operation of the air-conditioning system.

[0005] In the first aspect, this application provides an air-conditioning cooling load prediction method, and the air-conditioning cooling load prediction method includes: collecting raw data of indoor and outdoor temperature, humidity, and CO2 concentration through an environmental sensor network, performing outlier rejection and sliding window smoothing processing on the raw data to obtain standardized environmental parameters; establishing a time slice distribution matrix based on the standardized environmental parameters and personnel density data, and obtaining user behavior characteristics through hierarchical clustering calculation; constructing a wall heat storage model based on the user behavior characteristics and standardized environmental parameters, and obtaining building instantaneous heat flux data through dynamic thermal resistance calculation; establishing a spatial downscaling model based on the building instantaneous heat flux data and meteorological forecast data, and obtaining local meteorological parameters through microclimate correction; constructing a prediction feature matrix based on the local meteorological parameters and building instantaneous heat flux data, and obtaining the cooling load prediction value through time series feature extraction and attention weight calculation; calculating the refrigeration power adjustment coefficient of the air-conditioning system according to the cooling load prediction value, and generating a sub-period cooling capacity control instruction.

[0006] In a second aspect, the present application provides an air-conditioning cooling load prediction system, and the air-conditioning cooling load prediction system includes: An acquisition module, configured to collect raw data of indoor and outdoor temperature, humidity, and CO2 concentration through an environmental sensing network, remove outliers from the raw data, and perform sliding window smoothing processing to obtain standardized environmental parameters; A calculation module, configured to establish a time slice distribution matrix according to the standardized environmental parameters and personnel density data, and obtain user behavior characteristics through hierarchical clustering calculation; A construction module, configured to construct a wall heat storage model according to the user behavior characteristics and the standardized environmental parameters, and obtain building instantaneous heat flux data through dynamic thermal resistance calculation; A correction module, configured to establish a spatial downscaling model according to the building instantaneous heat flux data and meteorological forecast data, and obtain local meteorological parameters through microclimate correction; A prediction module, configured to construct a prediction feature matrix according to the local meteorological parameters and the building instantaneous heat flux data, and obtain a cooling load prediction value through time series feature extraction and attention weight calculation; A generation module, configured to calculate a refrigeration power adjustment coefficient of the air-conditioning system according to the cooling load prediction value, and generate a time-segmented cooling capacity control instruction.

[0007] In a third aspect of the present invention, there is provided a computer device, including: a memory and at least one processor, wherein instructions are stored in the memory; the at least one processor calls the instructions in the memory to enable the computer device to execute the above-mentioned air-conditioning cooling load prediction method.

[0008] In a fourth aspect of the present invention, there is provided a computer-readable storage medium, wherein instructions are stored in the computer-readable storage medium, and when the instructions are run on a computer, the computer is enabled to execute the above-mentioned air-conditioning cooling load prediction method.

[0009] In the technical solution provided by this application, the original data of indoor and outdoor temperature, humidity, and CO2 concentration are collected through an environmental sensing network, and outlier removal and sliding window smoothing processing are performed, improving the quality and reliability of environmental data and providing an accurate data basis for subsequent prediction. A time slice distribution matrix is established based on standardized environmental parameters and personnel density data, and user behavior characteristics are obtained through hierarchical clustering calculation, realizing the accurate identification of building usage patterns. A wall heat storage model is constructed based on user behavior characteristics and standardized environmental parameters, and the instantaneous heat flow data of the building are obtained through dynamic thermal resistance calculation, accurately reflecting the dynamic heat transfer characteristics of the building wall. A spatial downscaling model is established using the instantaneous heat flow data of the building and meteorological forecast data, and local meteorological parameters are obtained through microclimate correction, solving the problem that regional meteorological data cannot accurately reflect the environmental characteristics around the building. A prediction feature matrix is constructed using the local meteorological parameters and the instantaneous heat flow data of the building, and the cooling load prediction value is obtained through time series feature extraction and attention weight calculation, making full use of the advantages of deep learning algorithms in time series data processing and improving the prediction accuracy. The cooling power adjustment coefficient of the air conditioning system is calculated based on the cooling load prediction value, and a sub-period cooling capacity control instruction is generated, realizing the precise control of the air conditioning system. In a specific air conditioning load prediction field, this solution establishes a data processing and analysis process by introducing algorithms such as hierarchical clustering and deep learning, and combining the physical characteristics and usage characteristics of the building, significantly improving the prediction accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for description in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0011] Figure 1 It is a schematic diagram of an embodiment of the air conditioning cooling load prediction method in the embodiments of this application; Figure 2 It is a schematic diagram of the process of obtaining the cooling load prediction value through time series feature extraction and attention weight calculation in the embodiments of this application; Figure 3 It is a schematic diagram of the load level in the embodiments of this application; Figure 4 It is a schematic diagram of an embodiment of the air conditioning cooling load prediction system in the embodiments of this application; Figure 5 It is a schematic block diagram of the structure of a computer device in the embodiments of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0012] The embodiments of the present application provide an air-conditioning cooling load prediction method and system. The terms "first", "second", "third", "fourth", etc. (if any) in the specification, claims and the above-mentioned drawings of the present application are used to distinguish similar objects and do not necessarily describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the term "including" or "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0013] For ease of understanding, the specific process of the embodiments of the present application will be described below. Please refer to Figure 1 , an embodiment of the air-conditioning cooling load prediction method in the embodiments of the present application includes: Step S101: Collect the original data of indoor and outdoor temperature, humidity, and CO2 concentration through the environmental sensor network, perform outlier removal and sliding window smoothing processing on the original data to obtain standardized environmental parameters; Step S102: Establish a time slice distribution matrix based on the standardized environmental parameters and the personnel density data, and obtain user behavior characteristics through hierarchical clustering calculation; Step S103: Construct a wall heat storage model based on the user behavior characteristics and the standardized environmental parameters, and obtain the instantaneous heat flow data of the building through dynamic thermal resistance calculation; Step S104: Establish a spatial downscaling model based on the instantaneous heat flow data of the building and the meteorological forecast data, and obtain local meteorological parameters through microclimate correction; Step S105: Construct a prediction feature matrix based on the local meteorological parameters and the instantaneous heat flow data of the building, and obtain the cooling load prediction value through time series feature extraction and attention weight calculation; Step S106: Calculate the refrigeration power adjustment coefficient of the air-conditioning system according to the cooling load prediction value, and generate a time-segmented cooling capacity control instruction.

[0014] It can be understood that the execution subject of the present application can be an air-conditioning cooling load prediction system, or a terminal or a server. Specifically, it is not limited here. The embodiments of the present application are described by taking the server as the execution subject as an example.

[0015] Specifically, indoor and outdoor environmental data are collected at fixed time intervals through an environmental sensing network. Temperature sensors, humidity sensors, and CO2 concentration sensors are arranged indoors and outdoors in the environmental sensing network, and data is collected once every minute. Outlier rejection processing is performed on the collected raw data. By calculating the interquartile range of temperature, humidity, and CO2 concentration data, outlier data points outside the range are removed. The data after outlier rejection is then subjected to a moving average process with a 5-minute time window to reduce the impact of data fluctuations and obtain standardized environmental parameter data. Then, a time slice distribution matrix is constructed based on the standardized environmental parameters and personnel density data. The 24-hour period is divided into 48 time slices at 30-minute intervals, and statistical analysis is performed on the environmental parameters and personnel density data within each time slice. Through hierarchical clustering calculation, the time slice data is divided into different user behavior scenarios, such as working periods and rest periods, to obtain user behavior characteristics. User behavior characteristics include information such as the change rate of personnel density and the change trend of environmental parameters.

[0016] A wall heat storage model is constructed based on user behavior characteristics and standardized environmental parameters. The wall heat storage model is a mathematical model that describes the heat conduction characteristics of building walls. By dividing the building walls into multiple heat transfer nodes, calculating the heat conduction coefficient between the nodes, and establishing a node heat transfer network. The temperature field distribution of the wall is calculated according to the heat flow influence factor, and the change in heat capacity is calculated in combination with the characteristics of the wall material to obtain dynamic thermal resistance data. The instantaneous heat flow data of the building can be obtained through dynamic thermal resistance calculation. Then, a spatial downscaling model is established based on the instantaneous heat flow data of the building and meteorological forecast data. The spatial downscaling model is a spatial mapping relationship matrix that converts regional-scale meteorological forecast data into microscale meteorological data around the building. By dividing the space around the building into grids, analyzing the correlation between heat flow data and meteorological data, and establishing a microclimate influence weight, local meteorological parameters are obtained.

[0017] A prediction feature matrix is constructed based on local meteorological parameters and the instantaneous heat flow data of the building. The prediction feature matrix includes characteristic parameters such as the temperature change rate, heat flow change rate, and pressure gradient. By extracting time series characteristics to analyze the change law of characteristic parameters over time, calculating the characteristic importance index and performing attention weight calculation, the cooling load prediction value is obtained. The cooling power adjustment coefficient of the air conditioning system is calculated according to the cooling load prediction value. The cooling load prediction value is divided into multiple time periods over 24 hours, the cooling load of each time period is classified, and the corresponding cooling power reference value is calculated. The cooling power adjustment coefficient is obtained through dynamic compensation, and then the cooling capacity control command for each time period is generated.

[0018] For example, a certain office building uses this cooling load prediction method for air conditioning control. Among the original data collected by the environmental sensing network, there is an abnormal temperature value of 35°C due to sensor failure (the normal room temperature is between 20-26°C). This abnormal value is removed through quartile range calculation, and the remaining data is processed by a 5-minute moving average to obtain standardized environmental parameters. After dividing 24 hours into 48 time slices, it is found through analysis that the personnel density is relatively high during 9:00-18:00, and the temperature and humidity change significantly, and this period is clustered as the working period. Based on these user behavior characteristics and combined with the heat conduction characteristics of the multi-layer wall structure, a heat transfer network is established, and the dynamic thermal resistance changes in different time periods are calculated. By establishing a spatial downscaling model, the regional meteorological forecast data is converted into local meteorological parameters around the building. Based on these data, a prediction matrix including features such as temperature gradient and heat flow change is constructed, and the cooling load prediction values for each time period are obtained through time series analysis. According to the prediction values, corresponding refrigeration control instructions are generated to achieve intelligent adjustment of the air conditioning system.

[0019] In the embodiment of the present application, the original data of indoor and outdoor temperature, humidity, and CO2 concentration are collected through the environmental sensing network, and outlier removal and sliding window smoothing processing are performed, which improves the quality and reliability of the environmental data and provides an accurate data basis for subsequent predictions. A time slice distribution matrix is established according to the standardized environmental parameters and personnel density data, and user behavior characteristics are obtained through hierarchical clustering calculation, realizing the accurate identification of the building usage pattern. A wall heat storage model is constructed based on user behavior characteristics and standardized environmental parameters, and the instantaneous heat flow data of the building is obtained through dynamic thermal resistance calculation, accurately reflecting the dynamic heat transfer characteristics of the building wall. A spatial downscaling model is established using the instantaneous heat flow data of the building and meteorological forecast data, and local meteorological parameters are obtained through microclimate correction, solving the problem that regional meteorological data cannot accurately reflect the environmental characteristics around the building. A prediction feature matrix is constructed with local meteorological parameters and the instantaneous heat flow data of the building, and the cooling load prediction values are obtained through time series feature extraction and attention weight calculation, making full use of the advantages of deep learning algorithms in time series data processing and improving the prediction accuracy. The refrigeration power adjustment coefficient of the air conditioning system is calculated according to the cooling load prediction values and the sub-period cooling capacity control instructions are generated, realizing the precise control of the air conditioning system. In a specific air conditioning load prediction field, this solution establishes a data processing and analysis process by introducing algorithms such as hierarchical clustering and deep learning, and combining the physical characteristics and usage characteristics of the building, significantly improving the prediction accuracy.

[0020] In a specific embodiment, the process of executing step S101 may specifically include the following steps: (1) Within a preset 24-hour time interval, each hour is divided into 60 sampling moments, and the indoor and outdoor temperature, humidity, and CO2 concentration data are collected through the environmental sensing network at each sampling moment to generate an environmental original sequence; (2) The original environmental sequence is segmented according to the hourly unit, and the upper and lower limits of the quartile interval are calculated for the temperature, humidity, and CO2 concentration data in each time period. The abnormal data points that exceed the interval range are eliminated to obtain the abnormal value processing sequence; (3) The temperature, humidity, and CO2 concentration data of each time period in the outlier processing sequence are normalized according to the maximum and minimum value differences to obtain a normalized environmental sequence; (4) The temperature, humidity, and CO2 concentration data in the normalized environmental series are grouped according to a five-minute time window, and the cumulative average values ​​of the environmental parameters in each group are calculated to generate a smoothed environmental series; (5) The smoothed environmental sequence is reorganized according to the 24-hour time interval, converted into a time series environmental data matrix, and the standardized environmental parameters are obtained.

[0021] Specifically, within a 24-hour time interval, each hour is subdivided into 60 sampling moments, which is equivalent to collecting data once a minute. The temperature sensor, humidity sensor, and CO2 concentration sensor in the environmental sensor network collect data synchronously at each sampling moment. The original data includes indoor and outdoor temperature values ​​(unit: °C), relative humidity values ​​(unit: %), and CO2 concentration values ​​(unit: ppm). These data are arranged in the order of collection time to form the original environmental sequence.

[0022] The original environmental sequence is processed in segments per hour. For the temperature, humidity, and CO2 concentration data in each hourly time period, the quartile intervals are calculated respectively. The quartile interval sorts the data from small to large to find the Q1 value at the 25% position, the Q2 value (median) at the 50% position, and the Q3 value at the 75% position. The upper limit is defined as Q3+1.5IQR, and the lower limit is defined as Q1-1.5IQR, where IQR is the interquartile range, which is equal to Q3-Q1. The data in each time period is screened, and the data points that exceed the upper and lower limits are marked as outliers and removed to form an outlier processing sequence.

[0023] For the data in each time period of the outlier processing sequence, normalization processing is performed. The normalization processing adopts the maximum-minimum difference method to map data with different dimensions into the interval [0, 1]. The specific approach is to find the maximum and minimum values of temperature, humidity, and CO2 concentration respectively in each time period, and then perform a linear transformation on the original data. The relative magnitude relationship of the original data remains unchanged after the transformation, which is convenient for subsequent processing, and a normalized environment sequence is obtained. The sliding time window method is used to smooth the normalized environment sequence. Set 5 minutes as the time window length, and the window slides backward by 1 minute each time. For the temperature, humidity, and CO2 concentration data within each time window, calculate their arithmetic mean. The smoothing processing reduces the random fluctuations of the data and highlights the change trend of the environmental parameters, generating a smoothed environment sequence.

[0024] Reorganize the smoothed environment sequence according to a 24-hour time interval. Align the temperature, humidity, and CO2 concentration data in time to construct a three-dimensional data matrix, where each dimension represents an environmental parameter, and each time point corresponds to a set of environmental parameter values.

[0025] For example: The environmental sensing network collects data once per minute within 24 hours, generating a total of 1440 sets of original data. Taking the temperature data as an example, among the 60 temperature values collected within a certain hour, there is an abnormally high temperature of 35°C caused by a sensor failure. By calculating the quartiles of the temperature data for that hour, Q1 = 23°C, Q2 = 24°C, Q3 = 25°C, IQR = 2°C, the upper limit value is 28°C, and the lower limit value is 20°C. 35°C is significantly beyond the upper limit value, so it is removed. Perform normalization processing on the remaining temperature data, find the highest temperature of 26°C and the lowest temperature of 22°C, and convert the temperature values to the interval [0, 1] according to linear mapping. Then use a 5-minute sliding window to average the normalized data. Each window contains 5 data points, and after calculating the average value, a smoother temperature change curve is obtained. The same processing process is carried out for the humidity and CO2 concentration data. Reorganize the three processed environmental parameters in chronological order to form a standardized environmental parameter matrix, which serves as the basic data for subsequent cooling load prediction. The original environment sequence retains all the information collected by the sensor. Outlier processing avoids the influence of incorrect data, normalization processing solves the problem of inconsistent dimensions of different parameters, moving average eliminates the interference of short-term fluctuations, and data reorganization facilitates subsequent multi-parameter comprehensive analysis.

[0026] In a specific embodiment, the process of executing step S102 may specifically include the following steps: (1) Within a 24-hour time interval, divide the standardized environmental parameters into 48 time slices in units of 30 minutes to generate an environmental time slice sequence; (2) Statistically analyze the personnel density data according to the time slice sequence, calculate the cumulative value of the personnel density within each time slice, and obtain the personnel density distribution data; (3) Align the environmental time slice sequence with the personnel density distribution data in terms of time, construct an environment-personnel joint data matrix, and obtain the time slice distribution matrix; (4) Calculate the environmental change rate and the personnel density change rate for each time slice in the time slice distribution matrix, and generate a time-varying feature sequence; (5) Group the time-varying feature sequence according to weekdays and weekends, obtain the scene type identifier through hierarchical division, and generate the user behavior characteristics.

[0027] Specifically, for the environmental parameters at time t, its mathematical expression is:

[0028] where is the environmental parameter vector, is the standardized temperature value, is the standardized humidity value, is the standardized CO2 concentration value.

[0029] The calculation of the personnel density distribution data adopts:

[0030] where is the personnel density value of the kth time slice, is the number of personnel at the ith sampling point within the time slice, is the effective area of the building, is the time slice length (30 minutes), and n is the number of sampling points within the time slice.

[0031] The expression of the environment-personnel joint data matrix is:

[0032] where is the matrix element, is the environmental parameter value of the ith time slice at time and is the corresponding personnel density value.

[0033] The calculation formulas for the environmental change rate and the personnel density change rate are:

[0034]

[0035] where is the environmental change rate of the kth time slice, is the rate of change of personnel density.

[0036] Scene type feature vector is defined as:

[0037] where is the weekday identifier of the time slice (1 for weekdays, 0 for rest days).

[0038] For example, within the time slice from 8:00 to 8:30 in the morning, the rate of change of temperature is 0.4 °C / 30 min, and the rate of change of personnel density is 20 people / ( ). By calculating the feature vectors of 48 time slices and combining the data grouping of weekdays and rest days, a description of user behavior characteristics is formed. The features reflect the dynamic change rules of environmental parameters and personnel activities during the use of the building.

[0039] In a specific embodiment, the process of executing step S103 may specifically include the following steps: (1) Perform a time series combination of the indoor-outdoor temperature difference in the standardized environmental parameters and the rate of change of personnel density in the user behavior characteristics to generate a heat flow influence factor; (2) Map the heat flow influence factor layer by layer according to the building structure to generate multi-layer heat transfer nodes, and calculate the heat conduction coefficient between nodes for the time series data of the multi-layer heat transfer nodes to construct a node heat transfer network; (3) Perform an associative operation on the node heat transfer network and the building material parameters to generate a wall heat storage model, and calculate the temperature distribution gradient of each layer of the building wall according to the wall heat storage model and the heat flow influence factor to obtain the wall temperature field data; (4) Perform time step slicing on the wall temperature field data based on the wall heat storage model, and calculate the change in heat capacity of the wall material within each time step to obtain the heat storage characteristic parameters; (5) Perform a time-varying coefficient calculation on the heat storage characteristic parameters in combination with the wall heat storage model, and at the same time perform a thermal resistance base value operation in combination with the wall material thickness to obtain an initial thermal resistance sequence; (6) Dynamically correct the initial thermal resistance sequence with the temperature field change rate to obtain a thermal resistance distribution sequence; (7) Perform a heat flux density conversion on the thermal resistance distribution sequence and the indoor-outdoor temperature difference to obtain the instantaneous heat flux data of the building.

[0040] Specifically, consider the combined influence of the indoor-outdoor temperature difference and personnel activities on the heat flow. Through the heat flow influence factor calculation formula:

[0041] where is the heat flow influence factor at time t, is the temperature difference between indoor and outdoor (°C). is the change rate of personnel density ( ·h), and are the weight coefficients. The temperature factor and the personnel activity factor are weighted and combined to reflect their comprehensive influence on the heat transfer of the wall. Based on the heat flow influence factor, the building wall is analyzed layer by layer. Heat transfer nodes are set for each layer of material, and calculated through the thermal conductivity between nodes:

[0042] where is the thermal conductivity between the i-th and j-th layer nodes, is the thermal conductivity of the i-th layer material, is the heat transfer area, is the distance between nodes (m). It reflects the heat transfer ability between different material layers.

[0043] According to the heat transfer characteristics between nodes, establish the wall temperature field distribution equation:

[0044] where is the temperature field distribution at position x and time t (°C), is the heat flux density, is the material thermal conductivity. The temperature field distribution directly affects the heat storage characteristics of the wall.

[0045] Based on the temperature field distribution, calculate the heat storage characteristics of the wall:

[0046] where is the heat storage quantity at time t (J), is the material density, is the specific heat capacity, is the material volume, is the temperature change. The heat storage characteristics characterize the ability of the wall to store and release heat.

[0047] Consider the thermal resistance effect of each layer of wall material:

[0048] where is the thermal resistance value at time t, with the unit of , is the thickness of the i-th layer material (m), is the thermal conductivity, with the unit of . The thermal resistance value reflects the ability of the wall to hinder heat transfer.

[0049] Dynamically correct the thermal resistance according to the temperature field change:

[0050] Wherein is the corrected thermal resistance value, is the influence coefficient of the temperature change rate. Dynamic correction considers the influence of temperature change on the thermal resistance. Calculate the instantaneous heat flux of the building:

[0051] Wherein is the heat flux density at time t. It reflects the real-time heat transfer condition of the wall.

[0052] For example: The exterior wall of a certain building is composed of three layers of materials (from outside to inside are the insulation layer, block layer, and plaster layer). When the indoor-outdoor temperature difference is 5°C, it is found through the calculation of the heat flux influence factor that the heat flux contribution caused by human activities accounts for about 20%. The temperature gradients at each layer node show that the temperature drop of the insulation layer is the largest, accounting for 60% of the total temperature difference. The dynamic thermal resistance calculation shows that when the indoor temperature changes rapidly, the equivalent thermal resistance value of the wall will increase by 15 - 20%, and this change directly affects the calculation result of the instantaneous heat flux.

[0053] In a specific embodiment, the process of executing step S104 may specifically include the following steps: (1) Divide the instantaneous heat flux data of the building into multiple regional data segments according to the building orientation to obtain the heat flux regional distribution; (2) Perform time series decomposition on the temperature, humidity, and wind speed data in the weather forecast data to obtain the weather change trend; (3) Perform spatio-temporal corresponding mapping on the heat flux regional distribution and the weather change trend to generate the regional influence factor; (4) Construct a heat island intensity calculation matrix according to the regional influence factor to obtain the spatial correlation coefficient; (5) Combine and calculate the spatial correlation coefficient with the building geometric parameters to generate a spatial downscaling model, where the spatial downscaling model is a spatial mapping relationship matrix that converts the regional scale weather forecast data into the micro-meteorological data around the building; (6) Perform scale conversion on the weather forecast data according to the spatial downscaling model to obtain the initial microclimate data; (7) Perform correction calculation on the initial microclimate data based on the building shielding effect to obtain the microclimate correction coefficient; (8) Perform spatial weighted calculation on the microclimate correction coefficient and the initial microclimate data to generate the local meteorological parameters.

[0054] Specifically, the microclimate correction process needs to process the heat flux data of each orientation of the building. The instantaneous heat flux of the building is divided into four regions according to the east, south, west, and north orientations, and the heat flux distribution of each region can be expressed as:

[0055] where is the heat flux value (W / ) at the spatial point . is the orientation weight coefficient, is the heat flux distribution function for each orientation, reflecting the heat flux distribution characteristics of different orientations of the building.

[0056] The meteorological forecast data is processed by time series decomposition and expressed by the following formula:[[]]

[0057] where is the time series of meteorological parameters, is the trend term, is the periodic term, is the random term. The decomposition helps to identify the main characteristics of meteorological changes.

[0058] The spatio-temporal mapping relationship between the regional distribution of heat flux and the trend of meteorological changes is expressed by the regional influence factor:[[]]

[0059] where is the regional influence factor, is the meteorological trend adjustment coefficient, reflecting the coupling relationship between local heat flux and meteorological changes. The construction of the heat island intensity calculation matrix adopts:[[]]

[0060] where is the heat island intensity matrix element, is the influence factor of the k-th region, is the weight coefficient, is the spatial distance, describing the spatial distribution of the urban heat island effect. The mathematical expression of the spatial downscaling model is:[[]]

[0061] where is the local meteorological parameter after downscaling, is the spatial basis function, M is the number of discrete points in the horizontal grid direction, and N is the number of discrete points in the vertical grid direction. These two parameters determine the spatial downscaling accuracy and computational complexity.

[0062] describes the mapping relationship from the regional scale to the local scale. The occlusion effect correction of the initial microclimate data adopts:[[]]

[0063] where is the corrected microclimate data is the occlusion weight, is the occlusion function. The calculation expression of local meteorological parameters:

[0064] is the local meteorological parameter, is the spatial weighting coefficient. Taking an office building as an example: The solar radiation heat flux received by the south outer wall of the building at noon in summer reaches 800 W / , while the north outer wall is only 200 W / . The decomposition of meteorological data shows that the daily temperature variation range in this area is 10°C, with obvious periodic characteristics. Through the analysis of the spatial downscaling model, it is found that the actual temperature around the building is 2 - 3°C lower than the regional meteorological forecast value, which is mainly affected by building shading and greening. This refined microclimate analysis provides more accurate local meteorological parameters for air-conditioning load prediction. The microclimate correction process reflects the scale conversion law from regional meteorology to local meteorology, fully considering the influence of building orientation, heat island effect, and occlusion effect on the local meteorological environment.

[0065] In a specific embodiment, the process of executing step S105 may specifically include the following steps: (1) Perform time synchronization processing on the local meteorological parameters and the building instantaneous heat flux data to obtain time-series combined data; (2) Slide and segment the time-series combined data with a 24-hour period to generate a time window sequence; (3) Extract the temperature change rate, heat flux change rate, and pressure gradient from the time window sequence to construct a prediction feature matrix, where the prediction feature matrix is a set of associated data between the building surrounding environment and the change of cooling load; (4) Calculate the correlation degree between each feature based on the prediction feature matrix to generate a feature importance index; (5) Sort the feature importance index according to the correlation degree to obtain the attention weight coefficient; (6) Perform weighted operation on the prediction feature matrix of each time window and the attention weight coefficient to obtain a weighted feature sequence; (7) Map the weighted feature sequence according to the historical cooling load data to generate a cooling load prediction value.

[0066] Specifically, as Figure 2As shown in the figure, it is a schematic flowchart of obtaining the cold load prediction value through time series feature extraction and attention weight calculation in the embodiment of the present application. The local meteorological parameters and the instantaneous heat flow data of the building are subjected to time synchronization processing. The two types of data are aligned according to the same time stamp to form a time series combined data containing meteorological information and heat flow information. Time synchronization ensures the consistency of data from different sources in the time dimension and provides a basis for subsequent analysis. The time series combined data needs to be sliced by sliding with a basic period of 24 hours. The length of the sliding window is set to 24 hours and it slides backward by 1 hour each time, so that each time window contains a daily cycle data. This slicing method fully considers the periodicity of the building usage characteristics, and each time window sequence contains daily variation information.

[0067] Extracting features from each time window sequence includes three key indicators: temperature change rate, heat flow change rate, and air pressure gradient. The temperature change rate reflects the change speed of the ambient temperature and is calculated by the temperature difference between adjacent moments. The heat flow change rate represents the dynamic characteristics of the building heat transfer state and is obtained from the difference in heat flow density between adjacent moments. The air pressure gradient describes the spatial distribution characteristics of the local air pressure field and is calculated by the spatial difference of air pressure values at different positions. These features form a prediction feature matrix, which is a set of associated data between the building's surrounding environment and the cold load change. When calculating the correlation degree between features based on the prediction feature matrix, the Pearson correlation coefficient method is used. Calculate the correlation coefficient between each pair of features to form a correlation matrix. The larger the absolute value of the correlation coefficient, the more significant the influence of the feature on the cold load prediction, and thus a feature importance index is generated.

[0068] The feature importance index is sorted in descending order according to the correlation degree to obtain the attention weight coefficient reflecting the contribution degree of each feature. The feature with a larger weight coefficient receives more attention during the prediction process. This data-driven weight allocation method can adaptively adjust the importance of different features. Perform a weighted operation on the prediction feature matrix and the attention weight coefficient to obtain a weighted feature sequence that integrates the feature importance. The weighting process highlights the influence of key features, suppresses the interference of secondary features, and improves the effectiveness of feature expression.

[0069] Map the weighted feature sequence to the historical cold load data to establish a mapping relationship from the feature space to the load space, thereby generating the cold load prediction value. This mapping process considers the load change law in the historical data and makes the prediction result more in line with the actual situation.

[0070] Taking the air-conditioning load prediction of an office building as an example: The locally collected meteorological data (including temperature, humidity, air pressure, etc.) and the instantaneous heat flux data of the building are time-aligned to form a time-series combined data with one data point per minute. Then, it is segmented by a sliding window with a 24-hour cycle, and each window contains 1440 data points. Features are extracted from these window data, such as the hourly change rate of temperature, the heat flux change of each building orientation, and the spatial distribution characteristics of the surrounding air pressure field. Through correlation analysis, it is found that the correlation between the temperature change rate and the cooling load change is the highest, followed by the heat flux change rate, while the influence of the air pressure gradient is relatively small. Accordingly, the feature weights are determined, and the temperature change rate obtains a higher attention weight. The weighted feature sequence is mapped with the historical cooling load data to obtain the predicted cooling load values for the next 24 hours. The prediction process makes full use of the information of multi-source data and realizes the accurate prediction of the building's cooling load through feature extraction, weight calculation, and mapping analysis.

[0071] In a specific embodiment, the process of executing step S106 may specifically include the following steps: (1) Divide the predicted cooling load values into multiple time periods according to the 24-hour time interval to obtain a time-sharing cooling load sequence; (2) Discretize the cooling load data in the time-sharing cooling load sequence to generate cooling load level marks; (3) Calculate the percentage of the rated cooling capacity of the air-conditioning system according to the cooling load level marks to obtain the refrigeration power reference value; (4) Perform dynamic compensation calculation on the refrigeration power reference value to generate a refrigeration power adjustment coefficient; (5) Match the refrigeration power adjustment coefficient with the operating state of the air-conditioning system to obtain a refrigeration capacity adjustment parameter; (6) Control the air-conditioning system in segments according to the refrigeration capacity adjustment parameter to generate a time-sharing refrigeration capacity control instruction.

[0072] Specifically, the 24-hour predicted cooling load values are divided into multiple time periods. According to the usage characteristics of the building, it is divided with 1 hour as the basic unit to form 24 consecutive time periods. Each time period contains the predicted cooling load values within that hour, and these values are arranged in chronological order to form a time-sharing cooling load sequence. The time-sharing cooling load sequence needs to be discretized. The discretization process maps the continuous cooling load values into a preset level interval. For example, the cooling load values are divided into three levels: low load, medium load, and high load. The specific division method is based on the statistical analysis of historical operation data to determine the threshold range of different levels. Through this discretization process, the cooling load level mark corresponding to each time period is generated.

[0073] When converting the cold load level mark into the cooling capacity control parameter of the air conditioning system, it is necessary to calculate the percentage relative to the rated cooling capacity. Obtain the rated cooling capacity value of the air conditioning system, and then determine the corresponding cooling capacity percentage range according to different load levels. As Figure 3 shown, it is a schematic diagram of the load level in the embodiment of the present application. The load level of the air conditioning system is divided into three levels: high, medium, and low. Each level corresponds to different operating parameter configurations. At the high load level, the cooling capacity is 80%-100% of the rated value, the compressor operates at 50-60 Hz, and the fan runs at high speed; at the medium load level, the cooling capacity is 50%-80% of the rated value, the compressor operates at 35-50 Hz, and the fan maintains medium speed; at the low load level, the cooling capacity is controlled at 30%-50% of the rated value, the compressor operates at 20-35 Hz, and the fan runs at low speed.

[0074] The low load level corresponds to 30%-50% of the rated cooling capacity, the medium load level corresponds to 50%-80% of the rated cooling capacity, and the high load level corresponds to 80%-100% of the rated cooling capacity. This percentage division method obtains the initial cooling power reference value. The cooling power reference value also needs to be dynamically compensated considering the influence of the actual operating environment. The compensation calculation considers factors such as outdoor temperature and the operating duration of the air conditioning system to correct the reference value. When the outdoor temperature is high, the compensation coefficient is increased; when the equipment runs continuously for a long time, the compensation coefficient is appropriately reduced. Through this dynamic compensation calculation, a more accurate cooling power adjustment coefficient is generated.

[0075] The cooling power adjustment coefficient needs to be matched with the actual operating state of the air conditioning system. The operating state includes parameters such as the compressor operating frequency and the fan speed. There is a corresponding relationship between these parameters and the cooling power. Through table lookup or interpolation calculation, the cooling power adjustment coefficient is converted into specific equipment operating parameters to obtain the cooling capacity adjustment parameter. According to the cooling capacity adjustment parameter, a time-sharing cooling capacity control instruction is generated. The control instruction contains specific control quantities such as the compressor frequency setting value and the fan speed setting value for each time period. These instructions are executed in chronological order to achieve the time-sharing operation control of the air conditioning system.

[0076] Taking the air-conditioning control of an office building as an example: The prediction shows that the cooling load gradually increases from 9 am to 11 am on weekdays. This period is divided into one-hour intervals, and the cooling load data for the two intervals of 9:00 - 10:00 and 10:00 - 11:00 are obtained. Through discretization, it is found that the period from 9:00 - 10:00 belongs to the medium load level, corresponding to 65% of the rated cooling capacity; the period from 10:00 - 11:00 belongs to the high load level, corresponding to 85% of the rated cooling capacity. Considering that the outdoor temperature is expected to rise from 28°C to 30°C during these two periods, the reference value of the cooling power is compensated, and the adjustment coefficients are increased by 5% and 8% respectively. The generated control instructions require that the compressor operating frequency be set to 45 Hz and the fan speed be set to medium speed during the period from 9:00 - 10:00; the compressor frequency is increased to 52 Hz and the fan speed is adjusted to high speed during the period from 10:00 - 11:00. This time-segmented control method based on prediction fully considers the building load change characteristics and the equipment operating status, and realizes the efficient operation of the air-conditioning system.

[0077] The air-conditioning cooling load prediction method in the embodiment of the present application is described above. Next, the air-conditioning cooling load prediction system in the embodiment of the present application will be described. Please refer to Figure 4 , an embodiment of the air-conditioning cooling load prediction system in the embodiment of the present application includes: An acquisition module, configured to collect the original data of indoor and outdoor temperature, humidity, and CO2 concentration through an environmental sensing network, perform outlier rejection and sliding window smoothing processing on the original data, and obtain standardized environmental parameters; A calculation module, configured to establish a time slice distribution matrix according to the standardized environmental parameters and the personnel density data, and obtain user behavior characteristics through hierarchical clustering calculation; A construction module, configured to construct a wall heat storage model according to the user behavior characteristics and the standardized environmental parameters, and obtain the instantaneous heat flow data of the building through dynamic thermal resistance calculation; A correction module, configured to establish a spatial downscaling model according to the instantaneous heat flow data of the building and the meteorological forecast data, and obtain local meteorological parameters through microclimate correction; A prediction module, configured to construct a prediction feature matrix according to the local meteorological parameters and the instantaneous heat flow data of the building, and obtain the cooling load prediction value through time series feature extraction and attention weight calculation; A generation module, configured to calculate the cooling power adjustment coefficient of the air-conditioning system according to the cooling load prediction value, and generate a time-segmented cooling capacity control instruction.

[0078] Through the collaborative cooperation of the above-mentioned various components, the original data of indoor and outdoor temperature, humidity, and CO2 concentration are collected through the environmental sensing network, and outlier rejection and sliding window smoothing processing are performed, improving the quality and reliability of the environmental data and providing an accurate data basis for subsequent prediction. A time slice distribution matrix is established based on the standardized environmental parameters and personnel density data, and the user behavior characteristics are obtained through hierarchical clustering calculation, realizing the accurate identification of the building usage pattern. A wall heat storage model is constructed based on the user behavior characteristics and standardized environmental parameters, and the instantaneous heat flow data of the building are obtained through dynamic thermal resistance calculation, accurately reflecting the dynamic heat transfer characteristics of the building wall. A spatial downscaling model is established using the instantaneous heat flow data of the building and the meteorological forecast data, and the local meteorological parameters are obtained through microclimate correction, solving the problem that the regional meteorological data cannot accurately reflect the characteristics of the building's surrounding environment. A prediction feature matrix is constructed with the local meteorological parameters and the instantaneous heat flow data of the building, and the cooling load prediction value is obtained through time series feature extraction and attention weight calculation, making full use of the advantages of deep learning algorithms in time series data processing and improving the prediction accuracy. The cooling power adjustment coefficient of the air conditioning system is calculated based on the cooling load prediction value and the time-slice cooling capacity control instructions are generated, realizing the precise control of the air conditioning system. In a specific air conditioning load prediction field, this solution establishes a data processing and analysis process by introducing algorithms such as hierarchical clustering and deep learning, and combining the physical characteristics and usage characteristics of the building, significantly improving the prediction accuracy.

[0079] Referring to Figure 5 , an embodiment of the present invention also provides a computer device, which may be a server, and its internal structure may be as Figure 5 shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface, and a database connected through a system bus. Among them, the processor of the computer design is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0080] Those skilled in the art can understand that Figure 5 the structure shown in

[0081] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0082] Those of ordinary skill in the art can understand that all or part of the processes in the above-described method embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the above-described method embodiments. Among them, any reference to memory, storage, database, or other media provided by the present invention and used in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM, etc.

[0083] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described system, system, and unit can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0084] When 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 such an understanding, the technical solution of the present invention, in essence, 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 and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0085] As described above, the above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A method for predicting the cooling load of an air conditioner, characterized in that, The air-conditioning cooling load prediction method includes: Collecting the original data of indoor and outdoor temperature, humidity, and CO2 concentration through the environmental sensing network, removing outliers and performing sliding window smoothing on the original data to obtain standardized environmental parameters; Establishing a time slice distribution matrix based on the standardized environmental parameters and personnel density data, and calculating user behavior characteristics through hierarchical clustering; Constructing a wall heat storage model based on user behavior characteristics and standardized environmental parameters, and calculating the instantaneous heat flow data of the building through dynamic thermal resistance; Establishing a spatial downscaling model based on the instantaneous heat flow data of the building and meteorological forecast data, and obtaining local meteorological parameters through microclimate correction; Constructing a prediction feature matrix based on the local meteorological parameters and the instantaneous heat flow data of the building, and obtaining the cooling load prediction value through time series feature extraction and attention weight calculation; Calculating the refrigeration power adjustment coefficient of the air-conditioning system according to the cooling load prediction value, and generating a time-segmented cooling capacity control instruction.

2. The air-conditioning cooling load prediction method according to claim 1, characterized in that The step of collecting the original data of indoor and outdoor temperature, humidity, and CO2 concentration through the environmental sensing network, removing outliers and performing sliding window smoothing on the original data to obtain standardized environmental parameters includes: Within a preset 24-hour time interval, each hour is divided into 60 sampling moments. The environmental sensing network collects the indoor and outdoor temperature, humidity, and CO2 concentration data at each sampling moment to generate an environmental original sequence; Segmenting the environmental original sequence on an hourly basis, calculating the upper and lower quartile interval values of the temperature, humidity, and CO2 concentration data within each time period, and removing the abnormal data points outside the interval range to obtain an outlier-processed sequence; Performing normalization calculation on the temperature, humidity, and CO2 concentration data within each time period of the outlier-processed sequence according to the maximum-minimum value difference to obtain a normalized environmental sequence; Grouping the temperature, humidity, and CO2 concentration data in the normalized environmental sequence according to a five-minute time window, calculating the cumulative average value of the environmental parameters within each group, and generating a smoothed environmental sequence; Reorganizing the data of the smoothed environmental sequence according to a 24-hour time interval, converting it into a time series environmental data matrix, and obtaining standardized environmental parameters.

3. The air-conditioning cooling load prediction method according to claim 1, wherein The step of establishing a time slice distribution matrix based on the standardized environmental parameters and personnel density data, and calculating user behavior characteristics through hierarchical clustering includes: Within a 24-hour time interval, dividing the standardized environmental parameters into 48 time slices at 30-minute intervals to generate an environmental time slice sequence; Statistically analyzing the personnel density data according to the time slice sequence, calculating the cumulative value of the personnel density within each time slice, and obtaining the personnel density distribution data; Aligning the environmental time slice sequence with the personnel density distribution data in time, constructing an environment-personnel joint data matrix, and obtaining the time slice distribution matrix; Calculating the environmental change rate and personnel density change rate for each time slice in the time slice distribution matrix to generate a time-varying feature sequence; Grouping the time-varying feature sequence according to weekdays and weekends, and obtaining the scene type identifier through hierarchical division to generate user behavior characteristics.

4. The air-conditioning cooling load prediction method according to claim 1, wherein, Constructing a wall heat storage model based on user behavior characteristics and standardized environmental parameters, and obtaining building instantaneous heat flow data through dynamic thermal resistance calculation, including: Performing a time series combination on the indoor-outdoor temperature difference in the standardized environmental parameters and the change rate of personnel density in the user behavior characteristics to generate a heat flow influence factor; Mapping the heat flow influence factor layer by layer according to the building structure to generate multi-layer heat transfer nodes, calculating the heat conduction coefficient between nodes for the time series data of the multi-layer heat transfer nodes, and constructing a node heat transfer network; Performing an associative operation on the node heat transfer network and the building material parameters to generate a wall heat storage model, and calculating the temperature distribution gradient of each layer of the building wall according to the wall heat storage model and the heat flow influence factor to obtain wall temperature field data; Performing time step segmentation on the wall temperature field data based on the wall heat storage model, calculating the change in heat capacity of the wall material within each time step, and obtaining heat storage characteristic parameters; Performing a time-varying coefficient calculation on the heat storage characteristic parameters in combination with the wall heat storage model, and performing a thermal resistance base value operation in combination with the wall material thickness to obtain an initial thermal resistance sequence; Performing dynamic correction on the initial thermal resistance sequence and the temperature field change rate to obtain a thermal resistance distribution sequence; Performing a heat flow density conversion on the thermal resistance distribution sequence and the indoor-outdoor temperature difference to obtain building instantaneous heat flow data.

5. The air-conditioning cooling load prediction method according to claim 1, wherein Establishing a spatial downscaling model based on the building instantaneous heat flow data and meteorological forecast data, and obtaining local meteorological parameters through microclimate correction, including: Dividing the building instantaneous heat flow data into multiple regional data segments according to the building orientation to obtain a heat flow regional distribution; Performing time series decomposition on the temperature, humidity, and wind speed data in the meteorological forecast data to obtain a meteorological change trend; Performing a spatio-temporal correspondence mapping on the heat flow regional distribution and the meteorological change trend to generate a regional influence factor; Constructing a heat island intensity calculation matrix according to the regional influence factor to obtain a spatial correlation coefficient; Performing a combined operation on the spatial correlation coefficient and the building geometric parameters to generate a spatial downscaling model, where the spatial downscaling model is a spatial mapping relationship matrix that converts regional-scale meteorological forecast data into micro-meteorological data around the building; Performing scale conversion on the meteorological forecast data according to the spatial downscaling model to obtain initial microclimate data; Performing a correction calculation on the initial microclimate data based on the building occlusion effect to obtain a microclimate correction coefficient; Performing a spatial weighted calculation on the microclimate correction coefficient and the initial microclimate data to generate local meteorological parameters.

6. The air-conditioning cooling load prediction method according to claim 1, wherein Constructing a prediction feature matrix based on the local meteorological parameters and the building instantaneous heat flow data, and obtaining a cooling load prediction value through time series feature extraction and attention weight calculation, including: Performing time synchronization processing on the local meteorological parameters and the building instantaneous heat flow data to obtain time series combined data; Performing sliding segmentation on the time series combined data with a 24-hour period to generate a time window sequence; Extracting the temperature change rate, heat flow change rate, and pressure gradient from the time window sequence to construct a prediction feature matrix, where the prediction feature matrix is a set of associated data between the building surrounding environment and the cooling load change; Calculating the correlation degree between each feature based on the prediction feature matrix to generate a feature importance index; Sort the feature importance indicators according to the degree of relevance to obtain the attention weight coefficients; Perform weighted operations on the prediction feature matrices of each time window and the attention weight coefficients to obtain weighted feature sequences; Map the weighted feature sequences according to the historical cooling load data to generate cooling load prediction values.

7. The air-conditioning cooling load prediction method according to claim 1, wherein, The calculating the cooling power adjustment coefficient of the air conditioning system according to the cooling load prediction value and generating a sub-period cooling capacity control instruction includes: Divide the cooling load prediction value into multiple time periods according to a 24-hour time interval to obtain a sub-period cooling load sequence; Discretize the cooling load data in the sub-period cooling load sequence to generate cooling load level marks; Calculate the percentage of the rated cooling capacity of the air conditioning system according to the cooling load level marks to obtain the cooling power reference value; Perform dynamic compensation calculation on the cooling power reference value to generate a cooling power adjustment coefficient; Match the cooling power adjustment coefficient with the operating state of the air conditioning system to obtain the cooling capacity adjustment parameter; Control the air conditioning system in sub-periods according to the cooling capacity adjustment parameter to generate a sub-period cooling capacity control instruction.

8. An air-conditioning cooling load prediction system for implementing the air-conditioning cooling load prediction method according to any one of claims 1-7, characterized in that, The air conditioning cooling load prediction system includes: An acquisition module for collecting original data of indoor and outdoor temperature, humidity, and CO2 concentration through an environmental sensor network, removing outliers from the original data, and performing sliding window smoothing processing to obtain standardized environmental parameters; A calculation module for establishing a time slice distribution matrix according to the standardized environmental parameters and the personnel density data, and obtaining user behavior characteristics through hierarchical clustering calculation; A construction module for constructing a wall heat storage model according to the user behavior characteristics and the standardized environmental parameters, and obtaining the instantaneous heat flow data of the building through dynamic thermal resistance calculation; A correction module for establishing a spatial downscaling model according to the instantaneous heat flow data of the building and the meteorological forecast data, and obtaining local meteorological parameters through microclimate correction; A prediction module for constructing a prediction feature matrix according to the local meteorological parameters and the instantaneous heat flow data of the building, and obtaining a cooling load prediction value through time series feature extraction and attention weight calculation; A generation module for calculating the cooling power adjustment coefficient of the air conditioning system according to the cooling load prediction value and generating a sub-period cooling capacity control instruction.

9. A computer device, characterized in that, It includes a memory and a processor, and the memory stores a computer program that can be run on the processor. The characteristic is that when the processor executes the computer program, it implements the air conditioning cooling load prediction method described in any one of claims 1 to 7.

10. A computer-readable storage medium, on which a computer program is stored, and when the computer program is run by a processor, the processor is caused to execute the air conditioning cooling load prediction method described in any one of claims 1 to 7.

Images