Building power load prediction method, device and system applied to VPP scenario
By combining time series method and regression analysis method, using historical data to fit polynomial linear regression curve function, the inaccuracy problem of load prediction method when dealing with abnormal data and complex relationships is solved, and the power load prediction in VPP scenario is realized , provides better power resource allocation and scheduling strategies.
Patent Information
- Application Number
- CN202410235017.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-03-01
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-03-01
AI Technical Summary
The existing load prediction methods are inaccurate when dealing with abnormal data or complex relationships, and require a large amount of real-time data. The total power load data for building use is difficult to obtain, affecting the prediction accuracy.
Combining the time series method and regression analysis method, the polynomial linear regression curve function is fitted through historical power load data and weather data, similar historical days are selected, and the correction coefficient is calculated to predict the power load changes on the target day.
In the VPP scenario, use the most limited parameters to accurately predict power load changes, provide better power resource allocation and scheduling strategies, and improve prediction accuracy and computing efficiency.
Smart Images

Figure CN118100157B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of virtual electric field load peak regulation, and specifically relates to a building power load prediction method, device and system applied to a VPP scenario. Background Art
[0002] With the development of global energy transformation and intelligent power system, virtual power plant (VPP) as a new form of power system is gradually gaining attention and importance. Virtual power plant is a form of power system that integrates, optimizes and coordinates decentralized distributed energy systems through advanced information technology. It connects various distributed energy devices (such as photovoltaic, wind power and energy storage) through Internet technology to form a virtual power plant, realizing unified management and optimal allocation of power resources.
[0003] In the virtual power plant scenario, load peaking is a key link to ensure the stable operation of the power system. During the peak period of electricity consumption, the demand for electricity is large. If the power generation is insufficient, it will lead to power shortages, affecting normal production and life. Through load peaking, the power load can be adjusted, and the power demand during the peak period can be transferred to the trough period, so as to maintain the balance of power supply and demand. Load forecasting technology came into being in this scenario and plays an important role. Through load forecasting technology, the power demand of buildings in the future can be predicted, so as to formulate corresponding peaking plans in advance. This can not only improve the stability and reliability of the power system, but also reduce the economic losses caused by power shortages. In addition, load forecasting technology can also be combined with renewable energy generation to improve the utilization rate of renewable energy and the overall operation efficiency of the power system. In the virtual power plant scenario, load peaking and load forecasting technology are important means to achieve stable operation of the power system, which is of great significance to promoting sustainable development and promoting energy transformation.
[0004] At present, common load forecasting schemes include: time series method, regression analysis method and neural network method. Although these methods have certain forecasting capabilities, they also have some problems. For example, the time series method performs well when processing time series data, but when the data is abnormal or mutated, the forecast results may be inaccurate; the regression analysis method is suitable for explanatory forecasting, but when the relationship between the independent variable and the dependent variable is complex, the forecasting effect of regression analysis may not be good; although the neural network method can handle complex nonlinear relationships, it requires a large amount of training data and computing resources, and is sensitive to the selection and adjustment of parameters.
[0005] In addition, no matter what kind of load forecasting scheme is used, it needs to be supported by a large amount of real-time and accurate data, such as historical power load and historical weather data. However, the total power load data of the building is often not directly available, and needs to be calculated through the electric meter under each transformer, and the temperature data is obtained by installing a small meteorological station. The accuracy of both data cannot be effectively guaranteed, which will affect the accuracy of the forecast. Summary of the invention
[0006] The purpose of the present invention is to provide a building power load forecasting method, device, system, computer equipment, computer-readable storage medium and computer program product applied to a VPP scenario, which is used to combine the time series method and the regression analysis method, and use the most limited parameters in the VPP scenario to accurately and effectively predict the power load changes of the target building within the predicted target day, so as to provide a better power resource allocation and scheduling strategy for the power system in the load peak regulation process applied in the virtual power plant scenario.
[0007] In order to achieve the above object, the present invention adopts the following technical solutions:
[0008] In a first aspect, a building power load forecasting method applied to a VPP scenario is provided, comprising:
[0009] Obtaining historical power load data of a target building and historical weather data of the area where the target building is located, wherein the historical power load data contains the daily energy consumption of the target building on each historical day and the energy consumption of each unit time period within each day, and the historical weather data contains the daily temperature of the area where the target building is located on each historical day;
[0010] According to the historical power load data and the historical weather data, a polynomial linear regression curve function of daily energy consumption and daily temperature of each type of date among different types of dates is fitted;
[0011] Obtaining the predicted temperature of the area where the target building is located on the predicted target day;
[0012] According to the intra-day predicted temperature, the historical power load data and the historical weather data, a plurality of similar historical days belonging to a certain category of the plurality of categories of dates together with the predicted target day are selected, and the intra-day temperature of each of the similar historical days and the intra-segment energy consumption of each unit time period of each day are extracted from the plurality of similar historical days;
[0013] For each of the same historical days, the corresponding new daily energy consumption is calculated based on the corresponding daily temperature and the polynomial linear regression curve function of the certain type of date;
[0014] Calculating the intraday predicted energy consumption of the predicted target day according to the intraday predicted temperature and the polynomial linear regression curve function of the certain type of date;
[0015] According to the intraday predicted energy consumption of the predicted target day and the intraday new energy consumption of each similar historical day, the correction coefficient d is calculated according to the following formula:
[0016]
[0017] Wherein, N represents the total number of the same historical days, n represents a positive integer less than or equal to N, L tg represents the predicted energy consumption within the day of the predicted target day, L new,n represents the new intra-day energy consumption of the nth similar historical day among the multiple similar historical days;
[0018] For each unit time period within the forecast target day, the corresponding forecast energy consumption within the time period is calculated according to the following formula:
[0019]
[0020] In the formula, k is a positive integer, l tg,k represents the predicted energy consumption per unit time period on the kth day of the predicted target day, l n,k represents the energy consumption per unit time period in the kth day of the nth similar historical day;
[0021] The predicted energy consumption in each unit time period on each day of the predicted target day is output as the building power load prediction result of the target building on the predicted target day.
[0022] Based on the above invention content, a new solution for building power load forecasting combining time series method and regression analysis method is provided, that is, first, according to historical power load data and historical weather data, a polynomial linear regression curve function of the intraday energy consumption and the intraday temperature of each type of date is fitted, and then according to the intraday predicted temperature of the prediction target day, the historical power load data and the historical weather data, a plurality of similar historical days belonging to a certain type of date together with the prediction target day are selected, and the function is applied to calculate the intraday predicted energy consumption of the prediction target day and the new intraday energy consumption of each similar historical day and a correction coefficient, and finally the correction coefficient is applied to calculate the intra-segment predicted energy consumption of each unit time period on each day of the prediction target day, so that the most limited parameters can be used in the VPP scenario to accurately and effectively predict the power load changes of the target building on the prediction target day, and then in the load peak regulation process applied in the virtual power plant scenario, a better power resource allocation and scheduling strategy can be provided for the power system, which is convenient for practical application and promotion.
[0023] In a possible design, according to the historical power load data and the historical weather data, a polynomial linear regression curve function of the daily energy consumption and the daily temperature of each type of date in different types of dates is fitted, including the following steps S21 to S24:
[0024] S21. For any date in the different multiple types of dates, extract the daily energy consumption of the corresponding M recent historical dates from the historical power load data, and extract the daily temperature of the corresponding M recent historical dates from the historical weather data, and then execute step S22, where M represents a positive integer greater than or equal to 3;
[0025] S22. For any of the types of dates, according to the corresponding daily energy consumption and daily temperature of the M recent historical dates, a polynomial linear regression curve function of the corresponding daily energy consumption and daily temperature is fitted, and then step S23 is executed;
[0026] S23. For any of the types of dates, according to the corresponding daily temperatures of the M most recent historical dates, the corresponding polynomial linear regression curve function is applied to calculate the estimated daily energy consumption of the M most recent historical dates, and then step S24 is executed;
[0027] S24. For any type of date, determine whether the corresponding function accuracy evaluation index meets the preset qualification condition. If so, use the corresponding polynomial linear regression curve function as the corresponding final polynomial linear regression curve function. Otherwise, after reducing the value M, return to step S21, wherein the calculation formula of the function accuracy evaluation index is as follows:
[0028]
[0029] In the formula, R represents the accuracy evaluation index of the function, m represents a positive integer less than or equal to M, and L m represents the daily energy consumption of the mth most recent historical date among the M most recent historical dates, L em,m represents the estimated energy consumption on the mth most recent historical date, represents the average daily energy consumption of the M most recent historical dates.
[0030] In a possible design, based on the intra-day predicted temperature, the historical power load data and the historical weather data, multiple similar historical days that belong to a certain category of the multiple categories of dates together with the predicted target day are selected, including the following steps S41 to S47:
[0031] S41. Obtain the temperature similarity criterion threshold, the date similarity criterion threshold and the energy consumption average threshold, and then execute step S42;
[0032] S42. According to the intraday predicted temperature and the historical weather data, traverse each historical day that belongs to a certain category of the multiple categories of dates together with the predicted target day: if the absolute value of the difference between the intraday temperature of a certain historical day and the intraday predicted temperature is less than or equal to the temperature similarity criterion threshold, take the certain historical day as the first candidate historical day, and execute step S43 after the traversal is completed;
[0033] S43. Traverse each of the first candidate historical dates: if the number of days between the predicted target date and a first candidate historical date is less than or equal to the date proximity criterion threshold, use the first candidate historical date as the second candidate historical date, and execute step S44 after the traversal is completed;
[0034] S44. Calculate the average daily energy consumption of all the second candidate historical days according to the historical power load data, and then execute step S45;
[0035] S45. Traverse each of the second candidate historical days: if the absolute value of the difference between the intra-day energy consumption of a second candidate historical day divided by the average intra-day energy consumption and 1 is greater than the energy consumption average threshold, then eliminate the second candidate historical day, and execute step S46 after the traversal is completed;
[0036] S46. If a second candidate historical day is eliminated in the process of traversing each of the second candidate historical days, return to step S44, otherwise execute step S47;
[0037] S47. Determine whether the total number of remaining second alternative historical days is greater than or equal to the preset day threshold. If so, treat the remaining second alternative historical days as multiple similar historical days belonging to the certain type of date together with the predicted target day. Otherwise, after adjusting the temperature similarity criterion threshold, the date proximity criterion threshold and / or the energy consumption average threshold, return to execute step S41, wherein the preset day threshold is greater than or equal to 3.
[0038] In a possible design, the daily temperature includes the daily minimum temperature, the daily average temperature and / or the daily maximum temperature.
[0039] In a possible design, the multiple types of dates include non-working days, the first working day after the non-working day and / or ordinary working days.
[0040] In the second aspect, a building power load forecasting device applied to a VPP scenario is provided, including a historical data acquisition module, a curve function fitting module, a predicted temperature acquisition module, a similar day data extraction module, a daily energy consumption calculation module, a correction parameter calculation module, a segment energy consumption prediction module and a prediction result output module;
[0041] The historical data acquisition module is used to acquire the historical power load data of the target building and the historical weather data of the area where the target building is located, wherein the historical power load data contains the daily energy consumption of the target building on each historical day and the energy consumption of each unit time period within each day, and the historical weather data contains the daily temperature of the area where the target building is located on each historical day;
[0042] The curve function fitting module is communicatively connected to the historical data acquisition module, and is used to fit a polynomial linear regression curve function of daily energy consumption and daily temperature for each type of date among different types of dates according to the historical power load data and the historical weather data;
[0043] The predicted temperature acquisition module is used to obtain the predicted temperature of the area where the target building is located on the predicted target day;
[0044] The same-day data extraction module is respectively connected to the historical data acquisition module and the predicted temperature acquisition module in communication, and is used to select a plurality of similar historical days that belong to a certain category of the plurality of dates together with the predicted target day according to the predicted temperature within the day, the historical power load data and the historical weather data, and extract the intraday temperature of each of the same historical days and the intra-segment energy consumption of each unit time period within the day from the plurality of similar historical days;
[0045] The intra-day energy consumption calculation module is respectively connected to the curve function fitting module and the same-day data extraction module in communication, and is used to calculate the corresponding new intra-day energy consumption for each of the same-day historical days according to the corresponding intra-day temperature and the polynomial linear regression curve function of the certain type of date, and is also used to calculate the intra-day predicted energy consumption of the predicted target day according to the intra-day predicted temperature and the polynomial linear regression curve function of the certain type of date;
[0046] The correction parameter calculation module is communicatively connected to the intra-day energy consumption calculation module, and is used to calculate the correction coefficient d according to the intra-day predicted energy consumption of the predicted target day and the intra-day new energy consumption of each similar historical day according to the following formula:
[0047]
[0048] Wherein, N represents the total number of the same historical days, n represents a positive integer less than or equal to N, L tg represents the predicted energy consumption within the day of the predicted target day, L new,n represents the new intra-day energy consumption of the nth similar historical day among the multiple similar historical days;
[0049] The intra-segment energy consumption prediction module is respectively connected to the historical data acquisition module and the correction parameter calculation module for calculating the corresponding intra-segment predicted energy consumption according to the following formula for each intra-day unit time period of the predicted target day:
[0050]
[0051] In the formula, k is a positive integer, l tg,k represents the predicted energy consumption per unit time period on the kth day of the predicted target day, l n,k represents the energy consumption per unit time period in the kth day of the nth similar historical day;
[0052] The prediction result output module is communicatively connected to the intra-segment energy consumption prediction module, and is used to output the intra-segment predicted energy consumption of the unit time period in each day of the prediction target day as the building power load prediction result of the target building on the prediction target day.
[0053] In a third aspect, the present invention provides a building power load forecasting system applied to a VPP scenario, comprising an infrared meter reading device, a weather data interface and a data processing device, wherein the data processing device is communicatively connected to the infrared meter reading device and the weather data interface respectively;
[0054] The infrared meter reading device is used to achieve close-range communication with the infrared optical port of the gateway electric energy meter of the target building through an infrared probe, so as to regularly read the energy consumption record data of the gateway electric energy meter, and upload the energy consumption record data to the data processing device, so that the data processing device can obtain the historical power load data of the target building according to the energy consumption record data, wherein the historical power load data contains the daily energy consumption of the target building on each historical day and the intra-segment energy consumption of each unit time period within each day;
[0055] The weather data interface is used to obtain the historical weather data of the area where the target building is located and the predicted daily temperature on the predicted target day by connecting with the API interface of a third-party weather data service provider, and upload the historical weather data and the predicted daily temperature to the data processing device, wherein the historical weather data contains the daily temperature of the area where the target building is located on each day of the history;
[0056] The data processing device is used to execute the building power load forecasting method as described in the first aspect or any possible design of the first aspect.
[0057] In a fourth aspect, the present invention provides a computer device comprising a memory, a processor and a transceiver which are communicatively connected in sequence, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program to execute the building power load forecasting method as described in the first aspect or any possible design of the first aspect.
[0058] In a fifth aspect, the present invention provides a computer-readable storage medium having instructions stored thereon, which, when executed on a computer, executes the building power load forecasting method as described in the first aspect or any possible design of the first aspect.
[0059] In a sixth aspect, the present invention provides a computer program product, comprising a computer program or instructions, which, when executed by a computer, implements the building power load forecasting method as described in the first aspect or any possible design of the first aspect.
[0060] Beneficial effects of the above scheme:
[0061] (1) The present invention creatively provides a new scheme for building power load forecasting by combining the time series method and the regression analysis method, that is, firstly, according to the historical power load data and the historical weather data, a polynomial linear regression curve function of the daily energy consumption and the daily temperature of each type of date is fitted, and then according to the daily predicted temperature of the prediction target day, the historical power load data and the historical weather data, a plurality of similar historical days belonging to a certain type of date together with the prediction target day are selected, and the function is applied to calculate the daily predicted energy consumption of the prediction target day and the new daily energy consumption of each similar historical day and the correction coefficient, and finally the correction coefficient is applied to calculate the intra-segment predicted energy consumption of each unit time period of each day of the prediction target day, so that the most limited parameters can be used in the VPP scenario to accurately and effectively predict the power load changes of the target building within the prediction target day, and then in the load peak regulation process applied in the virtual power plant scenario, a better power resource allocation and scheduling strategy can be provided for the power system;
[0062] (2) It can also more accurately predict the changes in the building's power load and provide accurate data support for the load peak regulation of the virtual power plant. It also has the advantages of being easy to implement, having high computational efficiency and requiring low computing power, which facilitates practical application and promotion. BRIEF DESCRIPTION OF THE DRAWINGS
[0063] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0064] Figure 1 A flow chart of a building power load forecasting method applied to a VPP scenario provided in an embodiment of the present application.
[0065] Figure 2 A schematic diagram of the structure of a building power load prediction device applied to a VPP scenario provided in an embodiment of the present application.
[0066] Figure 3 A schematic diagram of the structure of a building power load forecasting system applied to a VPP scenario provided in an embodiment of the present application.
[0067] Figure 4 A schematic diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0068] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in combination with the drawings and the description of the embodiments or the prior art. Obviously, the following description of the structure of the drawings is only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.
[0069] It should be understood that although the terms first and second, etc. may be used herein to describe various objects, these objects should not be limited by these terms. These terms are only used to distinguish one object from another object. For example, a first object can be referred to as a second object, and similarly, a second object can be referred to as a first object without departing from the scope of the exemplary embodiments of the present invention.
[0070] It should be understood that the term "and / or" that may appear in this document is merely a description of the association relationship between associated objects, indicating that there may be three relationships. For example, A and / or B can indicate three situations: A exists alone, B exists alone, or A and B exist at the same time. For another example, A, B and / or C can indicate the existence of any one of A, B and C or any combination of them. The term " / and" that may appear in this document describes another type of association object relationship, indicating that there may be two relationships. For example, A / and B can indicate two situations: A exists alone or A and B exist at the same time. In addition, the character " / " that may appear in this document generally indicates that the previous and next associated objects are in an "or" relationship.
[0071] Example:
[0072] like Figure 1 As shown, the building power load forecasting method provided in the first aspect of this embodiment and applied to the VPP scenario can be, but is not limited to, executed by a computer device with certain computing resources, such as a cloud platform server or other computer device. Figure 1 As shown, the building power load prediction method may include but is not limited to the following steps S1 to S9.
[0073] S1. Obtain the historical power load data of the target building and the historical weather data of the area where the target building is located, wherein the historical power load data includes but is not limited to the daily energy consumption of the target building on each historical day and the energy consumption of each unit time period within each day, and the historical weather data includes but is not limited to the daily temperature of the area where the target building is located on each historical day.
[0074] In step S1, the target building is the object to be predicted, which may be, but is not limited to, an office building, a shopping mall, etc. The intraday unit time period may be, but is not limited to, an intraday hour period, such as an hour period from 0:00 to 1:00, an hour period from 1:00 to 2:00, ..., an hour period from 23:00 to 24:00, etc. Specifically, the intraday temperature includes, but is not limited to, the intraday minimum temperature, the intraday average temperature, and / or the intraday maximum temperature, etc. In addition, each historical day needs to be a historical day that has not undergone peak shaving regulation.
[0075] S2. According to the historical power load data and the historical weather data, a polynomial linear regression curve function of the daily energy consumption and the daily temperature of each type of date in different types of dates is fitted.
[0076] In step S2, specifically, the multiple types of dates include but are not limited to non-working days (such as Saturday and Sunday), the first working day after a non-working day (such as Monday), and / or ordinary working days (such as Tuesday, Wednesday, Thursday and Friday). In the polynomial linear regression curve function, the intraday temperature is the independent variable, and the intraday energy consumption is the dependent variable; for example, the polynomial linear regression curve function can be but is not limited to the following formula:
[0077] L=f(T)=a0+a1×T+a2×T 2 +a3×T 3
[0078] Wherein, T represents the daily temperature as the independent variable, L represents the daily energy consumption as the dependent variable, f() represents the polynomial linear regression curve function, and a0, a1, a2 and a3 represent the proposed coefficients respectively.
[0079] In the step S2, in order to ensure the accuracy of the fitting curve of the polynomial linear regression curve function of the various types of dates, preferably, based on the historical power load data and the historical weather data, a polynomial linear regression curve function of the daily energy consumption and the daily temperature of each type of date in different multiple types of dates is fitted, including but not limited to the following steps S21 to S24.
[0080] S21. For any date among the different multiple types of dates, extract the daily energy consumption of the corresponding M most recent historical dates from the historical power load data, and extract the daily temperature of the corresponding M most recent historical dates from the historical weather data, and then execute step S22, where M represents a positive integer greater than or equal to 3.
[0081] In step S21, for example, if any of the dates is a Sunday, it is necessary to extract the daily energy consumption and daily temperature of M recent historical Sundays. In addition, the initial value of M can be 30 or 10, for example.
[0082] S22. For any type of date, according to the corresponding daily energy consumption and daily temperature of the M recent historical dates, a polynomial linear regression curve function of the corresponding daily energy consumption and daily temperature is fitted, and then step S23 is executed.
[0083] In step S22, the specific fitting process is the process of determining the specific values of a0, a1, a2 and a3, which can be achieved by using existing linear regression methods.
[0084] S23. For any of the types of dates, according to the corresponding daily temperatures of the M most recent historical dates, the corresponding polynomial linear regression curve function is applied to calculate the estimated daily energy consumption of the M most recent historical dates, and then step S24 is executed.
[0085] In step S23, the specific application process is to sequentially substitute the daily temperatures of the M most recent historical dates into the polynomial linear regression curve function, and then sequentially obtain the estimated daily energy consumption of the M most recent historical dates.
[0086] S24. For any type of date, determine whether the corresponding function accuracy evaluation index meets the preset qualification condition. If so, use the corresponding polynomial linear regression curve function as the corresponding final polynomial linear regression curve function. Otherwise, after reducing the value M, return to step S21, wherein the calculation formula of the function accuracy evaluation index is as follows:
[0087]
[0088] In the formula, R represents the accuracy evaluation index of the function, m represents a positive integer less than or equal to M, and L m represents the daily energy consumption of the mth most recent historical date among the M most recent historical dates, L em,m represents the estimated energy consumption on the mth most recent historical date, represents the average daily energy consumption of the M most recent historical dates.
[0089] S3. Obtain the predicted temperature of the area where the target building is located on the predicted target day.
[0090] S4. Based on the predicted temperature within the day, the historical power load data and the historical weather data, select a plurality of similar historical days that belong to a certain category of dates among the multiple categories of dates together with the predicted target day, and extract the daily temperature of each similar historical day among the multiple similar historical days and the energy consumption within a unit time period of each day.
[0091] In step S4, for example, if the forecast target day is July 19, 2022 (Tuesday), which is an ordinary working day, then the multiple similar historical days must also be ordinary working days. The intraday temperature of each similar historical day and the intra-segment energy consumption of each unit time period within each day are used as reference data for subsequent predictions. In order to ensure the availability of these reference data, preferably, based on the intraday forecast temperature, the historical power load data and the historical weather data, multiple similar historical days that belong to a certain category of the multiple categories of dates together with the forecast target day are selected, including but not limited to the following steps S41 to S47.
[0092] S41. Obtain the temperature similarity criterion threshold, date similarity criterion threshold and energy consumption average threshold, and then execute step S42.
[0093] In the step S41, the unit of the temperature similarity criterion threshold is Celsius, the unit of the date similarity criterion threshold is day, and the unit of the energy consumption average threshold is %.
[0094] S42. Based on the predicted temperature for the day and the historical weather data, traverse each historical day that belongs to a category of dates among the multiple categories of dates together with the predicted target day: if the absolute value of the difference between the daily temperature of a historical day and the predicted temperature for the day is less than or equal to the temperature similarity criterion threshold, then the historical day is used as the first alternative historical day, and step S43 is executed after the traversal is completed.
[0095] S43. Traverse each of the first candidate historical days: if the number of days between the predicted target day and a certain first candidate historical day is less than or equal to the date proximity criterion threshold, then the certain first candidate historical day is used as the second candidate historical day, and step S44 is executed after the traversal is completed.
[0096] S44. Calculate the average daily energy consumption of all the second candidate historical days based on the historical power load data, and then execute step S45.
[0097] S45. Traverse each of the second alternative historical days: if the absolute value of the difference between the daily energy consumption of a second alternative historical day divided by the average daily energy consumption and 1 is greater than the energy consumption average threshold, then eliminate the second alternative historical day and execute step S46 after the traversal is completed.
[0098] In step S45, if the absolute value of the difference between the result of dividing the intraday energy consumption of a second candidate historical day by the average intraday energy consumption and 1 is greater than the energy consumption average threshold, it indicates that the second candidate historical day is an abnormal day and needs to be eliminated.
[0099] S46. If a second candidate historical day is eliminated in the process of traversing each of the second candidate historical days, return to step S44, otherwise execute step S47.
[0100] S47. Determine whether the total number of remaining second alternative historical days is greater than or equal to the preset day threshold. If so, treat the remaining second alternative historical days as multiple similar historical days belonging to the certain type of date together with the predicted target day. Otherwise, after adjusting the temperature similarity criterion threshold, the date proximity criterion threshold and / or the energy consumption average threshold, return to execute step S41, wherein the preset day threshold is greater than or equal to 3.
[0101] In step S47, the specific adjustment method of the temperature similarity criterion threshold, the date similarity criterion threshold and / or the energy consumption average threshold can be a manual adjustment method or an automatic adjustment method based on a certain existing optimization algorithm (such as the gray wolf optimization algorithm). In addition, the preset day threshold can be 8 for example.
[0102] S5. For each of the same historical days, the corresponding new daily energy consumption is calculated based on the corresponding daily temperature and the polynomial linear regression curve function of the certain type of date.
[0103] S6. Calculate the predicted daily energy consumption of the predicted target day based on the predicted daily temperature and the polynomial linear regression curve function of the certain type of date.
[0104] In the steps S5 to S6, the specific calculation process of the new energy consumption within the day and the predicted energy consumption within the day can be referred to the aforementioned step S23, which will not be repeated here.
[0105] S7. Based on the predicted energy consumption within the day of the predicted target day and the new energy consumption within the day of each similar historical day, the correction coefficient d is calculated according to the following formula:
[0106]
[0107] Wherein, N represents the total number of the same historical days, n represents a positive integer less than or equal to N, L tg represents the predicted energy consumption within the day of the predicted target day, L new,n It represents the new intra-day energy consumption of the nth similar historical day among the multiple similar historical days.
[0108] S8. For each unit time period within the forecast target day, the corresponding forecast energy consumption within the time period is calculated according to the following formula:
[0109]
[0110] In the formula, k is a positive integer, l tg,k represents the predicted energy consumption per unit time period on the kth day of the predicted target day, l n,k It represents the energy consumption within a unit time period within the kth day of the nth similar historical day.
[0111] S9. The predicted energy consumption in each unit time period of each day on the predicted target day is output as the building power load prediction result of the target building on the predicted target day.
[0112] In step S9, the predicted energy consumption within each unit time period on each day of the predicted target day can be output to the load peak-shaving strategy formulation server, so that the load peak-shaving strategy formulation server can formulate and execute the load peak-shaving plan on the predicted target day based on the predicted energy consumption within each unit time period on the predicted target day.
[0113] Therefore, based on the building power load forecasting method described in the aforementioned steps S1 to S9, a new scheme for building power load forecasting combining the time series method and the regression analysis method is provided, that is, first, according to the historical power load data and the historical weather data, a polynomial linear regression curve function of the intraday energy consumption and the intraday temperature of each type of date is fitted, and then according to the intraday predicted temperature of the prediction target day, the historical power load data and the historical weather data, a plurality of similar historical days belonging to a certain type of date together with the prediction target day are selected, and the function is applied to calculate the intraday predicted energy consumption of the prediction target day and the new intraday energy consumption of each similar historical day and the correction coefficient, and finally the correction coefficient is applied to calculate the intra-segment predicted energy consumption of each unit time period on each day of the prediction target day, so that the most limited parameters can be used in the VPP scenario to accurately and effectively predict the power load changes of the target building on the prediction target day, and then in the load peak regulation process applied in the virtual power plant scenario, a better power resource allocation and scheduling strategy can be provided for the power system, which is convenient for practical application and promotion.
[0114] like Figure 2 As shown, the second aspect of this embodiment provides a virtual device for implementing the building power load prediction method described in the first aspect, including a historical data acquisition module, a curve function fitting module, a predicted temperature acquisition module, a similar day data extraction module, a daily energy consumption calculation module, a correction parameter calculation module, a segment energy consumption prediction module and a prediction result output module;
[0115] The historical data acquisition module is used to acquire the historical power load data of the target building and the historical weather data of the area where the target building is located, wherein the historical power load data contains the daily energy consumption of the target building on each historical day and the energy consumption of each unit time period within each day, and the historical weather data contains the daily temperature of the area where the target building is located on each historical day;
[0116] The curve function fitting module is communicatively connected to the historical data acquisition module, and is used to fit a polynomial linear regression curve function of daily energy consumption and daily temperature for each type of date among different types of dates according to the historical power load data and the historical weather data;
[0117] The predicted temperature acquisition module is used to obtain the predicted temperature of the area where the target building is located on the predicted target day;
[0118] The same-day data extraction module is respectively connected to the historical data acquisition module and the predicted temperature acquisition module in communication, and is used to select a plurality of similar historical days that belong to a certain category of the plurality of dates together with the predicted target day according to the predicted temperature within the day, the historical power load data and the historical weather data, and extract the intraday temperature of each of the same historical days and the intra-segment energy consumption of each unit time period within the day from the plurality of similar historical days;
[0119] The intra-day energy consumption calculation module is respectively connected to the curve function fitting module and the same-day data extraction module in communication, and is used to calculate the corresponding new intra-day energy consumption for each of the same-day historical days according to the corresponding intra-day temperature and the polynomial linear regression curve function of the certain type of date, and is also used to calculate the intra-day predicted energy consumption of the predicted target day according to the intra-day predicted temperature and the polynomial linear regression curve function of the certain type of date;
[0120] The correction parameter calculation module is communicatively connected to the intra-day energy consumption calculation module, and is used to calculate the correction coefficient d according to the intra-day predicted energy consumption of the predicted target day and the intra-day new energy consumption of each similar historical day according to the following formula:
[0121]
[0122] Wherein, N represents the total number of the same historical days, n represents a positive integer less than or equal to N, L tg represents the predicted energy consumption within the day of the predicted target day, L new,n represents the new intra-day energy consumption of the nth similar historical day among the multiple similar historical days;
[0123] The intra-segment energy consumption prediction module is respectively connected to the historical data acquisition module and the correction parameter calculation module for calculating the corresponding intra-segment predicted energy consumption according to the following formula for each intra-day unit time period of the predicted target day:
[0124]
[0125] In the formula, k is a positive integer, l tg,k represents the predicted energy consumption per unit time period on the kth day of the predicted target day, l n,k represents the energy consumption per unit time period in the kth day of the nth similar historical day;
[0126] The prediction result output module is communicatively connected to the intra-segment energy consumption prediction module, and is used to output the intra-segment predicted energy consumption of the unit time period in each day of the prediction target day as the building power load prediction result of the target building on the prediction target day.
[0127] The working process, working details and technical effects of the aforementioned device provided in the second aspect of this embodiment can be referred to the building power load prediction method described in the first aspect, and will not be repeated here.
[0128] like Figure 3 As shown, the third aspect of this embodiment provides a physical system for implementing the building power load prediction method described in the first aspect, including an infrared meter reading device, a weather data interface and a data processing device, wherein the data processing device is respectively communicatively connected to the infrared meter reading device and the weather data interface;
[0129] The infrared meter reading device is used to achieve close-range communication with the infrared optical port of the gateway electric energy meter of the target building through an infrared probe, so as to regularly read the energy consumption record data of the gateway electric energy meter, and upload the energy consumption record data to the data processing device, so that the data processing device can obtain the historical power load data of the target building according to the energy consumption record data, wherein the historical power load data contains the daily energy consumption of the target building on each historical day and the intra-segment energy consumption of each unit time period within each day;
[0130] The weather data interface is used to obtain the historical weather data of the area where the target building is located and the predicted daily temperature on the predicted target day by connecting with the API (Application Programming Interface) interface of a third-party weather data service provider, and upload the historical weather data and the predicted daily temperature to the data processing device, wherein the historical weather data contains the daily temperature of the area where the target building is located on each day of the history;
[0131] The data processing device is used to execute the building power load forecasting method as described in the first aspect.
[0132] The working process, working details and technical effects of the aforementioned system provided in the third aspect of this embodiment can be referred to the building power load forecasting method described in the first aspect, and will not be repeated here.
[0133] like Figure 4As shown, the fourth aspect of this embodiment provides a computer device for executing the building power load prediction method as described in the first aspect, including a memory, a processor and a transceiver that are sequentially connected in communication, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program to execute the building power load prediction method as described in the first aspect. For example, the memory may include, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a flash memory, a first-in-first-out memory (FIFO) and / or a first-in-last-out memory (FILO), etc.; the processor may include, but is not limited to, a microprocessor of the STM32F105 series. In addition, the computer device may also include, but is not limited to, a power module, a display screen and other necessary components.
[0134] The working process, working details and technical effects of the aforementioned computer equipment provided in the fourth aspect of this embodiment can be found in the building power load forecasting method described in the first aspect, and will not be repeated here.
[0135] In a fifth aspect of this embodiment, there is provided a computer-readable storage medium storing instructions including the building power load forecasting method as described in the first aspect, that is, the computer-readable storage medium stores instructions, and when the instructions are executed on a computer, the building power load forecasting method as described in the first aspect is executed. The computer-readable storage medium refers to a carrier for storing data, which may include but is not limited to computer-readable storage media such as floppy disks, optical disks, hard disks, flash memories, USB flash drives, and / or memory sticks, and the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.
[0136] The working process, working details and technical effects of the aforementioned computer-readable storage medium provided in the fifth aspect of this embodiment can be referred to the building power load forecasting method described in the first aspect, and will not be repeated here.
[0137] In a sixth aspect of this embodiment, a computer program product is provided, including a computer program or an instruction, wherein when the computer program or the instruction is executed by a computer, the building power load forecasting method as described in the first aspect is implemented. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device.
[0138] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A building power load forecasting method applied to a VPP scenario, characterized in that: include: Obtaining historical power load data of a target building and historical weather data of the area where the target building is located, wherein the historical power load data contains the daily energy consumption of the target building on each historical day and the energy consumption of each unit time period within each day, and the historical weather data contains the daily temperature of the area where the target building is located on each historical day; According to the historical power load data and the historical weather data, a polynomial linear regression curve function of daily energy consumption and daily temperature of each type of date among different types of dates is fitted; Obtaining the predicted temperature of the area where the target building is located on the predicted target day; According to the intra-day predicted temperature, the historical power load data and the historical weather data, a plurality of similar historical days belonging to a certain category of the plurality of categories of dates together with the predicted target day are selected, and the intra-day temperature of each of the similar historical days and the intra-segment energy consumption of each unit time period of each day are extracted from the plurality of similar historical days; For each of the same historical days, the corresponding daily new energy consumption is calculated based on the corresponding daily temperature and the polynomial linear regression curve function of the certain type of date; Calculating the intraday predicted energy consumption of the predicted target day according to the intraday predicted temperature and the polynomial linear regression curve function of the certain type of date; According to the intraday predicted energy consumption of the predicted target day and the intraday new energy consumption of each similar historical day, the correction coefficient d is calculated according to the following formula: Wherein, N represents the total number of the same historical days, n represents a positive integer less than or equal to N, L tg represents the predicted energy consumption within the day of the predicted target day, L new,n represents the new intra-day energy consumption of the nth similar historical day among the multiple similar historical days; For each unit time period within the forecast target day, the corresponding forecast energy consumption within the time period is calculated according to the following formula: In the formula, k is a positive integer, l tg,k represents the predicted energy consumption per unit time period on the kth day of the predicted target day, l n,k represents the energy consumption per unit time period in the kth day of the nth similar historical day; The predicted energy consumption in each unit time period on each day of the predicted target day is output as the building power load prediction result of the target building on the predicted target day.
2. The building power load prediction method according to claim 1, characterized in that: According to the historical power load data and the historical weather data, a polynomial linear regression curve function of daily energy consumption and daily temperature of each type of date in different types of dates is fitted, including the following steps S21-S24: S21. For any date in the different multiple types of dates, extract the daily energy consumption of the corresponding M recent historical dates from the historical power load data, and extract the daily temperature of the corresponding M recent historical dates from the historical weather data, and then execute step S22, where M represents a positive integer greater than or equal to 3; S22. For any of the types of dates, according to the corresponding daily energy consumption and daily temperature of the M recent historical dates, a polynomial linear regression curve function of the corresponding daily energy consumption and daily temperature is fitted, and then step S23 is executed; S23. For any of the types of dates, according to the corresponding daily temperatures of the M most recent historical dates, the corresponding polynomial linear regression curve function is applied to calculate the estimated daily energy consumption of the M most recent historical dates, and then step S24 is executed; S24. For any type of date, determine whether the corresponding function accuracy evaluation index meets the preset qualification condition. If so, use the corresponding polynomial linear regression curve function as the corresponding final polynomial linear regression curve function. Otherwise, after reducing the value M, return to step S21, wherein the calculation formula of the function accuracy evaluation index is as follows: In the formula, R represents the accuracy evaluation index of the function, m represents a positive integer less than or equal to M, and L m represents the daily energy consumption of the mth most recent historical date among the M most recent historical dates, L em,m represents the estimated energy consumption on the mth most recent historical date, represents the average daily energy consumption of the M most recent historical dates.
3. The building power load prediction method according to claim 1, characterized in that: According to the intra-day predicted temperature, the historical power load data and the historical weather data, a plurality of similar historical days belonging to a certain category of the plurality of categories of dates together with the predicted target day are selected, including the following steps S41 to S47: S41. Obtain the temperature similarity criterion threshold, the date similarity criterion threshold and the energy consumption average threshold, and then execute step S42; S42. According to the intraday predicted temperature and the historical weather data, traverse each historical day that belongs to a certain category of the multiple categories of dates together with the predicted target day: if the absolute value of the difference between the intraday temperature of a certain historical day and the intraday predicted temperature is less than or equal to the temperature similarity criterion threshold, take the certain historical day as the first candidate historical day, and execute step S43 after the traversal is completed; S43. Traverse each of the first candidate historical dates: if the number of days between the predicted target date and a first candidate historical date is less than or equal to the date proximity criterion threshold, use the first candidate historical date as the second candidate historical date, and execute step S44 after the traversal is completed; S44. Calculate the average daily energy consumption of all the second candidate historical days according to the historical power load data, and then execute step S45; S45. Traverse each of the second candidate historical days: if the absolute value of the difference between the intra-day energy consumption of a second candidate historical day divided by the average intra-day energy consumption and 1 is greater than the energy consumption average threshold, then eliminate the second candidate historical day, and execute step S46 after the traversal is completed; S46. If a second candidate historical day is eliminated in the process of traversing each of the second candidate historical days, return to step S44, otherwise execute step S47; S47. Determine whether the total number of remaining second alternative historical days is greater than or equal to the preset day threshold. If so, treat the remaining second alternative historical days as multiple similar historical days belonging to the certain type of date together with the predicted target day. Otherwise, after adjusting the temperature similarity criterion threshold, the date proximity criterion threshold and / or the energy consumption average threshold, return to execute step S41, wherein the preset day threshold is greater than or equal to 3.
4. The building power load prediction method according to claim 1, characterized in that: The daily temperature includes the daily minimum temperature, the daily average temperature and / or the daily maximum temperature.
5. The building power load prediction method according to claim 1, characterized in that: The multiple types of dates include non-working days, the first working day after the non-working day and / or ordinary working days.
6. A building power load forecasting device applied to a VPP scenario, characterized in that: It includes a historical data acquisition module, a curve function fitting module, a predicted temperature acquisition module, a similar day data extraction module, a daily energy consumption calculation module, a correction parameter calculation module, a segment energy consumption prediction module and a prediction result output module; The historical data acquisition module is used to acquire the historical power load data of the target building and the historical weather data of the area where the target building is located, wherein the historical power load data contains the daily energy consumption of the target building on each historical day and the energy consumption of each unit time period within each day, and the historical weather data contains the daily temperature of the area where the target building is located on each historical day; The curve function fitting module is communicatively connected to the historical data acquisition module, and is used to fit a polynomial linear regression curve function of daily energy consumption and daily temperature for each type of date among different types of dates according to the historical power load data and the historical weather data; The predicted temperature acquisition module is used to obtain the predicted temperature of the area where the target building is located on the predicted target day; The same-day data extraction module is respectively connected to the historical data acquisition module and the predicted temperature acquisition module in communication, and is used to select a plurality of similar historical days that belong to a certain category of the plurality of dates together with the predicted target day according to the predicted temperature within the day, the historical power load data and the historical weather data, and extract the intraday temperature of each of the same historical days and the intra-segment energy consumption of each unit time period within the day from the plurality of similar historical days; The intra-day energy consumption calculation module is respectively connected to the curve function fitting module and the same-day data extraction module in communication, and is used to calculate the corresponding new intra-day energy consumption for each of the same-day historical days according to the corresponding intra-day temperature and the polynomial linear regression curve function of the certain type of date, and is also used to calculate the intra-day predicted energy consumption of the predicted target day according to the intra-day predicted temperature and the polynomial linear regression curve function of the certain type of date; The correction parameter calculation module is communicatively connected to the intra-day energy consumption calculation module, and is used to calculate the correction coefficient d according to the intra-day predicted energy consumption of the predicted target day and the intra-day new energy consumption of each similar historical day according to the following formula: Wherein, N represents the total number of the same historical days, n represents a positive integer less than or equal to N, L tg represents the predicted energy consumption within the day of the predicted target day, L new,n represents the new intra-day energy consumption of the nth similar historical day among the multiple similar historical days; The intra-segment energy consumption prediction module is respectively connected to the historical data acquisition module and the correction parameter calculation module for calculating the corresponding intra-segment predicted energy consumption according to the following formula for each intra-day unit time period of the predicted target day: In the formula, k is a positive integer, l tg,k represents the predicted energy consumption per unit time period on the kth day of the predicted target day, l n,k represents the energy consumption per unit time period in the kth day of the nth similar historical day; The prediction result output module is communicatively connected to the intra-segment energy consumption prediction module, and is used to output the intra-segment predicted energy consumption of the unit time period in each day of the prediction target day as the building power load prediction result of the target building on the prediction target day.
7. A building power load forecasting system applied to VPP scenarios, characterized in that: It includes an infrared meter reading device, a weather data interface and a data processing device, wherein the data processing device is respectively connected to the infrared meter reading device and the weather data interface for communication; The infrared meter reading device is used to achieve close-range communication with the infrared optical port of the gateway electric energy meter of the target building through an infrared probe, so as to regularly read the energy consumption record data of the gateway electric energy meter, and upload the energy consumption record data to the data processing device, so that the data processing device can obtain the historical power load data of the target building according to the energy consumption record data, wherein the historical power load data contains the daily energy consumption of the target building on each historical day and the intra-segment energy consumption of each unit time period within each day; The weather data interface is used to obtain the historical weather data of the area where the target building is located and the predicted daily temperature on the predicted target day by connecting with the API interface of a third-party weather data service provider, and upload the historical weather data and the predicted daily temperature to the data processing device, wherein the historical weather data contains the daily temperature of the area where the target building is located on each day of the history; The data processing device is used to execute the building power load forecasting method as described in any one of claims 1 to 5.
8. A computer device, characterized in that: It includes a memory, a processor and a transceiver which are communicatively connected in sequence, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program to execute the building power load forecasting method as described in any one of claims 1 to 5.
9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores instructions, and when the instructions are executed on the computer, the building power load forecasting method as described in any one of claims 1 to 5 is executed.
10. A computer program product comprising a computer program or instructions, characterized in that When the computer program or the instruction is executed by a computer, the building power load forecasting method according to any one of claims 1 to 5 is implemented.
