A method for non-intrusive identification of central air conditioning operating load

Abstract

This invention provides a non-intrusive method for identifying the operating load of central air conditioning systems. It selects central air conditioning units of different brands, records their complete operating electrical waveforms, performs time-frequency domain feature analysis to extract common features, and selects the user's total daily electricity load data and daily temperature data from the past year. The method analyzes the correlation between the user's daily electricity load curve and daily temperature curve to obtain the heating load area, the base load area, and the cooling load area. Then, it performs linear data fitting with the temperature to obtain the load curve. Subtracting the cooling load curve from the base load curve yields the formula for the relationship between the user's central air conditioning steady-state power and temperature. Furthermore, it determines whether the central air conditioning system is in a start-up / stop state. During user air conditioning start-up and stop events, it uses partial power estimation and integrates the steady-state power to obtain the power consumption of the central air conditioning equipment during operation, thus achieving central air conditioning operating status tracking and power consumption estimation.

Description

Technical Field

[0001] This invention relates to the field of power load monitoring in power systems, and more particularly to a non-intrusive identification method for central air conditioning operating load that provides accurate characteristics of its operating status and power consumption breakdown. Background Technology

[0002] Summer and winter are the peak seasons for electricity consumption in my country. Among them, air conditioning load accounts for a relatively high proportion of the electricity consumption and largely overlaps with the peak electricity consumption of the power grid. Therefore, in the work of demand response, it is necessary to optimize the allocation of power resources by adjusting the air conditioning load.

[0003] Traditional methods of adjusting air conditioning load rely heavily on monitoring on-site / surveying the actual situation of users to statistically analyze and investigate load resources. This method obviously cannot obtain real-time information on air conditioning load and start / stop status on a large scale, lacks precise control over air conditioning resources, and causes data distortion.

[0004] Current technology typically employs the method of installing interventional monitoring equipment on the equipment side to monitor air conditioning load. However, practical field operations have revealed that this method involves significant investment and maintenance, resulting in high costs. Therefore, most electricity users opt for non-interventional load monitoring solutions.

[0005] However, existing non-intrusive load monitoring solutions suffer from the problem of a wide variety of air conditioner types, brands, and power ranges on the market, making it impossible to establish a complete load feature library. Consequently, load identification is quite difficult. Therefore, existing non-intrusive load identification technologies can only identify air conditioning equipment on the market that has completed on-site waveform recording and feature library establishment. This results in incomplete models, poor versatility and inaccuracy, and unsatisfactory practical effects.

[0006] Based on the above problems, a new method for extracting air conditioning load features is needed to enable the widespread application of the identification model. Summary of the Invention

[0007] To address the issues of high cost of existing interventional monitoring equipment and the shortcomings of non-interventional load monitoring solutions, such as a lack of load feature databases, high difficulty in load identification, poor versatility, and inaccuracy, this invention proposes a non-interventional load identification method suitable for central air conditioning operating loads. This method improves the completeness of the load identification model by adopting a composite mode based on the operating characteristics of each household's air conditioning load and electrical feature extraction. Furthermore, it can complete parameter training for each household without conducting on-site surveys, effectively enhancing the versatility of the identification model while achieving accurate and efficient identification.

[0008] The present invention provides a non-intrusive method for identifying the operating load of a central air conditioning system, the specific steps of which are as follows:

[0009] 1) First, select central air conditioning units of different brands, record their complete operating electrical waveforms, and perform time-frequency domain feature analysis to extract common features present in the operation of different brands of air conditioning units, namely the startup peak value. Steady-state power value Stop function change Downtime and current distortion rate A total of 5 common characteristic parameters are collected, and a typical air conditioner operating power curve is formed; in this step, 5 or more central air conditioners of different brands are selected, and the electrical waveform of the complete operation process is recorded 1 or more times.

[0010] It should be noted that the steady-state power of different air conditioning loads is closely related to the air conditioner's horsepower and the current temperature difference.

[0011] 2) This step first selects the user's total daily electricity load data and daily temperature data from the past year for different users and different central air conditioning models, and analyzes the correlation between the user's daily electricity load curve and daily temperature curve by region. Specifically:

[0012] Based on the daily electricity load curve data collected by users at 96:00 AM and daily temperature data, the load aggregation density at different temperatures is calculated. ,as follows:

[0013]

[0014] In the formula: This represents the number of times temperature t and power p appear in the daily load curve data for one year;

[0015] Based on actual conditions, heating load areas, base load areas, and cooling load areas were divided. Linear data fitting was performed between the data of different load areas and temperature, and regression clustering was used to obtain load regression curves for these three areas.

[0016]

[0017] In the formula: p is power, t is temperature, These are the slope and intercept of the k-th fitted curve, respectively, where k = 1, 2, 3;

[0018] The load regression curve is divided into cooling load curve. Basic load curve and heating load There are three curves in total. By subtracting the cooling load curve from the base load curve, the relationship between the steady-state power and temperature of the central air conditioning system for different users can be obtained.

[0019] 3) Finally, the load identification module updates the identification program based on the obtained steady-state power parameters, and calculates the power characteristic value using the original high-frequency voltage and current data. The obtained power characteristic data is further calculated using the sliding window bilateral CUSUM algorithm to determine whether the central air conditioning is in the start / stop state. During the user's air conditioning start-up and stop events, the power consumption is estimated by sub-items, and the power consumption of the central air conditioning equipment during operation is obtained by steady-state power integration. Finally, the operation status tracking and power consumption estimation of the central air conditioning are obtained.

[0020] According to a method for non-intrusive identification of central air conditioning operating load based on the present invention, the typical air conditioning operating power curve in step 1) specifically includes the following values: peak value at startup. The maximum power reached during the air conditioner's startup process, and the steady-state power value. The power average during stable operation of the air conditioner, and the power change when the air conditioner is off. The total power change during the air conditioner shutdown process, and the shutdown duration. The total duration of the air conditioner shutdown process, and the current distortion rate. This is the effective value of all non-fundamental current components - including harmonics and interharmonics - divided by the effective value of the fundamental current at 50Hz.

[0021] According to a method for non-intrusive identification of central air conditioning operating load based on the present invention, the starting peak value is characterized in that... Choose a steady-state power value of 1.2 to 3 times. And shutdown function change Then select a steady-state power value of 0.8 to 1.2 times. .

[0022] According to a method for non-intrusive identification of central air conditioning operating load according to the present invention, the "adopting regression clustering" in step 2) specifically adopts a Gaussian mixture regression clustering model, sets the number of clusters to 3, and simultaneously calculates the fitted load regression curve.

[0023] According to a method for non-intrusive identification of central air conditioning operating loads according to the present invention, step 2) includes data preprocessing, specifically:

[0024] Set the load aggregation density threshold e1=0.6, when the load aggregation density When <= e1, the corresponding sample data point (t, p) is removed.

[0025] According to a method for non-intrusive identification of central air conditioning operating load based on the present invention, the linear data fitting in step 2) is specifically optimized using a Gaussian mixture regression clustering model, as follows:

[0026] 1) Optimization objective function:

[0027]

[0028] In the formula: Let be the slope and intercept terms of the k-th fitted curve, where k = 1, 2, 3;

[0029] It represents the proportion of the k-th regression fit class in the total number of classes;

[0030] It is the standard deviation of the error distribution of the k-th fitted curve;

[0031] The sample size is N, (t i ,p i () represents the weighted data sample points;

[0032] Let be the probability density function of the normal distribution, which is as follows:

[0033]

[0034] Where e is the base of the natural logarithm, with a value of 2.718;

[0035] 2) Initialize parameters:

[0036]

[0037] 3) E-step expectation maximization:

[0038] Calculate the probability p of fitting the curve for the i-th sample, i=1,2,…,N to the k-th sample, k=1,2,3. ik ,

[0039] Specific gravity πk, as follows:

[0040] ;

[0041] 4) Calculate the regression coefficients of the new fitted regression curve. And calculate the standard deviation of the regression fit. ,in, Specifically as follows:

[0042] ;

[0043] 5) Determine the new Compared with the previous step If the absolute value of the difference between the two is greater than the stopping error limit η = 10^(-5), continue repeating steps 7.3) and 7.4). Otherwise, stop the loop and output the final fitted regression curve. .

[0044] According to a method for non-intrusive identification of central air conditioning operating load based on the present invention, the sliding window bilateral CUSUM algorithm in step 3) is characterized in that:

[0045] Let the power characteristic time series be The mean calculation window and the transient detection window are W, respectively. m and W d The window lengths are defined as m and n, where currently m=n=128, and the average power within the two windows is M. m and M d The calculation methods are as follows:

[0046] ;

[0047] ;

[0048] In the formula, L is the first sampling point of the mean calculation window.

[0049] According to a method for non-intrusive identification of central air conditioning operating load based on the present invention, the identification method for "determining whether the central air conditioning is in a start / stop state" in step 3) is specifically as follows:

[0050] 1) Air conditioner start-up status identification, defining the cumulative start-up event sum. The calculation formula is:

[0051] ,

[0052] In the formula, β represents the fluctuation level when the power is stable, and when the detection window M d >M m When +3*β, it is determined to be a suspected device startup event. Start by incrementing from 0. If after 5 minutes... The value is 1.2 to 3 times the steady-state power of the user's central air conditioning system, which is used to determine that the user's central air conditioning equipment has started.

[0053] 2) Air conditioner shutdown status identification, defining the cumulative shutdown event sum. The calculation formula is:

[0054] ,

[0055] In the formula, β represents the fluctuation level when the power is stable, and when the detection window M d <M m When the value is +3*β, it is determined to be a suspected equipment shutdown event. Start by incrementing from 0. If after 1 minute... If the value is 0.8 to 1.2 times the steady-state power of the user's central air conditioning system, it is determined that the user's central air conditioning equipment is shut down.

[0056] According to a method for non-intrusive identification of central air conditioning operating load based on the present invention, the method is characterized in that, in step 3), "obtaining the power consumption of the central air conditioning equipment during operation using steady-state power integration" specifically involves estimating the power consumption of individual components during user air conditioning start-up and shutdown events, and then integrating the steady-state power. The electricity consumption of the central air conditioning equipment during operation is obtained.

[0057] The following beneficial effects were achieved by using the non-intrusive identification method for central air conditioning operating load according to the present invention:

[0058] 1. The present invention provides a non-intrusive identification method for the operating load of central air conditioning, which provides a novel feature extraction dimension and feature extraction and identification method, can achieve the integrity of the load identification model, and can complete the parameter training of "one household, one policy" without the need for on-site surveys and waveform recording.

[0059] 2. The present invention provides a non-intrusive identification method for the operating load of central air conditioning systems, which solves the problems of poor generalization of load identification models and large and difficult-to-obtain training data for the equipment to be identified. It can effectively improve the versatility of the identification model, and is accurate and efficient. Attached Figure Description

[0060] Figure 1This is a typical power curve of a central air conditioning unit, which is a method for non-intrusive identification of the operating load of a central air conditioning unit according to the present invention.

[0061] Figure 2 This is a typical diagram showing the relationship between power and temperature of a central air conditioning unit, which is applicable to a non-intrusive identification method for the operating load of a central air conditioning unit according to the present invention. Detailed Implementation

[0062] The following description, in conjunction with the accompanying drawings and embodiments, further describes the technical means, creative features, objectives, and effects of a method for non-intrusive identification of central air conditioning operating load according to the present invention.

[0063] A method for non-intrusive identification of central air conditioning operating load, the specific steps of which are as follows:

[0064] 1) First, select central air conditioning units of different brands, record their complete operating electrical waveforms, and perform time-frequency domain feature analysis to extract common features present in the operation of different brands of air conditioning units, namely the startup peak value. Steady-state power value Stop function change Downtime and current distortion rate A total of 5 common characteristic parameters are collected, and a typical air conditioner operating power curve is formed; in this step, 5 or more central air conditioners of different brands are selected, and the electrical waveform of the complete operation process is recorded 1 or more times.

[0065] 2) This step first selects the user's total daily electricity load data and daily temperature data from the past year for different users and different central air conditioning models, and analyzes the correlation between the user's daily electricity load curve and daily temperature curve by region. Specifically:

[0066] Based on the daily electricity load curve data collected by users at 96:00 AM and daily temperature data, the load aggregation density at different temperatures is calculated. ,as follows:

[0067]

[0068] In the formula: This represents the number of times temperature t and power p appear in the daily load curve data for one year;

[0069] Based on actual conditions, heating load areas, base load areas, and cooling load areas were divided. Linear data fitting was performed between the data of different load areas and temperature, and regression clustering was used to obtain load regression curves for these three areas.

[0070]

[0071] In the formula: p is power, t is temperature, These are the slope and intercept of the k-th fitted curve, respectively, where k = 1, 2, 3;

[0072] The load regression curve is divided into cooling load curve. Basic load curve and heating load There are three curves in total. By subtracting the cooling load curve from the base load curve, the relationship between the steady-state power and temperature of the central air conditioning system for different users can be obtained.

[0073] 3) Finally, the load identification module updates the identification program based on the obtained steady-state power parameters, and calculates the power characteristic value using the original high-frequency voltage and current data. The obtained power characteristic data is further calculated using the sliding window bilateral CUSUM algorithm to determine whether the central air conditioning is in the start / stop state. During the user's air conditioning start-up and stop events, the power consumption is estimated by sub-items, and the power consumption of the central air conditioning equipment during operation is obtained by steady-state power integration. Finally, the operation status tracking and power consumption estimation of the central air conditioning are obtained.

[0074] The typical air conditioner operating power curve in step 1) has the following specific values: peak value at startup. The maximum power reached during the air conditioner's startup process, and the steady-state power value. The power average during stable operation of the air conditioner, and the power change when the air conditioner is off. The total power change during the air conditioner shutdown process, and the shutdown duration. The total duration of the air conditioner shutdown process, and the current distortion rate. This is the effective value of all non-fundamental current components divided by the effective value of the fundamental current at 50Hz. Non-fundamental current components include harmonics and interharmonics.

[0075] Startup peak Choose a steady-state power value of 1.2 to 3 times. And shutdown function change Then select a steady-state power value of 0.8 to 1.2 times. .

[0076] Step 2) “Use regression clustering” specifically means using a Gaussian mixture regression clustering model, setting the number of clusters to 3, and simultaneously calculating the fitted load regression curve.

[0077] Step 2) includes data preprocessing, which specifically includes:

[0078] Set the load aggregation density threshold e1=0.6, when the load aggregation density When <= e1, the corresponding sample data point (t, p) is removed.

[0079] Step 2) involves linear data fitting, specifically using a Gaussian mixture regression clustering model for fitting optimization, as follows:

[0080] 1) Optimization objective function:

[0081]

[0082] In the formula: Let be the slope and intercept terms of the k-th fitted curve, where k = 1, 2, 3;

[0083] It represents the proportion of the k-th regression fit class in the total number of classes;

[0084] It is the standard deviation of the error distribution of the k-th fitted curve;

[0085] The sample size is N, (t i ,p i () represents the weighted data sample points;

[0086] Let be the probability density function of the normal distribution, which is as follows:

[0087]

[0088] Where e is the base of the natural logarithm, with a value of 2.718;

[0089] 2) Initialize parameters:

[0090]

[0091] 3) E-step expectation maximization:

[0092] Calculate the probability p of fitting the curve for the i-th sample, i=1,2,…,N to the k-th sample, k=1,2,3. ik ,

[0093] Specific gravity πk, as follows:

[0094] ;

[0095] 4) Calculate the regression coefficients of the new fitted regression curve. And calculate the standard deviation of the regression fit. ,in, Specifically as follows:

[0096] ;

[0097] 5) Determine the new Compared with the previous step If the absolute value of the difference between the two is greater than the stopping error limit η = 10^(-5), continue repeating steps 7.3) and 7.4). Otherwise, stop the loop and output the final fitted regression curve. .

[0098] Step 3) of the sliding window bilateral CUSUM algorithm is as follows:

[0099] Let the power characteristic time series be The mean calculation window and the transient detection window are W, respectively. m and W d The window lengths are defined as m and n, where currently m=n=128, and the average power within the two windows is M. m and M d The calculation methods are as follows:

[0100] ;

[0101] ;

[0102] In the formula, L is the first sampling point of the mean calculation window.

[0103] The specific method for identifying whether the central air conditioning system is in start / stop state in step 3) is as follows:

[0104] 1) Air conditioner start-up status identification, defining the cumulative start-up event sum. The calculation formula is:

[0105] ,

[0106] In the formula, β represents the fluctuation level when the power is stable, and when the detection window M d >M m When +3*β, it is determined to be a suspected device startup event. Start by incrementing from 0. If after 5 minutes... The value is 1.2 to 3 times the steady-state power of the user's central air conditioning system, which is used to determine that the user's central air conditioning equipment has started.

[0107] 2) Air conditioner shutdown status identification, defining the cumulative shutdown event sum. The calculation formula is:

[0108] ,

[0109] In the formula, β represents the fluctuation level when the power is stable, and when the detection window M d <M m When the value is +3*β, it is determined to be a suspected equipment shutdown event. Start by incrementing from 0. If after 1 minute... If the value is 0.8 to 1.2 times the steady-state power of the user's central air conditioning system, it is determined that the user's central air conditioning equipment is shut down.

[0110] Step 3), "obtaining the power consumption of the central air conditioning equipment during operation using steady-state power integration," specifically involves estimating the power consumption for each item during user air conditioning start-up and shutdown events, and then integrating the results using steady-state power integration. The electricity consumption of the central air conditioning equipment during operation is obtained.

[0111] Example

[0112] Based on the above steps, load identification results were verified using data from a user from June 11th to July 7th. The results (as shown in Table 1) were compared with the data from the guide rail meter installed on the air conditioner side. Analysis revealed that the user's air conditioning equipment actually started and stopped 23 times, and the load identification model identified 23 start and stop times. The accuracy rates for load type identification and load status identification were both 100%. During the air conditioning operation, the guide rail meter measured 9036 kWh of electricity, and the algorithm estimated 9282 kWh of electricity, with an estimation deviation of 3%. This demonstrates that the identification model of the non-intrusive load identification method for central air conditioning operation of this invention has strong versatility, high accuracy, high efficiency, and small error.

[0113] date Rail meter power (kWh) Load identification power (kWh) Identify battery level / Actual battery level Actual number of start-stops Identify the number of stops June 11 267 282 106% 1 1 June 12 301 263 87% 1 1 June 13 407 417 102% 2 2 June 14 445 411 92% 1 1 June 15 157 145 92% 1 1 June 16 0 0 / / / June 18 437 433 99% 1 1 June 19 353 375 106% 1 1 June 20 388 375 97% 1 1 June 21 410 397 97% 1 1 June 22 219 228 104% 1 1 June 23 151 65 43% 1 1 June 25 324 337 96% 1 1 June 26 358 383 107% 1 1 June 27 303 289 105% 1 1 June 28 379 379 100% 1 1 June 29 0 0 / / / June 30 0 0 / / / July 2 717 899 125% 1 1 July 3 540 609 89% 1 1 July 4th 733 721 102% 1 1 July 5 996 1155 116% 2 2 July 6 570 569 100% 1 1 July 7 581 550 95% 1 1

[0114] Table 1 - Comparison of Central Air Conditioning Identification and Verification Results for a Certain User

[0115] This invention provides a non-intrusive method for identifying the operating load of central air conditioning systems. It offers a novel feature extraction dimension and a feature extraction and identification method, enabling the completeness of the load identification model and allowing for parameter training tailored to each household without the need for on-site surveys or waveform recording. Furthermore, this invention solves the problems of poor generalization of the load identification model and the large and difficult-to-obtain training data for the devices to be identified. It effectively improves the versatility of the identification model, while also providing accurate and efficient identification.

[0116] This invention is applicable to the field of power load monitoring and power consumption analysis of central air conditioning operating load.

[0117] Furthermore, those skilled in the art should recognize that the above embodiments are merely illustrative of this application and are not intended to limit this application. Any variations or modifications to the above embodiments that fall within the spirit and essence of this application will fall within the scope of the claims of this application.

Claims

1. A method for non-intrusive identification of central air conditioning operating load, the specific steps of which are as follows: 1) First, select central air conditioning units of different brands, record their complete operating electrical waveforms, and perform time-frequency domain feature analysis to extract common features present in the operation of different brands of air conditioning units, namely the startup peak value. Steady-state power value Stop function change Downtime and current distortion rate A total of 5 common characteristic parameters are collected, and a typical air conditioner operating power curve is formed; in this step, 5 or more central air conditioners of different brands are selected, and the electrical waveform of the complete operation process is recorded 1 or more times. 2) This step first selects the user's total daily electricity load data and daily temperature data from the past year for different users and different central air conditioning models, and analyzes the correlation between the user's daily electricity load curve and daily temperature curve by region. Specifically: Based on the daily electricity load curve data collected by users at 96:00 AM and daily temperature data, the load aggregation density at different temperatures is calculated. ,as follows: , In the formula: This represents the number of times temperature t and power p appear in the daily load curve data for one year; Based on actual conditions, heating load areas, base load areas, and cooling load areas were divided. Linear data fitting was performed between the data of different load areas and temperature, and regression clustering was used to obtain load regression curves for these three areas. , In the formula: p is power, t is temperature, These are the slope and intercept of the k-th fitted curve, respectively, where k = 1, 2, 3; The load regression curve is divided into cooling load curve. Basic load curve and heating load There are three curves in total. By subtracting the cooling load curve from the base load curve, the relationship between the steady-state power and temperature of the central air conditioning system for different users can be obtained. 3) Finally, the load identification module updates the identification program based on the obtained steady-state power parameters, and calculates the power characteristic value using the original high-frequency voltage and current data. The obtained power characteristic data is further calculated using the sliding window bilateral CUSUM algorithm to determine whether the central air conditioning is in the start / stop state. During the user's air conditioning start-up and stop events, the power consumption is estimated by sub-items, and the power consumption of the central air conditioning equipment during operation is obtained by steady-state power integration. Finally, the operation status tracking and power consumption estimation of the central air conditioning are obtained.

2. The method for non-intrusive identification of central air conditioning operating load as described in claim 1, characterized in that, The typical air conditioner operating power curve in step 1) has the following specific values: peak value at startup. The maximum power reached during the air conditioner's startup process, and the steady-state power value. The power average during stable operation of the air conditioner, and the power change when the air conditioner is off. The total power change during the air conditioner shutdown process, and the shutdown duration. The total duration of the air conditioner shutdown process, and the current distortion rate. This is the effective value of all non-fundamental current components divided by the effective value of the fundamental current at 50Hz. Non-fundamental current components include harmonics and interharmonics.

3. The method for non-intrusive identification of central air conditioning operating load as described in claim 1, characterized in that, The aforementioned startup peak Choose a steady-state power value of 1.2 to 3 times. And shutdown function change Then select a steady-state power value of 0.8 to 1.2 times. .

4. The method for non-intrusive identification of central air conditioning operating load as described in claim 1, characterized in that, In step 2), "using regression clustering" specifically means using a Gaussian mixture regression clustering model, setting the number of clusters to 3, and simultaneously calculating the fitted load regression curve.

5. The method for non-intrusive identification of central air conditioning operating load as described in claim 1, characterized in that, Step 2) includes data preprocessing, specifically: Set the load aggregation density threshold e1=0.6, when the load aggregation density When <= e1, the corresponding sample data point (t, p) is removed.

6. The method for non-intrusive identification of central air conditioning operating load as described in claim 1, characterized in that, The linear data fitting in step 2) specifically involves fitting optimization using a Gaussian mixture regression clustering model, as follows: 1) Optimization objective function: , In the formula: Let be the slope and intercept terms of the k-th fitted curve, where k = 1, 2, 3; It represents the proportion of the k-th regression fit class in the total number of classes; It is the standard deviation of the error distribution of the k-th fitted curve; The sample size is N, (t i ,p i () represents the weighted data sample points; Let be the probability density function of the normal distribution, which is as follows: , Where e is the base of the natural logarithm, with a value of 2.718; 2) Initialize parameters: , 3) E-step expectation maximization: Calculate the probability p of fitting the curve for the i-th sample, i=1,2,…,N to the k-th sample, k=1,2,3. ik , Specific gravity π k ,as follows: ； 4) Calculate the regression coefficients of the new fitted regression curve. And calculate the standard deviation of the regression fit. ,in, Specifically as follows: ； 5) Determine the new Compared with the previous step If the absolute value of the difference between the two is greater than the stopping error limit η = 10^(-5), continue repeating steps 7.3) and 7.4). Otherwise, stop the loop and output the final fitted regression curve. .

7. The method for non-intrusive identification of central air conditioning operating load as described in claim 1, characterized in that, The sliding window bilateral CUSUM algorithm in step 3) is specifically as follows: Let the power characteristic time series be The mean calculation window and the transient detection window are W, respectively. m and W d The window lengths are defined as m and n, where currently m=n=128, and the average power within the two windows is M. m and M d The calculation methods are as follows: ； ； In the formula, L is the first sampling point of the mean calculation window.

8. The method for non-intrusive identification of central air conditioning operating load as described in claim 7, characterized in that, The specific method for identifying whether the central air conditioning system is in start-up / stop state in step 3) is as follows: 1) Air conditioner start-up status identification, defining the cumulative start-up event sum. The calculation formula is: ， In the formula, β represents the fluctuation level when the power is stable, and when the detection window M d >M m When +3*β, it is determined to be a suspected device startup event. Start by incrementing from 0. If after 5 minutes... The value is 1.2 to 3 times the steady-state power of the user's central air conditioning system, which is used to determine that the user's central air conditioning equipment has started. 2) Air conditioner shutdown status identification, defining the cumulative shutdown event sum. The calculation formula is: ， In the formula, β represents the fluctuation level when the power is stable, and when the detection window M d <M m When the value is +3*β, it is determined to be a suspected equipment shutdown event. Start by incrementing from 0. If after 1 minute... If the value is 0.8 to 1.2 times the steady-state power of the user's central air conditioning system, it is determined that the user's central air conditioning equipment is shut down.

9. The method for non-intrusive identification of central air conditioning operating load as described in claim 1, characterized in that, Step 3) of the above, "obtaining the power consumption of the central air conditioning equipment during operation using steady-state power integration," specifically involves estimating the power consumption for each item during user air conditioning start-up and shutdown events, and then using steady-state power integration. The electricity consumption of the central air conditioning equipment during operation is obtained.

Images