Air conditioner datum line selection method, system, equipment and medium

Abstract

The invention discloses an air conditioner datum line selection method, system and device and a medium. The method comprises the steps that weather parameters of the day and historical air conditioner load data are obtained; on the basis of the weather parameters of the day and the historical air conditioner load data, a pre-constructed optimal datum line selection model is combined for calculation, and the optimal datum line of the air conditioner of the day is obtained; wherein the optimal datum line selection model is obtained by correcting and constructing a candidate optimal datum line obtained by performing temperature difference weighting and Frechet distance calculation on historical air conditioner load data by using a dynamic correction function; according to the invention, the temperature difference weighted calculation can quantify the influence of temperature deviation on the load; the Frechet distance calculation can more accurately capture the load curve form similarity; the dynamic correction function is combined with the reference line adjusted by the real-time temperature, so that the accuracy is

Description

Technical Field

[0001] The present application relates to the field of power system data analysis, and specifically to an air conditioning baseline selection method, system, equipment and medium. Background Art

[0002] With the development of society and economic growth, power system load management has become increasingly complex, especially for air conditioning loads. As the primary energy consumer in modern buildings, the operating efficiency and energy consumption of air conditioning systems directly impact the overall load distribution of the power system. Therefore, accurate analysis and effective management of air conditioning loads have become crucial.

[0003] Air conditioning load data preprocessing involves a large amount of time-series data, including load values ​​at different time points and corresponding environmental parameters such as temperature and humidity. To better regulate electricity demand, a stable and efficient algorithm for mining air conditioning load patterns is needed. This is particularly important as air conditioning loads increasingly contribute to electricity consumption across various industries.

[0004] The load baseline is a key metric for evaluating the effectiveness of demand response programs, representing normal electricity usage behavior in the absence of demand response interventions. Traditional baseline determination methods include the high-X-day average method, the regression baseline method, and the matching day method. The high-X-day average method averages the X days with the highest load from a preselected Y days; the regression baseline method considers the impact of factors such as temperature on load through a regression model; and the matching day method constructs a baseline based on historical data from similar days during the same period. These methods each have their advantages and disadvantages in practical applications. For example, the high-X-day average method is simple to use but has limited accuracy, while the regression model can reflect environmental factors but is computationally complex.

[0005] In recent years, with the development of machine learning technology, several novel baseline determination methods have been proposed, such as baseline selection methods based on cluster analysis, prediction models based on neural networks, and ensemble learning methods that consider multidimensional features. These methods improve the accuracy and representativeness of baselines by mining the deep features of historical load data. However, these methods still face challenges in dealing with the highly time-varying and nonlinear characteristics of air conditioning loads.

[0006] Currently, common air conditioning load analysis methods are primarily based on statistical analysis and simple linear regression models. However, these methods struggle to fully account for the complex interactions between temperature variations, humidity, and time series characteristics. Furthermore, traditional baseline selection methods often focus solely on average load values, ignoring the morphological characteristics of the load curve. This results in the selected baseline not accurately reflecting typical air conditioning usage patterns.

[0007] Therefore, how to solve the problem that the air-conditioning baseline selected by traditional methods cannot truly reflect the typical mode of air-conditioning use has become an urgent problem to be solved in this field. Summary of the Invention

[0008] In order to solve the problem that the air conditioning baseline selected by the conventional method in the prior art cannot truly reflect the typical usage pattern of the air conditioner, this application proposes an air conditioning baseline selection method, including:

[0009] Obtain the weather parameters of the day and historical air conditioning 96-point load data (historical air conditioning load data);

[0010] Calculating the optimal air conditioning baseline for the day based on the weather parameters of the day and historical air conditioning load data in combination with a pre-built optimal baseline selection model;

[0011] The optimal baseline selection model is constructed by correcting the candidate optimal baselines obtained by performing temperature difference weighting and Fréchet distance calculation based on historical air-conditioning load data using a dynamic correction function.

[0012] Preferably, the process of constructing the optimal baseline selection model includes:

[0013] The historical air conditioning load data is filtered and outliers are processed according to the set outdoor temperature and season, and a load data set is constructed based on the processed data;

[0014] Performing a weighted temperature difference calculation based on the load data set to obtain weighted load curves;

[0015] Calculating the Fleche distance based on the historical air-conditioning load data and the weighted load curves, and selecting the curve with the smallest Fleche distance as a candidate optimal baseline;

[0016] The candidate optimal baseline is modified based on the weather parameters of the day to obtain the optimal baseline of the day.

[0017] Preferably, performing a weighted temperature difference calculation based on the load data set to obtain each weighted load curve includes:

[0018] Based on the difference calculation between the outdoor temperature corresponding to each load data in the load data set and the reference temperature, each temperature difference is obtained;

[0019] The average value is calculated based on the load data with the outdoor temperature as the reference temperature in the load data set to obtain the average load at the reference temperature;

[0020] Performing a difference calculation based on each load data in the load data set and the average load at the reference temperature to obtain each load difference;

[0021] Each load difference is weighted based on the number of temperature difference types and the adjusted weight coefficient to obtain each weighted load difference;

[0022] Based on each weighted load difference and the average load at the reference temperature, the weighted load curves are calculated and drawn.

[0023] Preferably, the calculation process of the adjusted weight coefficient includes:

[0024] The weight coefficient is obtained by calculating based on the absolute value of the temperature difference and the weight attenuation coefficient;

[0025] The weather similarity index is calculated based on the temperature, humidity and wind speed of the current day and the temperature, humidity and wind speed of the reference day;

[0026] The weight coefficient is adjusted and calculated based on the weather similarity index to obtain an adjusted weight coefficient.

[0027] Preferably, the Fréchet distance calculation based on the historical air-conditioning load data and the weighted load curves, and selecting the curve with the smallest Fréchet distance as the candidate optimal baseline, includes:

[0028] Filtering summer air-conditioning load data according to season based on the historical air-conditioning load data, and plotting summer load curves based on the summer air-conditioning load data;

[0029] performing a weighted Fréchet distance calculation based on each of the summer load curves and each of the weighted load curves in combination with a preset time weight, to obtain a Fréchet distance from each of the summer load curves to each of the weighted load curves;

[0030] The average value of each Fréchet distance between each load curve in summer and each weighted load curve is calculated according to the date to obtain the average value of each Fréchet distance;

[0031] The curve with the smallest average value based on the Fréchet distance is selected as the candidate optimal baseline.

[0032] Preferably, the calculation formula of the weighted Fréchet distance is as follows:

[0033]

[0034] Where, F w (A, B) is the weighted Fréchet distance between curve A and curve B; w k is the weight coefficient at time point k; a k b is the load data corresponding to curve A at time point K; k is the load data corresponding to curve B at time point k; d is the distance function used to calculate a k with bk k is the time point index, representing different moments; mn is the total number of time points, that is, the number of moments covered by the load curve.

[0035] Preferably, the step of correcting the candidate optimal baseline based on the weather parameters of the day to obtain the optimal baseline of the day includes:

[0036] Calculate based on the highest and lowest temperatures of the day to get the average temperature of the day;

[0037] Based on the average temperature, average relative humidity and reference relative humidity of the day, the candidate optimal baseline is corrected using the temperature-load correction function to obtain the optimal baseline of the day.

[0038] Preferably, the calculation formula of the temperature-load correction function is as follows:

[0039] L current =L min ×(1+k1×(T avg -26℃)+k2×(RH avg -RH ref ) 2 +k3×(T avg -26℃)×(RH avg -RH ref ))

[0040] Where, L current is the optimal baseline at the current temperature; L min is the candidate optimal baseline; k1 is the temperature linear correction coefficient; k2 is the humidity quadratic correction coefficient; k3 is the temperature and humidity interaction correction coefficient; T avg is the average temperature of the day; RH avg The average relative humidity of the day; RH ref is the reference relative humidity.

[0041] Preferably, the calculation based on the weather parameters of the day and the historical air conditioning load data in combination with a pre-built optimal baseline selection model to obtain the optimal air conditioning baseline for the day includes:

[0042] The load data set is obtained after filtering and processing outliers based on historical air conditioning load data;

[0043] Based on the load data set, the temperature difference weighting and Fréchet distance calculation are performed to obtain the candidate optimal baseline;

[0044] The candidate optimal baseline is modified based on the weather parameters of the day to obtain the optimal baseline for air conditioning on that day.

[0045] Based on the same application concept, this application also proposes an air conditioning baseline selection system, including:

[0046] Data acquisition module, used to obtain the weather parameters of the day and historical air conditioning load data;

[0047] A model solving module, configured to calculate the optimal air conditioning baseline for the day based on the weather parameters of the day and historical air conditioning load data in combination with a pre-built optimal baseline selection model;

[0048] The optimal baseline selection model is constructed by correcting the candidate optimal baselines obtained by performing temperature difference weighting and Fréchet distance calculation based on historical air-conditioning load data using a dynamic correction function.

[0049] Preferably, it further comprises a model construction module, wherein the model construction module:

[0050] The dataset construction submodule is used to filter and process the historical air conditioning load data according to the set outdoor temperature and season, and construct a load dataset based on the processed data;

[0051] A temperature difference weighted calculation submodule is used to perform temperature difference weighted calculation based on the load data set to obtain each weighted load curve;

[0052] a Fréchet distance calculation submodule, configured to calculate the Fréchet distance based on the historical air-conditioning load data and the weighted load curves, and select the curve with the smallest Fréchet distance as a candidate optimal baseline;

[0053] The baseline correction submodule is used to correct the candidate optimal baseline based on the weather parameters of the day to obtain the optimal baseline of the day.

[0054] Preferably, the temperature difference weighted calculation submodule is specifically used to:

[0055] Based on the difference calculation between the outdoor temperature corresponding to each load data in the load data set and the reference temperature, each temperature difference is obtained;

[0056] The average value is calculated based on the load data with the outdoor temperature as the reference temperature in the load data set to obtain the average load at the reference temperature;

[0057] Performing a difference calculation based on each load data in the load data set and the average load at the reference temperature to obtain each load difference;

[0058] Each load difference is weighted based on the number of temperature difference types and the adjusted weight coefficient to obtain each weighted load difference;

[0059] Based on each weighted load difference and the average load at the reference temperature, the weighted load curves are calculated and drawn.

[0060] Preferably, the calculation process of the re-adjusted weight coefficient of the temperature difference weighted calculation submodule includes:

[0061] The weight coefficient is obtained by calculating based on the absolute value of the temperature difference and the weight attenuation coefficient;

[0062] The weather similarity index is calculated based on the temperature, humidity and wind speed of the current day and the temperature, humidity and wind speed of the reference day;

[0063] The weight coefficient is adjusted and calculated based on the weather similarity index to obtain an adjusted weight coefficient.

[0064] Preferably, the Fréchet distance calculation submodule is specifically used to:

[0065] Filtering summer air-conditioning load data according to season based on the historical air-conditioning load data, and plotting summer load curves based on the summer air-conditioning load data;

[0066] performing a weighted Fréchet distance calculation based on each of the summer load curves and each of the weighted load curves in combination with a preset time weight, to obtain a Fréchet distance from each of the summer load curves to each of the weighted load curves;

[0067] The average value of each Fréchet distance between each load curve in summer and each weighted load curve is calculated according to the date to obtain the average value of each Fréchet distance;

[0068] The curve with the smallest average value based on the Fréchet distance is selected as the candidate optimal baseline.

[0069] Preferably, the calculation formula of the weighted Fréchet distance of the Fréchet distance calculation submodule is as follows:

[0070]

[0071] Where, F w (A, B) is the weighted Fréchet distance between curve A and curve B; w k is the weight coefficient at time point k; a k b is the load data corresponding to curve A at time point K; k is the load data corresponding to curve B at time point k; d is the distance function used to calculate a k with b k k is the time point index, representing different moments; mn is the total number of time points, that is, the number of moments covered by the load curve.

[0072] Preferably, the baseline correction submodule is specifically used to:

[0073] Calculate based on the highest and lowest temperatures of the day to get the average temperature of the day;

[0074] Based on the average temperature, average relative humidity and reference relative humidity of the day, the candidate optimal baseline is corrected using the temperature-load correction function to obtain the optimal baseline of the day.

[0075] Preferably, the calculation formula of the temperature-load correction function in the baseline correction submodule is as follows:

[0076] L current =L min ×(1+k1×(T avg -26℃)+k2×(RH avg -RH ref ) 2 +k3×(T avg -26℃)×(RH avg -RH ref ))

[0077] Where, L current is the optimal baseline at the current temperature; L min is the candidate optimal baseline; k1 is the temperature linear correction coefficient; k2 is the humidity quadratic correction coefficient; k3 is the temperature and humidity interaction correction coefficient; T avg is the average temperature of the day; RH avg The average relative humidity of the day; RH ref is the reference relative humidity.

[0078] Preferably, the model solving module is specifically used to:

[0079] The load data set is obtained after filtering and processing outliers based on historical air conditioning load data;

[0080] Based on the load data set, the temperature difference weighting and Fréchet distance calculation are performed to obtain the candidate optimal baseline;

[0081] The candidate optimal baseline is modified based on the weather parameters of the day to obtain the optimal baseline for air conditioning on that day.

[0082] On the other hand, the present application also proposes an electronic device, comprising: at least one processor and a memory; the memory and the processor are connected via a bus;

[0083] The memory is used to store one or more programs;

[0084] When the one or more programs are executed by the at least one processor, the air-conditioning baseline selection method described above is implemented.

[0085] On the other hand, the present application also proposes a readable storage medium having an execution program stored thereon. When the execution program is executed, an air-conditioning baseline selection method as described above is implemented.

[0086] Compared with the prior art, the present invention has the following advantages:

[0087] A method, system, device and medium for selecting an air-conditioning baseline, comprising: obtaining weather parameters for the day and historical air-conditioning load data; performing calculations based on the weather parameters for the day and the historical air-conditioning load data in combination with a pre-constructed optimal baseline selection model to obtain an optimal air-conditioning baseline for the day; wherein the optimal baseline selection model is constructed by correcting a candidate optimal baseline obtained by performing temperature difference weighting and Fréchet distance calculation based on historical air-conditioning load data using a dynamic correction function; the temperature difference weighting calculation of the present application can quantify the impact of temperature offset on load; the Fréchet distance calculation can more accurately capture the similarity of load curve morphology; the dynamic correction function combined with the real-time temperature adjustment of the baseline is more accurate, allowing the air-conditioning baseline to adapt to real-time changing environmental and operating conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] Figure 1 This is a flow chart of an air conditioning baseline selection method for this application;

[0089] Figure 2 This is a flow chart of the air conditioning optimal baseline selection model based on temperature difference weighting and Fréchet distance in this application;

[0090] Figure 3 This is a flow chart of the Fréchet distance calculation algorithm of this application;

[0091] Figure 4 Schematic diagram of the baseline correction and Master Wen correction functions for this application;

[0092] Figure 5 This is a structural diagram of an air conditioning baseline selection system for this application;

[0093] Figure 6 This is a diagram of an electronic device for this application. DETAILED DESCRIPTION

[0094] The present invention proposes an air-conditioning baseline selection method, which aims to solve the baseline selection problem in air-conditioning load analysis. By introducing a temperature difference weighting mechanism and a Fréchet distance calculation method, the method can accurately capture the morphological characteristics of the air-conditioning load curve and comprehensively consider the influence of environmental factors such as temperature and humidity, providing a scientific basis for selecting the optimal baseline; this is of great significance for accurately evaluating the operating efficiency of the air-conditioning system, optimizing energy consumption, and achieving energy conservation and emission reduction goals; in order to better understand the present application, the content of the present application is further explained below in conjunction with the drawings and embodiments of the specification.

[0095] Example 1:

[0096] A method for selecting an air conditioning baseline, the specific process is as follows Figure 1 Shown, including:

[0097] Step 1: Obtain the weather parameters of the day and historical air conditioning load data;

[0098] Step 2: Calculating the optimal baseline for air conditioning on that day based on the weather parameters of the day and historical air conditioning load data in combination with a pre-built optimal baseline selection model;

[0099] The optimal baseline selection model is constructed by correcting the candidate optimal baselines obtained by performing temperature difference weighting and Fréchet distance calculation based on historical air-conditioning load data using a dynamic correction function.

[0100] The following combination Figure 2 To further illustrate this embodiment, before step 1, the process of constructing an optimal baseline selection model is also included. The process of constructing the optimal baseline selection model includes:

[0101] The historical air conditioning load data is filtered and outliers are processed according to the set outdoor temperature and season, and a load dataset is constructed based on the processed data:

[0102] The historical air conditioning load data of this embodiment is 96-point air conditioning load data within a year.

[0103] From the 96 air conditioning load data points in a year, the air conditioning load data when the air conditioning is turned on in summer (June to September) and the outdoor temperature is 26℃±0.5℃ are selected to form the load data set D=L1,L2,...,L n , where L1 is the load data filtered out in the first item, L2 is the load data filtered out in the second item, L n The nth load data is selected, and each load data contains the load values ​​of 96 time points;

[0104] Outlier detection and processing are achieved through the following steps:

[0105] 1) Calculate the interquartile range IQR = Q3 - Q1 of the 96-point load data, where Q1 is the first quartile and Q3 is the third quartile;

[0106] 2) Define the upper and lower bounds, Lower = Q1 - 1.5 × IQR, Upper = Q3 + 1.5 × IQR;

[0107] 3) For any load point L, if l < Lower or l > Upper, it is marked as an outlier;

[0108] 4) Smooth the outliers by the moving average method: where w = 2v + 1 is the window size, v is the window radius, L j is the load value corresponding to the index j in the original load data sequence, i is the index of the position of the outlier to be processed, j is the index of the load point traversed within the window for moving average calculation, and L′i is the corrected load value obtained after moving average smoothing of the outlier load value at position i.

[0109] Based on the load data set, perform temperature difference weighted calculation to obtain each weighted load curve:

[0110] Based on the outdoor temperature T corresponding to each load data in the load data set i Perform a difference calculation with the reference temperature of 26°C to obtain each temperature difference dif i = T i - 26°C;

[0111] Based on the load data with the outdoor temperature as the reference temperature in the load data set, perform an average value calculation to obtain the average load at the reference temperature n 26 is the number of load data with the outdoor temperature of 26°C, L k is the specific value of the k-th load point in the load data with the outdoor temperature of 26°C, and k is the index used to traverse the load data with the outdoor temperature of 26°C;

[0112] Based on each load data L in the load data set i Perform a difference calculation with the average load at the reference temperature to obtain each load difference

[0113] Based on the number of types of temperature differences n diffi and the adjusted weight coefficient w′ in the load data set, perform an average value calculation to obtain the average load at the reference temperature diff Perform a weighted processing on each load difference ΔL i to obtain each weighted load difference where N is the total number of samples;

[0114] Based on the weighted load difference and the average load at the reference temperature, the weighted load curves are calculated and drawn.

[0115] Weight coefficient w diff The calculation formula is:

[0116]

[0117] Where |diff| is the absolute value of the temperature difference; α is the weight attenuation coefficient, which ranges from [1, 2] and is used to adjust the attenuation rate of the weight as the temperature difference increases; when α = 1, the weight coefficient can be simplified to

[0118] Taking into account the influence of weather factors, the Weather Similarity Index (WSI) is introduced to adjust the weights:

[0119]

[0120] Where, T act is the current temperature; RH act is the current humidity; v act is the current wind speed; T ref is the temperature of the reference day; RH ref is the humidity on the reference day; v ref is the wind speed on the reference day; σ t is the first normalization parameter; σ rh is the second normalization parameter; σ v is the third normalization parameter.

[0121] Adjusted weight coefficient w′ diff =w diff ×WSI.

[0122] Based on the historical air conditioning load data and the weighted load curves, the Fréchet distance is calculated, and the curve with the smallest Fréchet distance is selected as the candidate optimal baseline:

[0123] Calculate the air conditioning load curve L for all dates throughout the summer total ={l1,l2,...,l t} and multiple weighted load curves L weighti The Fréchet distance F(l i ,l weighti ), the Fleche distance used here takes time sensitivity into consideration, assigns different weights to different time periods, and uses the weighted Fleche distance F w (l i ,L weighti ), and then calculate the average Fréchet distance for all dates And select the curve with the smallest average value As a candidate optimal baseline. i is the index used to traverse different dates in summer; t is the total number of days included in summer; l t is the air conditioning load curve on day t.

[0124] The Fréchet distance operator is Figure 3 As shown:

[0125]

[0126] Among them, the two curves A=a1,a2,...,a m and B=b1,b2,...,b n The calculation steps of the Fréchet distance F(A,B) between m is the mth point on curve A, b n is the nth point on curve B,

[0127] Construct an m×n Euclidean distance matrix d, where Represents point a in curve A i and point b on curve B j The Euclidean distance between

[0128] Construct the cumulative distance matrix D, which is recursively defined as:

[0129] D(1,1)=d(1,1), for i,j=1

[0130] D(i,1)=d(i,1)+D(i-1,1), for i>1

[0131] D(1,j)=d(1,j)+D(1,j-1), for j>

[0132] D(i,j)=d(i,j)+min(D(i-1,j),D(i,j-1),D(i-1,j-1)), for i,j>1.

[0133] The calculation formula of weighted Fréchet distance is as follows:

[0134]

[0135] Where, F w (A, B) is the weighted Fréchet distance between curve A and curve B; w k is the weight coefficient of time point k, the weight of working time period (9:00-18:00) is 1.5, and the weight of non-working time period is 1.0; a k b is the load data corresponding to curve A at time point K; kis the load data corresponding to curve B at time point k; d is the distance function used to calculate a k with b k k is the time point index, representing different moments; mn is the total number of time points, that is, the number of moments covered by the load curve.

[0136] Based on the weather parameters of the day, the candidate optimal baseline is corrected to obtain the optimal baseline of the day, where ΔT is the difference between the mean temperature and the reference temperature; ΔRH is the difference between the average relative humidity and the reference relative humidity; Figure 4 As shown:

[0137] Combined with the maximum temperature of the day, T max With the minimum temperature T min , for the candidate optimal baseline L min Make corrections to obtain the optimal baseline L at the current temperature current =L min ×f(T avg ,RH avg ,v avg ), where f(·) is the temperature-humidity correction function, T avg is the average temperature of the day, RH avg is the average relative humidity of the day, v avg The average wind speed for the day.

[0138] The temperature-load correction function is as follows:

[0139] L current =L min ×(1+k1×(T avg -26℃)+k2×(RH avg -RH ref ) 2 +k3×(T avg

[0140] -26℃)×(RH avg -RH ref ))

[0141] Where, L current is the optimal baseline at the current temperature; L min is the candidate optimal baseline; k1 is the temperature linear correction coefficient; k2 is the humidity quadratic correction coefficient; k3 is the temperature and humidity interaction correction coefficient; T avg is the average temperature of the day; T max is the maximum temperature of the day, T min is the lowest temperature of the day; RH avg The average relative humidity of the day; RH ref is the reference relative humidity (usually 50%).

[0142] The temperature linear correction coefficient k1, humidity quadratic correction coefficient k2, and temperature-humidity interaction correction coefficient k3 are determined by multiple linear regression:

[0143] [k1,k2,k3] T =(X T X) -1 X t Y

[0144] Where X is the design matrix, X=[x1,c2,...,x n ] T , x i =[(T avgi -26℃),(RH avgi -RH ref ) 2 (T avgi -26℃)×(RH avgi -RH ref )];x n is the nth element in the matrix;

[0145] Y is the response vector, L actual_i is the actual air conditioning load on day i; L min_i is the candidate optimal benchmark load corresponding to the i-th day, and the value of i ranges from 1 to n; L actual_n is the actual air conditioning load on the nth day; L min_n is the candidate optimal benchmark load corresponding to the nth day, T avgi is the average temperature on day i, RH avgi is the average relative humidity on day i.

[0146] In this embodiment, the optimal baseline is defined as the baseline with the lowest air conditioning load consumed when the indoor temperature is maintained in the human comfort range [22°C, 28°C] and the work efficiency of personnel is not affected by temperature (that is, when the indoor temperature fluctuates within the range of ±2°C, the work status of personnel does not change significantly).

[0147] Step 1 obtains the weather parameters of the day and historical air conditioning load data, specifically including:

[0148] Obtain the day's weather parameters and historical air conditioning load data;

[0149] Among them, weather parameters include: temperature, humidity and wind speed.

[0150] The calculation in step 2 based on the weather parameters of the day and the historical air conditioning load data is combined with the pre-built optimal baseline selection model to obtain the optimal air conditioning baseline for the day, specifically including:

[0151] The load data set is obtained after filtering and processing outliers based on historical air conditioning load data;

[0152] Based on the load data set, the temperature difference weighting and Fréchet distance calculation are performed to obtain the candidate optimal baseline;

[0153] The candidate optimal baseline is modified based on the weather parameters of the day to obtain the optimal baseline for air conditioning on that day.

[0154] This paper proposes a model for selecting an optimal baseline. This model considers the impact of temperature changes on air conditioning load and combines it with the Fréchet distance metric to effectively identify typical air conditioning power usage patterns. The model first collects air conditioning system operating data, calculates temperature difference weighting coefficients, and applies the Fréchet distance algorithm to evaluate load curve similarity, thereby determining the optimal baseline. This method overcomes the problems of traditional baseline selection, such as the neglect of temperature factors and the inaccurate distance calculation, significantly improving the accuracy of air conditioning energy efficiency assessment and abnormal operation identification. Experimental results demonstrate that the model exhibits excellent performance under different climate conditions and user types, providing a reliable basis for energy-saving optimization and fault diagnosis of air conditioning systems.

[0155] This embodiment implements a dynamic function to adjust the baseline. The core is to make the air-conditioning baseline adapt to the real-time changing environment and operating conditions. Due to fixed assumptions, the traditional baseline is difficult to accurately reflect the true optimal operating state of the air conditioner. Dynamic adjustment uses the weather parameters of the day and the historical load pattern to correct the baseline in real time to ensure that it fits the system energy efficiency baseline, providing an accurate basis for energy efficiency evaluation and anomaly identification, and solving the problem of mismatch between static baselines and dynamic scenarios. In industrial applications, the dynamically adjusted load baseline can provide support for coordinated scheduling on the power generation side, predict air-conditioning load fluctuations based on the baseline, assist in optimizing power generation plans, and trigger demand response; it can also help optimize industrial air-conditioning energy efficiency and fault warnings, by comparing baseline deviations to troubleshoot problems and diagnose faults; it can also serve the intelligent management and control of regional energy systems, coordinate cooling capacity distribution, match green electricity consumption, promote energy management to adapt to various scenarios, and provide key basis for power supply and demand balance, industrial energy conservation, and intelligent operation of energy systems from a macro perspective.

[0156] Example 2:

[0157] An air conditioning baseline selection system, the results are as follows Figure 5 Shown, including:

[0158] Data acquisition module, used to obtain the weather parameters of the day and historical air conditioning load data;

[0159] A model solving module, configured to calculate the optimal air conditioning baseline for the day based on the weather parameters of the day and historical air conditioning load data in combination with a pre-built optimal baseline selection model;

[0160] The optimal baseline selection model is constructed by correcting the candidate optimal baselines obtained by performing temperature difference weighting and Fréchet distance calculation based on historical air-conditioning load data using a dynamic correction function.

[0161] Also included is a model building module, the model building module:

[0162] The dataset construction submodule is used to filter and process the historical air conditioning load data according to the set outdoor temperature and season, and construct a load dataset based on the processed data;

[0163] A temperature difference weighted calculation submodule is used to perform temperature difference weighted calculation based on the load data set to obtain each weighted load curve;

[0164] a Fréchet distance calculation submodule, configured to calculate the Fréchet distance based on the historical air-conditioning load data and the weighted load curves, and select the curve with the smallest Fréchet distance as a candidate optimal baseline;

[0165] The baseline correction submodule is used to correct the candidate optimal baseline based on the weather parameters of the day to obtain the optimal baseline of the day.

[0166] The temperature difference weighted calculation submodule is specifically used to:

[0167] Based on the difference calculation between the outdoor temperature corresponding to each load data in the load data set and the reference temperature, each temperature difference is obtained;

[0168] The average value is calculated based on the load data with the outdoor temperature as the reference temperature in the load data set to obtain the average load at the reference temperature;

[0169] Performing a difference calculation based on each load data in the load data set and the average load at the reference temperature to obtain each load difference;

[0170] Each load difference is weighted based on the number of temperature difference types and the adjusted weight coefficient to obtain each weighted load difference;

[0171] Based on each weighted load difference and the average load at the reference temperature, the weighted load curves are calculated and drawn.

[0172] The calculation process of the re-adjusted weight coefficient of the temperature difference weighted calculation submodule includes:

[0173] The weight coefficient is obtained by calculating based on the absolute value of the temperature difference and the weight attenuation coefficient;

[0174] The weather similarity index is calculated based on the temperature, humidity and wind speed of the current day and the temperature, humidity and wind speed of the reference day;

[0175] The weight coefficient is adjusted and calculated based on the weather similarity index to obtain an adjusted weight coefficient.

[0176] The Fréchet distance calculation submodule is specifically used to:

[0177] Filtering summer air-conditioning load data according to season based on the historical air-conditioning load data, and plotting summer load curves based on the summer air-conditioning load data;

[0178] performing a weighted Fréchet distance calculation based on each of the summer load curves and each of the weighted load curves in combination with a preset time weight, to obtain a Fréchet distance from each of the summer load curves to each of the weighted load curves;

[0179] The average value of each Fréchet distance between each load curve in summer and each weighted load curve is calculated according to the date to obtain the average value of each Fréchet distance;

[0180] The curve with the smallest average value based on the Fréchet distance is selected as the candidate optimal baseline.

[0181] The calculation formula of the weighted Fréchet distance of the Fréchet distance calculation submodule is as follows:

[0182]

[0183] Where, F w (A, B) is the weighted Fréchet distance between curve A and curve B; w k is the weight coefficient at time point k; a k b is the load data corresponding to curve A at time point K; k is the load data corresponding to curve B at time point k; d is the distance function used to calculate a k with b k k is the time point index, representing different moments; mn is the total number of time points, that is, the number of moments covered by the load curve.

[0184] The baseline correction submodule is specifically used to:

[0185] Calculate based on the highest and lowest temperatures of the day to get the average temperature of the day;

[0186] Based on the average temperature, average relative humidity and reference relative humidity of the day, the candidate optimal baseline is corrected using the temperature-load correction function to obtain the optimal baseline of the day.

[0187] The calculation formula of the temperature-load correction function in the baseline correction submodule is as follows:

[0188] L current =L min ×(1+k1×(T avg -26℃)+k2×(RH avg -RH ref ) 2 +k3×(T avg -26℃)×(RH avg -RH ref ))

[0189] Where, L current is the optimal baseline at the current temperature; L min is the candidate optimal baseline; k1 is the temperature linear correction coefficient; k2 is the humidity quadratic correction coefficient; k3 is the temperature and humidity interaction correction coefficient; T avg is the average temperature of the day; RH avg The average relative humidity of the day; RH ref is the reference relative humidity.

[0190] The model solving module is specifically used for:

[0191] The load data set is obtained after filtering and processing outliers based on historical air conditioning load data;

[0192] Based on the load data set, the temperature difference weighting and Fréchet distance calculation are performed to obtain the candidate optimal baseline;

[0193] The candidate optimal baseline is modified based on the weather parameters of the day to obtain the optimal baseline for air conditioning on that day.

[0194] Example 3:

[0195] like Figure 6 As shown, the present invention also provides an electronic device, which may be a computer, a single-chip microcomputer, a smart mobile device, or the like. The electronic device in this embodiment may include a processor, a memory, a transceiver component, and the like. The memory, processor, and transceiver component are connected via a bus; the memory may be used to store an execution program, which may include instructions; and the processor may be used to execute the instructions stored in the memory. The memory may also be used to store data, which may be accessed and / or modified during the execution of the instructions.

[0196] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in the storage medium to implement the corresponding method flow or corresponding function, so as to implement the steps of an air-conditioning baseline selection method in the above embodiment.

[0197] Example 4

[0198] Based on the same inventive concept, the present invention also provides a readable storage medium, specifically an electronic device readable storage medium (Memory), which is a memory device in an electronic device for storing programs and data. It can be understood that the storage medium here can include both built-in storage media in the electronic device and, of course, extended storage media supported by the electronic device. The storage medium provides a storage space that stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space. These instructions can be one or more execution programs (including program codes). It should be noted that the storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The processor loads and executes one or more instructions stored in the storage medium to implement the steps of an air-conditioning baseline selection method in the above embodiment.

[0199] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0200] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0201] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0202] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0203] The above are merely embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application are included in the scope of the claims of the present application to be approved.

Claims

1. A method for selecting an air conditioning baseline, characterized in that: include: Obtain the day's weather parameters and historical air conditioning load data; Calculating the optimal air conditioning baseline for the day based on the weather parameters of the day and historical air conditioning load data in combination with a pre-built optimal baseline selection model; The optimal baseline selection model is constructed by correcting the candidate optimal baselines obtained by performing temperature difference weighting and Fréchet distance calculation based on historical air-conditioning load data using a dynamic correction function.

2. The method according to claim 1, characterized in that The construction process of the optimal baseline selection model includes: The historical air conditioning load data is filtered and outliers are processed according to the set outdoor temperature and season, and a load data set is constructed based on the processed data; Performing a weighted temperature difference calculation based on the load data set to obtain weighted load curves; Calculating the Fleche distance based on the historical air-conditioning load data and the weighted load curves, and selecting the curve with the smallest Fleche distance as a candidate optimal baseline; The candidate optimal baseline is modified based on the weather parameters of the day to obtain the optimal baseline of the day.

3. The method according to claim 2, characterized in that The weighted calculation of the temperature difference based on the load data set to obtain each weighted load curve includes: Based on the difference calculation between the outdoor temperature corresponding to each load data in the load data set and the reference temperature, each temperature difference is obtained; The average value is calculated based on the load data with the outdoor temperature as the reference temperature in the load data set to obtain the average load at the reference temperature; Performing a difference calculation based on each load data in the load data set and the average load at the reference temperature to obtain each load difference; Each load difference is weighted based on the number of temperature difference types and the adjusted weight coefficient to obtain each weighted load difference; Based on each weighted load difference and the average load at the reference temperature, the weighted load curves are calculated and drawn.

4. The method according to claim 3, characterized in that The calculation process of the adjusted weight coefficient includes: The weight coefficient is obtained by calculating based on the absolute value of the temperature difference and the weight attenuation coefficient; The weather similarity index is calculated based on the temperature, humidity and wind speed of the current day and the temperature, humidity and wind speed of the reference day; The weight coefficient is adjusted and calculated based on the weather similarity index to obtain an adjusted weight coefficient.

5. The method according to claim 2, characterized in that The Fréchet distance calculation based on the historical air-conditioning load data and the weighted load curves is performed, and the curve with the smallest Fréchet distance is selected as the candidate optimal baseline, including: Filtering summer air-conditioning load data according to season based on the historical air-conditioning load data, and plotting summer load curves based on the summer air-conditioning load data; performing a weighted Fréchet distance calculation based on each of the summer load curves and each of the weighted load curves in combination with a preset time weight, to obtain a Fréchet distance from each of the summer load curves to each of the weighted load curves; The average value of each Fréchet distance between each load curve in summer and each weighted load curve is calculated according to the date to obtain the average value of each Fréchet distance; Based on the average value of each Fréchet distance, the curve with the smallest value is selected as the candidate optimal baseline.

6. The method according to claim 5, characterized in that The calculation formula of the weighted Fréchet distance is as follows: Where, F w (A, B) is the weighted Fréchet distance between curve A and curve B; w k is the weight coefficient at time point k; a k is the load data corresponding to curve A at time point K; b k is the load data corresponding to curve B at time point k; d is the distance function used to calculate a k with b k k is the time point index, representing different moments; mn is the total number of time points, that is, the number of moments covered by the load curve.

7. The method according to claim 2, characterized in that The step of correcting the candidate optimal baseline based on the weather parameters of the day to obtain the optimal baseline of the day includes: Calculate based on the highest and lowest temperatures of the day to get the average temperature of the day; Based on the average temperature, average relative humidity and reference relative humidity of the day, the candidate optimal baseline is corrected using the temperature-load correction function to obtain the optimal baseline of the day.

8. The method according to claim 7, characterized in that The calculation formula of the temperature-load correction function is as follows: L current =L min ×(1+k1×(T avg -26℃)+k2×(RH avg -HR ref ) 2 +k3×(T avg -26℃)×(RH avg -HR ref )) Where, L current is the optimal baseline at the current temperature; L min is the candidate optimal baseline; k1 is the temperature linear correction coefficient; k2 is the humidity quadratic correction coefficient; k3 is the temperature and humidity interaction correction coefficient; T avg is the average temperature of the day; RH avg The average relative humidity of the day; RH ref is the reference relative humidity.

9. The method according to claim 1, characterized in that The calculation based on the weather parameters of the day and the historical air conditioning load data is combined with a pre-built optimal baseline selection model to obtain the optimal air conditioning baseline for the day, including: The load data set is obtained after filtering and processing outliers based on historical air conditioning load data; Based on the load data set, the temperature difference weighting and Fréchet distance calculation are performed to obtain the candidate optimal baseline; The candidate optimal baseline is modified based on the weather parameters of the day to obtain the optimal baseline for air conditioning on that day.

10. An air conditioning baseline selection system, characterized in that: include: Data acquisition module, used to obtain the weather parameters of the day and historical air conditioning load data; A model solving module, configured to calculate the optimal air conditioning baseline for the day based on the weather parameters of the day and historical air conditioning load data in combination with a pre-built optimal baseline selection model; The optimal baseline selection model is constructed by correcting the candidate optimal baselines obtained by performing temperature difference weighting and Fréchet distance calculation based on historical air-conditioning load data using a dynamic correction function.

11. The system according to claim 10, wherein: Also included is a model building module, the model building module: The dataset construction submodule is used to filter and process the historical air conditioning load data according to the set outdoor temperature and season, and construct a load dataset based on the processed data; A temperature difference weighted calculation submodule is used to perform temperature difference weighted calculation based on the load data set to obtain each weighted load curve; a Fréchet distance calculation submodule, configured to calculate the Fréchet distance based on the historical air-conditioning load data and the weighted load curves, and select the curve with the smallest Fréchet distance as a candidate optimal baseline; The baseline correction submodule is used to correct the candidate optimal baseline based on the weather parameters of the day to obtain the optimal baseline of the day.

12. The system according to claim 11, wherein: The temperature difference weighted calculation submodule is specifically used to: Based on the difference calculation between the outdoor temperature corresponding to each load data in the load data set and the reference temperature, each temperature difference is obtained; The average value is calculated based on the load data with the outdoor temperature as the reference temperature in the load data set to obtain the average load at the reference temperature; Performing a difference calculation based on each load data in the load data set and the average load at the reference temperature to obtain each load difference; Each load difference is weighted based on the number of temperature difference types and the adjusted weight coefficient to obtain each weighted load difference; Based on each weighted load difference and the average load at the reference temperature, the weighted load curves are calculated and drawn.

13. The system according to claim 12, wherein: The calculation process of the re-adjusted weight coefficient of the temperature difference weighted calculation submodule includes: The weight coefficient is obtained by calculating based on the absolute value of the temperature difference and the weight attenuation coefficient; The weather similarity index is calculated based on the temperature, humidity and wind speed of the current day and the temperature, humidity and wind speed of the reference day; The weight coefficient is adjusted and calculated based on the weather similarity index to obtain an adjusted weight coefficient.

14. The system according to claim 11, wherein: The Fréchet distance calculation submodule is specifically used to: Filtering summer air-conditioning load data according to season based on the historical air-conditioning load data, and plotting summer load curves based on the summer air-conditioning load data; performing a weighted Fréchet distance calculation based on each of the summer load curves and each of the weighted load curves in combination with a preset time weight, to obtain a Fréchet distance from each of the summer load curves to each of the weighted load curves; The average value of each Fréchet distance between each load curve in summer and each weighted load curve is calculated according to the date to obtain the average value of each Fréchet distance; Based on the average value of each Fréchet distance, the curve with the smallest value is selected as the candidate optimal baseline.

15. The system according to claim 14, wherein: The calculation formula of the weighted Fréchet distance of the Fréchet distance calculation submodule is as follows: Where, F w (A, B) is the weighted Fréchet distance between curve A and curve B; w k is the weight coefficient at time point k; a k is the load data corresponding to curve A at time point K; b k is the load data corresponding to curve B at time point k; d is the distance function used to calculate a k with b k k is the time point index, representing different moments; mn is the total number of time points, that is, the number of moments covered by the load curve.

16. The system according to claim 11, wherein: The baseline correction submodule is specifically used to: Calculate based on the highest and lowest temperatures of the day to get the average temperature of the day; Based on the average temperature, average relative humidity and reference relative humidity of the day, the candidate optimal baseline is corrected using the temperature-load correction function to obtain the optimal baseline of the day.

17. The system according to claim 16, wherein: The calculation formula of the temperature-load correction function in the baseline correction submodule is as follows: L current =L min ×(1+k1×(T avg -26℃)+k2×(RH avg -HR ref ) 2 +k3×(T avg -26℃)×(RH avg -HR ref )) Where, L current is the optimal baseline at the current temperature; L min is the candidate optimal baseline; k1 is the temperature linear correction coefficient; k2 is the humidity quadratic correction coefficient; k3 is the temperature and humidity interaction correction coefficient; T avg is the average temperature of the day; RH avg The average relative humidity of the day; RH ref is the reference relative humidity.

18. The system according to claim 10, wherein: The model solving module is specifically used for: The load data set is obtained after filtering and processing outliers based on historical air conditioning load data; Based on the load data set, the temperature difference weighting and Fréchet distance calculation are performed to obtain the candidate optimal baseline; The candidate optimal baseline is modified based on the weather parameters of the day to obtain the optimal baseline for air conditioning on that day.

19. An electronic device, characterized in that: include: at least one processor and memory; The memory and the processor are connected via a bus; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, an air-conditioning baseline selection method according to any one of claims 1 to 9 is implemented.

20. A readable storage medium, characterized in that An execution program is stored thereon, and when the execution program is executed, an air-conditioning baseline selection method according to any one of claims 1 to 9 is implemented.

Images