A virtual power plant baseline calculation method introducing user confidence

Abstract

The present application relates to a kind of virtual power plant baseline calculation method of introducing user confidence, comprising the following steps: based on user load prediction curve, the historical predicted load value before event release day and the user load prediction value of event release day are obtained by analysis;User confidence is calculated using the historical predicted load value and its corresponding actual load value;Filter out several typical days before demand response day;According to the actual load value of the typical day, the historical baseline load value is calculated;Based on the user confidence, the historical baseline load value and the user load prediction value, the virtual power plant load baseline is calculated.The present application can optimize the load baseline of virtual power plant, reduce the influence of load prediction deviation on baseline, better encourage user to adjust energy behavior.

Description

Technical Field

[0001] This invention relates to the field of virtual power plant baseline calculation technology, and in particular to a virtual power plant baseline calculation method that incorporates user confidence. Background Technology

[0002] The baseline is a crucial basis for evaluating the regulatory market. It typically refers to the expected electricity consumption pattern or load level of power users before load management measures are implemented and the ancillary market is activated. Baseline load measures the change in a user's electricity load curve after the implementation of demand response and other measures. It serves as a benchmark for calculating the regulation volume of new operating entities such as virtual power plants participating in peak-shaving ancillary services and demand response transactions, and is a vital basis for calculating their compensation costs. Current technologies often use the average load over a specified number of days before the transaction to serve as the baseline. However, due to the influence of weather, production plans, and other factors on user energy consumption behavior, there is a significant deviation between the user's actual operating plan and the calculated baseline. This can lead to situations where responses are not calculated or where revenue is obtained even without actual responses. Furthermore, if there are many demand responses, the time period corresponding to the baseline calculation data deviates significantly from the current time, and the more responses there are, the more equal the baseline becomes. There is also the possibility of users maliciously declaring demand to obtain revenue. Therefore, a more scientific consideration of the relationship between historical data, user forecast curves, and user credibility is needed to optimize the baseline calculation method.

[0003] Document CN119518740A discloses a load baseline forecasting method, which mainly considers calculating the baseline based on historical data. However, as the number of demand responses increases, this method has the contradiction of long data push-forward time and large deviation from user operation plans. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a virtual power plant baseline calculation method that incorporates user confidence, which can more scientifically consider the relationship between historical data, user prediction curves and user confidence, and optimize the baseline calculation method.

[0005] The technical solution adopted by this invention to solve its technical problem is: to provide a virtual power plant baseline calculation method that incorporates user confidence, comprising the following steps:

[0006] Based on the user load forecast curve, the historical forecast load value before the event release date and the user load forecast value on the event release date are analyzed.

[0007] The user confidence level is calculated using the historical predicted load values ​​and their corresponding actual load values.

[0008] Select several typical days prior to the demand response date;

[0009] The historical baseline load value is calculated based on the actual load value of the typical day;

[0010] The virtual power plant load baseline is calculated based on the user confidence level, the historical baseline load value, and the user load forecast value.

[0011] Furthermore, the user confidence level is calculated based on data from several normal days prior to the event publication date, excluding demand response days, emergency peak avoidance days, and power outage maintenance days.

[0012] Furthermore, the historical baseline load value is a weighted average of the actual load values ​​for each of the typical days.

[0013] Furthermore, the virtual power plant load baseline is represented as follows:

[0014] ,

[0015] in, Indicates the first The virtual power plant load baseline mentioned on the day, This indicates the user's confidence level. Indicates the first The historical predicted load value for the day, Indicates the first The historical baseline load value for that day.

[0016] Furthermore, it also includes the step of introducing a first adjustment factor to correct the virtual power plant load baseline, wherein the first adjustment factor is the ratio of the average load of a first set number of time periods before the event to the historical average load, wherein the historical average load is the average of the total load of all the typical days of the same time period relative to the total number of time periods.

[0017] Furthermore, the virtual power plant load baseline after being corrected by the first adjustment factor is the product of the virtual power plant load baseline and the first adjustment factor.

[0018] Furthermore, it also includes the step of introducing a second adjustment factor to correct the virtual power plant load baseline, wherein the second adjustment factor is a set percentage value less than 1.

[0019] Furthermore, the revised virtual power plant load baseline is represented as follows:

[0020] ,

[0021] in, Indicates the first The virtual power plant load baseline mentioned on the day, This indicates the correction after introducing the second adjustment factor. The virtual power plant load baseline mentioned on the day, This represents the second adjustment factor. Indicates the first Actual load value at the time of the incident.

[0022] Furthermore, the second adjustment factor is set to 90%.

[0023] Furthermore, the selection of multiple typical days prior to the demand response date includes:

[0024] If the demand response date is a working day, then a first set number of first typical days are selected before the demand response date, and the first typical days satisfy the first set condition.

[0025] If the demand response date is a non-working day, then a second set number of second typical days prior to the demand response date are selected, and the second typical days satisfy the second set conditions.

[0026] Furthermore, the first set condition excludes non-working days, maintenance days, demand response days, and ancillary service days, and the daily load fluctuation rate does not exceed the threshold.

[0027] Furthermore, the second setting condition excludes working days, maintenance days, demand response days, and ancillary service days, and the daily load fluctuation rate does not exceed the threshold.

[0028] Beneficial effects

[0029] By adopting the above-mentioned technical solution, this invention has the following advantages and positive effects compared with the prior art: This invention can better optimize the load baseline and reduce the impact of load forecast deviation by combining and correcting the predicted load reported by virtual power plants and other entities based on the accuracy of users' historical load forecasts; furthermore, by introducing a first correction factor to further correct the optimized load baseline, the impact of weather changes can be reduced; by introducing a second correction factor for users with strong load fluctuations, post-correction is carried out according to the load fluctuation correction results, which on the one hand avoids the situation where users adjust their plans and do not receive subsidies, and on the other hand, it also avoids the adverse effects of users' speculative declarations. Attached Figure Description

[0030] Figure 1 This is a flowchart of an embodiment of the present invention;

[0031] Figure 2 This is a flowchart of a preferred embodiment of the present invention. Detailed Implementation

[0032] The present invention will be further illustrated below with reference to specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. Furthermore, it should be understood that after reading the teachings of this invention, those skilled in the art can make various alterations or modifications to the invention, and these equivalent forms also fall within the scope defined by the appended claims.

[0033] The technical terms used in this embodiment include:

[0034] Embodiments of the present invention relate to a method for calculating the baseline of a virtual power plant by incorporating user confidence, such as... Figure 1 As shown, it includes the following steps:

[0035] S0 obtains the user load forecast curve;

[0036] S1 calculates the historical load forecast accuracy as the user confidence level based on the historical predicted load value and its corresponding actual load value before the event release date.

[0037] S2 filters out several typical days prior to the demand response date;

[0038] S3 calculates the historical baseline load value based on the actual load value of a typical day;

[0039] S4 calculates the virtual power plant load baseline based on historical load forecast accuracy, historical baseline load value, and user load forecast value.

[0040] In some implementations, the load forecasting results involved in step S0 can be determined autonomously based on the operating plan itself, or the conclusion can be obtained based on the load forecasting method, such as using the LSTM algorithm for load forecasting.

[0041] Step S2 involves selecting typical days, which can be divided into two cases: weekdays and non-working days, each requiring the fulfillment of corresponding set conditions. Specifically, the following methods can be used:

[0042] 1) If demand for regulatory market execution, such as demand response, occurs on a working day, then select the execution date before the working day. sky( (5, 7, or 10 can be selected), but non-working days, maintenance days, days requiring user participation in demand response, ancillary service days, and days with load fluctuation exceeding the threshold should be excluded. If the result is insufficient after exclusion, the following conditions must be met. The "sky" portion should be selected sequentially from the front; it should be supplemented. Heaven, from the above After further removing the two days with the highest and lowest daily power loads from the central region, the remaining ( -2) A day is called a typical day;

[0043] 2) If demand for regulatory market execution, such as demand response, occurs on non-working days, the three most recent non-working days before the execution date shall be selected as typical days. Maintenance days, user participation in demand response, ancillary service days, and load fluctuation exceeding the threshold shall be excluded. Any days less than three days after exclusion shall be selected sequentially forward and supplemented to three days.

[0044] Step S3 involves baseline calculation using the weighted average load curve for the corresponding response period on a typical day as the historical baseline load, as shown in the following formula:

[0045]

[0046] In the formula:

[0047] The baseline load value for day j;

[0048] The load value is the value of day d before day j.

[0049] This represents the total number of typical days.

[0050] The weighting coefficient for each typical day can be set according to the local load characteristics. For example, if there are 5 typical days, the weighting coefficient A can be set in ascending order from 1 day to 5 days before the typical day, with values ​​of 50%, 25%, 12.5%, 6.25%, and 6.25% respectively.

[0051] Step S4 multiplies the predicted load value reported by the user at each time point by the historical load prediction accuracy, and adds (1 - historical load prediction accuracy) multiplied by the baseline load to obtain the new baseline load value at each time point. The higher the user's load prediction accuracy, the closer the baseline load will be to the user's prediction curve; conversely, the lower the accuracy, the closer it will be to the baseline load calculated by conventional methods. The algorithm design is as follows:

[0052] ,

[0053] In the formula:

[0054] For the accuracy of historical load forecast on day j The revised new baseline load;

[0055] The baseline load value for day j;

[0056] This refers to the accuracy of historical load forecasts.

[0057] In some preferred embodiments, when encountering extreme weather or other situations where weather changes need to be considered, the virtual power plant load baseline obtained in step S4 can be corrected by introducing a weather adjustment factor.

[0058] Specifically, the weather adjustment factor can be selected as the ratio of the average load several hours before the event to the average total load for the corresponding period on a typical day. The time period closest to the load at the time of the event, and where the likelihood of user speculation is low, can be chosen, such as the average load two hours before the event. The adjusted baseline load for users should be derived by multiplying the calculated uncorrected baseline load by the weather adjustment factor.

[0059] For users with highly volatile loads, post-event adjustments can be made based on the load fluctuation correction results. Specifically, a percentage coefficient X can be introduced as the adjustment factor, with X as the weight of the baseline load and (1-X) as the weight of the load value at the time of the event. The weighted sum of the two is then calculated as the new baseline load. The initial base load should account for a relatively large proportion to curb speculative behavior by users.

[0060] It should be noted that the load baseline can be calculated at different time granularities such as minutes, hours, and days, and those skilled in the art can make corresponding adjustments according to actual needs.

[0061] A preferred embodiment of this implementation is, for example... Figure 2 As shown, it includes:

[0062] 1. User load curve prediction results are generated from virtual power plant applications or through load forecasting algorithms;

[0063] 2. Calculate the accuracy of historical load forecasts reported by virtual power plants;

[0064] 3. Select a typical day;

[0065] 4. Baseline calculation method based on typical daily historical data;

[0066] 5. Determine the optimized virtual power plant baseline based on the load curve forecast results and the conventional baseline calculation results;

[0067] 6. Revise as needed based on weather conditions;

[0068] 7. Revise as needed based on load fluctuation rate;

[0069] 8. Publish baseline load.

[0070] The selection of typical days can be divided into two cases: weekdays and non-weekdays, each requiring the fulfillment of specific conditions. The following methods can be used for this purpose:

[0071] 1) If demand for regulatory market execution, such as demand response, occurs on a working day, then select the execution date before the working day. sky( (5, 7, or 10 can be selected), but non-working days, maintenance days, days requiring user participation in demand response, ancillary service days, and days with load fluctuation exceeding the threshold should be excluded. If the result is insufficient after exclusion, the following conditions must be met. The "sky" portion should be selected sequentially from the front; it should be supplemented. Heaven, from the above After further removing the two days with the highest and lowest daily power loads from the central region, the remaining ( -2) A day is called a typical day;

[0072] 2) If demand for regulatory market execution, such as demand response, occurs on non-working days, the three most recent non-working days before the execution date shall be selected as typical days. Maintenance days, user participation in demand response, ancillary service days, and load fluctuation exceeding the threshold shall be excluded. Any days less than three days after exclusion shall be selected sequentially forward and supplemented to three days.

[0073] Baseline calculation uses the weighted average load curve for the corresponding response event period on a typical day as the historical baseline load. It can be specified according to actual needs, as shown in the following formula:

[0074]

[0075] In the formula:

[0076] For the event period corresponding to day j The baseline load value;

[0077] For the event period corresponding to day d before day j. The average load;

[0078] This represents the total number of typical days.

[0079] The weighting coefficient for each typical day can be set according to the local load characteristics. For example, if there are 5 typical days, the weighting coefficient A can be set in ascending order from 1 day to 5 days before the typical day, with values ​​of 50%, 25%, 12.5%, 6.25%, and 6.25% respectively.

[0080] The algorithm multiplies the predicted load value reported by the user at each time point by the historical load prediction accuracy, and then adds (1 - historical load prediction accuracy) multiplied by the baseline load to obtain the new baseline load value at each time point. The higher the user's load prediction accuracy, the closer the baseline load will be to the user's prediction curve; conversely, the lower the accuracy, the closer it will be to the baseline load calculated using conventional methods. The algorithm design is as follows:

[0081] ,

[0082] In the formula:

[0083] For the accuracy of historical load forecast on day j Corrected corresponding event time period Baseline load;

[0084] For the event period corresponding to day j The baseline load value;

[0085] This refers to the accuracy of historical load forecasts.

[0086] The selection of correction factors is optional. However, when encountering extreme weather or other situations where weather changes need to be considered, a meteorological adjustment factor should be introduced for correction.

[0087] Step 6 selects the ratio of the average load of the K hours prior to the event to the average total load of the corresponding period on a selected typical day as the weather adjustment factor. The selection of the K hours prior to the event should consider that the load at this time is closest to the load at the time of the event, and that the possibility of user speculation is low; a reference value for K is 2. The user-adjusted baseline load should be obtained by multiplying the calculated uncorrected baseline load by the weather adjustment factor. The algorithm design is as follows:

[0088]

[0089] The event period corresponding to day j after further adjustment by meteorological adjustment factors. Baseline load;

[0090] For the accuracy of historical load forecast on day j Corrected corresponding event time period The baseline load value;

[0091] For the event period on day j The average load in the K hours prior to the event;

[0092] The event period corresponding to day d before day j (i.e., the selected typical day). The historical average load value in the K hours prior to the occurrence can be calculated using the following formula:

[0093] .

[0094] Step 7: For users with high load volatility, a post-event correction can be made based on the load volatility correction result. Specifically, the new baseline load can be calculated by adding the baseline load volatility correction percentage factor X to the percentage factor (1-X) of the load value at the time of the event. The initial base load should have a relatively large proportion to curb speculative behavior by users; for example, it could be set to 90%. The algorithm design is as follows:

[0095]

[0096] For the event period corresponding to the load fluctuation correction on day j. The new baseline load;

[0097] For the accuracy of historical load forecast on day j Corrected corresponding event time period The baseline load value;

[0098] The load value at the time the event occurs on day j;

[0099] X is a correction factor, which is a percentage value.

[0100] In this embodiment, the software carrier is the corresponding system module. The system consists of two parts: the baseline publisher and the user submitter. The baseline publisher is the main system and connects with the user submitter system through an interface. The user submitter can be carried out in the form of mobile terminal, web page, etc.

[0101] The baseline publisher's system has functions such as data acquisition, user load forecast data acquisition, forecast accuracy assessment, selection of typical daily data, baseline calculation based on historical data, optimization of virtual power plant baseline calculation, meteorological factor correction, volatility correction, and baseline publication;

[0102] The user reporting system has functions such as load forecast data reporting and baseline viewing.

Claims

1. A method for calculating the baseline of a virtual power plant by incorporating user confidence, characterized in that, Includes the following steps: Based on the user load forecast curve, the historical forecast load value before the event release date and the user load forecast value on the event release date are analyzed. The user confidence level is calculated using the historical predicted load values ​​and their corresponding actual load values. Select several typical days prior to the demand response date; The historical baseline load value is calculated based on the actual load value of the typical day; The virtual power plant load baseline is calculated based on the user confidence level, the historical baseline load value, and the user load forecast value. The virtual power plant load baseline is represented as follows: ， in, Indicates the first The virtual power plant load baseline mentioned on the day, This indicates the user's confidence level. Indicates the first The predicted user load value for the day, Indicates the first The historical baseline load value for that day.

2. The method according to claim 1, characterized in that, The user confidence level is calculated based on data from several normal days prior to the event publication date. These normal days do not include demand response days, emergency peak avoidance days, or power outage maintenance days.

3. The method according to claim 1, characterized in that, The historical baseline load value is a weighted average of the actual load values ​​for each of the typical days.

4. The method according to claim 1, characterized in that, It also includes the step of introducing a first adjustment factor to correct the virtual power plant load baseline, wherein the first adjustment factor is the ratio of the average load of a first set number of time periods before the event to the historical average load, wherein the historical average load is the average of the total load of all the typical days of the same time period relative to the total number of time periods.

5. The method according to claim 4, characterized in that, The virtual power plant load baseline after being corrected by the first adjustment factor is the product of the virtual power plant load baseline and the first adjustment factor.

6. The method according to claim 1, characterized in that, It also includes the step of introducing a second adjustment factor to correct the virtual power plant load baseline, wherein the second adjustment factor is a set percentage value less than 1.

7. The method according to claim 6, characterized in that, The revised virtual power plant load baseline is represented as follows: ， in, Indicates the first The virtual power plant load baseline mentioned on the day, This indicates the first adjustment factor after correction. The virtual power plant load baseline mentioned on the day, This represents the second adjustment factor. Indicates the first Actual load value at the time of the incident.

8. The method according to claim 7, characterized in that, The second adjustment factor is set to 90%.

Images