Electric power system climbing demand calculation method, system and device considering new energy uncertainty and storage medium

Abstract

The invention discloses an electric power system climbing demand calculation method, system and device considering new energy uncertainty and a storage medium, and the method comprises the steps: calculating the statistical distribution characteristics of a load and a new energy prediction deviation based on historical data, and building a probability model of a net load prediction deviation; system operation information is obtained, and a deterministic climbing demand and an uncertain climbing demand are calculated; and finally, the up-down climbing requirement of the system is generated through superposition. According to the method, the climbing demand of the power system can be scientifically quantified, the uncertainty caused by new energy output fluctuation is effectively dealt with, and the running efficiency of the climbing market and the real-time balance capability of the power system are improved.

Description

Technical Field

[0001] This invention belongs to the technical field of electricity spot market, and in particular relates to a method, system, device and storage medium for calculating the ramp-up demand of power systems that takes into account the uncertainty of new energy sources. Background Technology

[0002] With the continuous advancement of the construction of the electricity market, many places across the country have established spot electricity markets and ancillary service markets such as peak shaving and frequency regulation. However, there is still a lack of flexible incentive mechanisms for participating entities. The active power reserve function of the system at various time scales has not yet been clearly classified and effectively implemented. In actual operation, the system regulation capacity often fails to meet the real-time power balance requirements.

[0003] Building a ramp-up market has become an important direction for improving system flexibility and promoting the consumption of new energy. Constructing a scientific and reasonable method for calculating ramp-up demand is also a key prerequisite for designing an effective market mechanism and forming accurate price signals.

[0004] Existing methods for calculating ramp demand have problems such as being unable to effectively quantify and address the risks brought about by the uncertainty of renewable energy output. The ramp demand values ​​they output differ significantly from the actual operating demand of a high-proportion renewable energy power market, making it difficult to guarantee dispatch feasibility under real fluctuation scenarios. This results in the system either facing the safety risk of insufficient regulation capacity or reducing economic operating efficiency due to overly conservative reserves. Summary of the Invention

[0005] Purpose of the Invention: This invention provides a method, system, device, and storage medium for calculating power system ramp demand that takes into account the uncertainty of renewable energy sources. It aims to solve the problem that existing technologies can only provide deterministic ramp amounts and cannot convert the random fluctuations of renewable energy into probabilistic ramp demand signals that can be executed in the market. On a day-to-day timescale, based on the statistical distribution characteristics of historical forecast deviations for load, wind power, and photovoltaics, a probabilistic model of net load forecast deviation is established. The uncertainty of renewable energy output is quantified into calculable uncertain ramp demand using a confidence interval approach, and then superimposed with deterministic ramp demand to generate a system ramp demand curve with upper and lower boundaries.

[0006] Technical solution: This invention provides a method for calculating the ramp demand of a power system that takes into account the uncertainty of new energy sources, including: S1. Based on the ramp-up market data of the period to be calculated in history, establish a probability distribution model of the single prediction error of load, wind power and photovoltaic, and establish a probability distribution model of net load deviation based on the output of the probability distribution model of the single prediction error. S2. Obtain the ramp-up market data for the period to be calculated during the calculation day, and calculate the deterministic ramp-up demand for that period. S3. Input the ramp market data for the period to be calculated during the day into the probability distribution model of net load deviation, and calculate the uncertain ramp demand for that period. S4. Overlay the deterministic ramping demand and the uncertain ramping demand for this period to generate the system ramping / down ramping demand for this period.

[0007] Furthermore, the probability distribution model for the single prediction error of load, wind power, and photovoltaic power, based on historical ramp-up market data, as described in S1, includes: S1-1. Obtain the historical scheduled number of days for the period to be calculated, including system load forecast data, wind power forecast total data, photovoltaic forecast total data, and system load measured data, wind power measured total data, and photovoltaic measured total data. Subtract the measured data from the forecast data to obtain system load forecast deviation data, wind power forecast deviation data, and photovoltaic forecast deviation data, respectively. S1-2. Normalize the system load forecast data, wind power forecast total data, and photovoltaic forecast total data respectively. S1-3. Based on the system load forecast deviation data, wind power forecast deviation data, and photovoltaic forecast deviation data, obtain the upper and lower bounds of the predetermined confidence intervals for historical load forecast deviation, historical wind power forecast deviation, and historical photovoltaic forecast deviation. Using normalized system load forecast data, total wind power forecast data, and total photovoltaic forecast data as inputs, and the upper and lower bounds of the predetermined confidence intervals for historical load forecast deviations, historical wind power forecast deviations, and historical photovoltaic forecast deviations as outputs, a probability distribution model for a single forecast error is established. The first distribution parameter that fits the inputs and outputs is then solved. The probability distribution model for the single forecast error adopts a quadratic function form.

[0008] Furthermore, the probability distribution model for establishing the net load deviation based on the output of the probability distribution model of a single prediction error described in S1 includes: S1-4. Input the historical ramp-up market data of the period to be calculated into the probability distribution model of the single prediction error using the first distribution parameter, and calculate the upper and lower bounds of the predetermined confidence interval of the load prediction deviation, the upper and lower bounds of the predetermined confidence interval of the wind power prediction deviation, and the upper and lower bounds of the predetermined confidence interval of the photovoltaic prediction deviation. The net load forecast deviation data is calculated using the following formula: Net load forecast deviation data = Load forecast deviation data - Wind power forecast deviation data - Photovoltaic forecast deviation data; Calculate the upper and lower bounds of the predetermined confidence interval for the net load forecast deviation based on the net load forecast deviation data; Using the upper and lower bounds of the predetermined confidence intervals for load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation as inputs, and the upper and lower bounds of the predetermined confidence intervals for net load forecast deviation as outputs, a probability distribution model for net load deviation is established. The second distribution parameter that fits the inputs and the outputs is then solved. The probability distribution model for net load deviation adopts a multiple linear regression structure.

[0009] Furthermore, S2 calculates the revised value of the new energy forecast for this period, including: S2-1. Obtain the ramp-up market data for the target time period t to be calculated and the next time period t+1 during the calculation day, including system load forecast data, new energy forecast data, tie line plan data, fixed output data, and the actual measured new energy data for the calculation time period a and the previous time period a-1. S2-2. Calculate the dynamic deviation correction and the new energy forecast correction value for the target time period t. The formula is: ; New energy forecast correction value [t] = New energy forecast data [t] - Dynamic deviation correction amount [t] - Initial correction deviation; in, Measured change = Measured data of new energy sources in period a - Measured data of new energy sources in period a-1; Initial correction deviation = New energy forecast data for period t - New energy measured data for period a; The predicted change [t] = the new energy prediction data for time period t+1 - the new energy prediction data for time period t.

[0010] Furthermore, the deterministic ramp requirement for this period is calculated in S2, including: S2-3. Based on system load forecast data, renewable energy forecast correction values, tie-line plan data, and fixed output data, calculate the deterministic ramp-up demand for time period t. The formula is: The deterministic ramp-up demand for time period t = the net system load for time period t+1 - the net system load for time period t, where the net system load = system load forecast data - new energy forecast correction value - tie line plan data - fixed output data.

[0011] Furthermore, the uncertain ramp-up demand in S3 for this period includes the upper and lower bounds of the predetermined confidence interval for the net load forecast deviation, which is obtained by inputting the load forecast data, wind power forecast total data, and photovoltaic forecast total data from the ramp-up market data for the period to be calculated into a probability distribution model of the net load deviation using the second distribution parameter.

[0012] Furthermore, in S4, the deterministic ramping demand and the uncertain ramping demand for this period are superimposed to generate the system's up / down ramping demand for this period, as shown in the formula: System ramp demand = max(0, deterministic ramp demand + upper bound of the predetermined confidence interval for net load forecast deviation); System ramp demand = min(0, deterministic ramp demand + lower bound of the predetermined confidence interval for net load forecast deviation); Where max represents the maximum value and min represents the minimum value.

[0013] This invention also provides a power system ramping demand calculation system that takes into account the uncertainty of new energy sources, comprising: The probability distribution model building module is used to build probability distribution models of single prediction errors for load, wind power, and photovoltaic based on the ramp-up market data of the period to be calculated in history, and to build probability distribution models of net load deviation based on the output of the probability distribution models of single prediction errors. The deterministic ramp-up demand calculation module is used to obtain ramp-up market data for the period to be calculated during the calculation day and calculate the deterministic ramp-up demand for that period. The Uncertainty Ramp Demand Calculation Module is used to input ramp market data for the period to be calculated during the calculation day into the probability distribution model of net load deviation, and calculate the uncertain ramp demand for that period. The demand generation module is used to overlay the deterministic ramping demand and the uncertain ramping demand for the current period to generate the system ramping / down ramping demand for that period.

[0014] The present invention also provides an apparatus comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method.

[0015] The present invention also provides a storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above method.

[0016] Beneficial Effects: This invention provides a method, system, device, and storage medium for calculating power system ramp demand that takes into account the uncertainty of new energy sources. Compared with existing technologies, this invention uses the upper and lower bounds of a predetermined confidence interval for historical prediction errors as the upper and lower boundaries, transforming the random fluctuations of wind and solar power into quantifiable ramp demand, thus solving the problem that traditional methods cannot reflect the random disturbances of new energy sources. By introducing a probabilistic calculation method for confidence intervals, the scientific nature and accuracy of the quantified ramp demand results are improved, effectively overcoming the limitations of traditional static estimation methods. This provides a more reliable dynamic ramp demand signal for the ramp market, ensuring the safe and stable operation of a high-proportion new energy power grid while optimizing the resource allocation efficiency of ramp ancillary services, improving the system's new energy absorption capacity, and also improving the operational efficiency of the ramp market and the real-time balance capability of the power system. Attached Figure Description

[0017] Figure 1 This is a flowchart of the method of the present invention.

[0018] Figure 2 This is a flowchart of the predictive model generation module of the power system ramp demand calculation system of the present invention.

[0019] Figure 3 This is a flowchart of the power system ramp demand calculation system demand calculation module of the present invention. Detailed Implementation

[0020] The present invention will be further illustrated below with reference to the accompanying drawings and specific embodiments. It should be understood that the following specific embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. After reading the present invention, any modifications of the present invention in various equivalent forms by those skilled in the art will fall within the scope defined by the appended claims. Example

[0021] Please see Figures 1 to 3 As shown, this invention provides a method for calculating the ramp demand of a power system that takes into account the uncertainty of new energy sources, including: S1. Based on the ramp-up market data of the period to be calculated in history, establish a probability distribution model of the single prediction error of load, wind power and photovoltaic, and establish a probability distribution model of net load deviation based on the output of the probability distribution model of the single prediction error. S2. Obtain the ramp-up market data for the period to be calculated during the calculation day, and calculate the deterministic ramp-up demand for that period. S3. Input the ramp market data for the period to be calculated during the day into the probability distribution model of net load deviation, and calculate the uncertain ramp demand for that period. S4. Overlay the deterministic ramping demand and the uncertain ramping demand for this period to generate the system ramping / down ramping demand for this period.

[0022] In this embodiment, S1 includes: S1-1. Obtain ramp-up market data for time period t in the historical scheduled days, including system load forecast data, wind power forecast total data, photovoltaic forecast total data, and system load measured data, wind power measured total data, and photovoltaic measured total data. Subtract the measured data from the forecast data to obtain system load forecast deviation data, wind power forecast deviation data, and photovoltaic forecast deviation data, respectively. In a preferred embodiment, the predetermined number of days in this step and subsequent steps is specifically 30 days; S1-2. Normalize the system load forecast data, total wind power forecast data, and total photovoltaic forecast data respectively to eliminate the influence of data dimensions and highlight statistical regularity. Taking the system load forecast data as an example, the normalization calculation formula is as follows: ;in, This represents the system load forecast data for time period t on day i. s represents the average load forecast for period t over 30 days, and s represents the standard deviation of the load forecast for period t over 30 days. This represents the standardized load forecast data for time period t on day i. S1-3. Based on the system load forecast deviation data, wind power forecast deviation data, and photovoltaic forecast deviation data, obtain the upper and lower bounds of the predetermined confidence intervals for historical load forecast deviation, historical wind power forecast deviation, and historical photovoltaic forecast deviation. In a preferred embodiment, in this step and subsequent steps, the upper bound of the predetermined confidence interval is specifically 97.5%, and the lower bound of the predetermined confidence interval is specifically 2.5%, that is, specifically obtaining the 97.5% quantile of historical load forecast deviation, the 97.5% quantile of historical wind power forecast deviation, the 97.5% quantile of historical photovoltaic forecast deviation, the 2.5% quantile of historical load forecast deviation, the 2.5% quantile of historical wind power forecast deviation, and the 2.5% quantile of historical photovoltaic forecast deviation; Using normalized system load forecast data, total wind power forecast data, and total photovoltaic forecast data as inputs, and the 97.5% quantile, 97.5% quantile, 97.5% quantile, 2.5% quantile, 2.5% quantile, and 2.5% quantile of historical load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation as outputs, a probability distribution model for a single forecast error is established. The first distribution parameter that fits the inputs and outputs is then calculated. The probability distribution model for the single forecast error adopts a quadratic function form, with distribution parameters A, B, and C, where A is the constant term coefficient, B is the linear term coefficient, and C is the quadratic term coefficient. The formula is as follows: ; ; ; ; ; ; Where A1, A2, and A3 are the constant term coefficients of the 97.5th percentile for load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation, respectively; B1, B2, and B3 are the linear term coefficients of the 97.5th percentile for load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation, respectively; and C1, C2, and C3 are the linear term coefficients of the 97.5th percentile for load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation, respectively. The coefficients for the quadratic term are: A4, A5, and A6 are the constant term coefficients for the 2.5% quantiles of load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation, respectively; B4, B5, and B6 are the linear term coefficients for the 2.5% quantiles of load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation, respectively; and C4, C5, and C6 are the quadratic term coefficients for the 2.5% quantiles of load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation, respectively. S1-4. Input the normalized historical 30-day t-period ramp-up market data, including system load forecast data, wind power forecast total data, and photovoltaic forecast total data, into the probability distribution model of the single forecast error using the first distribution parameter, and calculate the 97.5% quantile of load forecast deviation, wind power forecast deviation, photovoltaic forecast deviation, 2.5% quantile of load forecast deviation, 2.5% quantile of wind power forecast deviation, and 2.5% quantile of photovoltaic forecast deviation. The net load forecast deviation data is calculated using the formula: Net load forecast deviation data = Load forecast deviation data - Wind power forecast deviation data - Photovoltaic forecast deviation data; Based on the net load forecast deviation data, the 97.5th percentile and 2.5th percentile of the net load forecast deviation are calculated. Using the 97.5% quantile of load forecast deviation, wind power forecast deviation, photovoltaic forecast deviation, and the 2.5% quantile of load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation as inputs, and the 97.5% quantile and 2.5% quantile of net load forecast deviation as outputs, a probability distribution model of net load deviation is established. The second distribution parameter that fits the inputs and outputs is then calculated. This probability distribution model of net load deviation employs a multiple linear regression structure. The model comprehensively analyzes the coupled impact of load and renewable energy forecast deviations on the system's net load. The second distribution parameter includes... ,in The coefficients are for the 97.5th and 2.5th percentiles of the load forecast deviation. These are the coefficients for the 97.5th and 2.5th percentiles of the wind power forecast deviation. Here are the coefficients for the 97.5% and 2.5% quantiles of the photovoltaic prediction deviation, where D is the coefficient of the constant term. The formula is: ; ;in, , These are the coefficients of the 97.5th and 2.5th quantiles of the net load forecast deviation. , The coefficients for the 97.5% and 2.5% quantiles of the net load forecast deviation and the net load forecast deviation, respectively, represent the coefficients for the wind power forecast deviation. , The coefficients for the 97.5% quantile and 2.5% quantile of the net load forecast deviation, and the photovoltaic forecast deviation. , The coefficients of the constant terms in the 97.5th percentile and 2.5th percentile of the net load forecast deviation.

[0023] The deterministic ramp requirement for this period is calculated in S2, including: S2-1. Obtain the ramp-up market data for the target time period t to be calculated and the next time period t+1 during the calculation day, including system load forecast data, new energy forecast data, tie line plan data, fixed output data, and the actual measured new energy data for the calculation time period a and the previous time period a-1. S2-2. Calculate the dynamic deviation correction and the new energy forecast correction value for time period t. The formula is: The revised value of the new energy forecast [t] = the new energy forecast data [t] - the dynamic deviation correction amount [t] - the initial correction deviation; in, Measured change = Measured data of new energy sources in period a - Measured data of new energy sources in period a-1; Initial correction deviation = New energy forecast data for period t - New energy measured data for period a; The predicted change [t] = the new energy prediction data for time period t+1 - the new energy prediction data for time period t; S2-3. Based on system load forecast data, renewable energy forecast correction values, tie-line plan data, and fixed output data, calculate the deterministic ramp-up demand for time period t. The formula is: The deterministic ramp-up demand for time period t = the net system load for time period t+1 - the net system load for time period t, where the net system load = system load forecast data - new energy forecast correction value - tie line plan data - fixed output data.

[0024] The uncertain ramp-up demand in S3 for this period includes the 97.5th percentile and the 2.5th percentile of the net load forecast deviation. These are obtained by inputting the load forecast data, wind power forecast total data, and photovoltaic forecast total data from the ramp-up market data for the period to be calculated into a probability distribution model of the net load deviation using the second distribution parameter.

[0025] In S4, the deterministic ramping demand and the uncertain ramping demand for this period are superimposed to generate the system's up / down ramping demand for this period, as shown in the formula: System ramp demand = max(0, deterministic ramp demand + 97.5th percentile of net load forecast deviation); System ramp demand = min(0, deterministic ramp demand + 2.5% quantile of net load forecast deviation); Where max represents the maximum value and min represents the minimum value.

[0026] Example 2 Please see Figures 1 to 3 As shown in Embodiment 1, this invention provides a power system ramping demand calculation system that takes into account the uncertainty of new energy sources, comprising: The probability distribution model building module is used to build probability distribution models of single prediction errors for load, wind power, and photovoltaic based on the ramp-up market data of the period to be calculated in history, and to build probability distribution models of net load deviation based on the output of the probability distribution models of single prediction errors. The deterministic ramp-up demand calculation module is used to obtain ramp-up market data for the period to be calculated during the calculation day and calculate the deterministic ramp-up demand for that period. The Uncertainty Ramp Demand Calculation Module is used to input ramp market data for the period to be calculated during the calculation day into the probability distribution model of net load deviation, and calculate the uncertain ramp demand for that period. The demand generation module is used to overlay the deterministic ramping demand and the uncertain ramping demand for the current period to generate the system ramping / down ramping demand for that period.

[0027] In this embodiment, the probability distribution model establishment module includes: The historical data acquisition module is used to acquire ramp-up market data for time period t within the historical scheduled days, including system load forecast data, wind power forecast total data, photovoltaic forecast total data, and system load measured data, wind power measured total data, and photovoltaic measured total data. The module also obtains system load forecast deviation data, wind power forecast deviation data, and photovoltaic forecast deviation data by subtracting the measured data from the forecast data. In a preferred embodiment, the predetermined number of days in this step and subsequent steps is specifically 30 days; The normalization module is used to normalize the system load forecast data, the total wind power forecast data, and the total photovoltaic forecast data respectively, eliminating the influence of data dimensions and highlighting statistical regularities. Taking the system load forecast data as an example, the normalization calculation formula is as follows: ;in, This represents the system load forecast data for time period t on day i. s represents the average load forecast for period t over 30 days, and s represents the standard deviation of the load forecast for period t over 30 days. This represents the standardized load forecast data for time period t on day i. The module for establishing the probability distribution of a single prediction error is used to obtain the upper and lower bounds of the predetermined confidence intervals for historical load prediction deviations, historical wind power prediction deviations, and historical photovoltaic prediction deviations based on system load prediction deviation data, wind power prediction deviation data, and photovoltaic prediction deviation data. In a preferred embodiment, in this step and subsequent steps, the upper bound of the predetermined confidence interval is specifically 97.5%, and the lower bound of the predetermined confidence interval is specifically 2.5%, that is, specifically obtaining the 97.5% quantile of historical load forecast deviation, the 97.5% quantile of historical wind power forecast deviation, the 97.5% quantile of historical photovoltaic forecast deviation, the 2.5% quantile of historical load forecast deviation, the 2.5% quantile of historical wind power forecast deviation, and the 2.5% quantile of historical photovoltaic forecast deviation; Using normalized system load forecast data, total wind power forecast data, and total photovoltaic forecast data as inputs, and the 97.5% quantile, 97.5% quantile, 97.5% quantile, 2.5% quantile, 2.5% quantile, and 2.5% quantile of historical load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation as outputs, a probability distribution model for a single forecast error is established. The first distribution parameter that fits the inputs and outputs is then calculated. The probability distribution model for the single forecast error adopts a quadratic function form, with distribution parameters A, B, and C, where A is the constant term coefficient, B is the linear term coefficient, and C is the quadratic term coefficient. The formula is as follows: ; ; ; ; ; ; Where A1, A2, and A3 are the constant term coefficients of the 97.5th percentile for load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation, respectively; B1, B2, and B3 are the linear term coefficients of the 97.5th percentile for load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation, respectively; and C1, C2, and C3 are the linear term coefficients of the 97.5th percentile for load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation, respectively. The coefficients for the quadratic term are: A4, A5, and A6 are the constant term coefficients for the 2.5% quantiles of load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation, respectively; B4, B5, and B6 are the linear term coefficients for the 2.5% quantiles of load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation, respectively; and C4, C5, and C6 are the quadratic term coefficients for the 2.5% quantiles of load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation, respectively. The net load deviation probability distribution module is used to input the normalized historical 30-day t-period ramp-up market data, including system load forecast data, wind power forecast total data, and photovoltaic forecast total data, into a probability distribution model of a single forecast error using the first distribution parameter. The module calculates the 97.5% quantile of the load forecast deviation, wind power forecast deviation, photovoltaic forecast deviation, and the 2.5% quantile of the load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation. The net load forecast deviation data is calculated using the formula: Net load forecast deviation data = Load forecast deviation data - Wind power forecast deviation data - Photovoltaic forecast deviation data; Based on the net load forecast deviation data, the 97.5th percentile and 2.5th percentile of the net load forecast deviation are calculated. Using the 97.5% quantile of load forecast deviation, wind power forecast deviation, photovoltaic forecast deviation, and the 2.5% quantile of load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation as inputs, and the 97.5% quantile and 2.5% quantile of net load forecast deviation as outputs, a probability distribution model of net load deviation is established. The second distribution parameter that fits the inputs and outputs is then calculated. This probability distribution model of net load deviation employs a multiple linear regression structure. The model comprehensively analyzes the coupled impact of load and renewable energy forecast deviations on the system's net load. The second distribution parameter includes... ,in The coefficients are for the 97.5th and 2.5th percentiles of the load forecast deviation. These are the coefficients for the 97.5th and 2.5th percentiles of the wind power forecast deviation. Here are the coefficients for the 97.5% and 2.5% quantiles of the photovoltaic prediction deviation, where D is the coefficient of the constant term. The formula is: ; ;in, , These are the coefficients of the 97.5th and 2.5th quantiles of the net load forecast deviation. , The coefficients for the 97.5% and 2.5% quantiles of the net load forecast deviation and the net load forecast deviation, respectively, represent the coefficients for the wind power forecast deviation. , The coefficients for the 97.5% quantile and 2.5% quantile of the net load forecast deviation, and the photovoltaic forecast deviation. , The coefficients of the constant terms in the 97.5th percentile and 2.5th percentile of the net load forecast deviation.

[0028] The deterministic ramp demand calculation module calculates the deterministic ramp demand for this time period, including: The data acquisition module is used to acquire the ramp-up market data for the target time period t and the next time period t+1 during the calculation day, including system load forecast data, new energy forecast data, tie line plan data, fixed output data, and new energy measured data for the calculation time period a and the previous time period a-1. The data acquisition module is used to calculate the dynamic deviation correction and the new energy forecast correction value for time period t. The formula is as follows: The revised value of the new energy forecast [t] = the new energy forecast data [t] - the dynamic deviation correction amount [t] - the initial correction deviation; in, Measured change = Measured data of new energy sources in period a - Measured data of new energy sources in period a-1; Initial correction deviation = New energy forecast data for period t - New energy measured data for period a; The predicted change [t] = the new energy prediction data for time period t+1 - the new energy prediction data for time period t; The deterministic ramp-up demand calculation module for a specific time period is used to calculate the deterministic ramp-up demand for time period t based on system load forecast data, renewable energy forecast correction values, tie-line plan data, and fixed output data. The formula is as follows: The deterministic ramp-up demand for time period t = the net system load for time period t+1 - the net system load for time period t, where the net system load = system load forecast data - new energy forecast correction value - tie line plan data - fixed output data.

[0029] The uncertainty ramp-up demand calculation module includes the 97.5th percentile and 2.5th percentile of the net load forecast deviation for this period. It is obtained by inputting the load forecast data, wind power forecast total data, and photovoltaic forecast total data from the ramp-up market data for the period to be calculated into the probability distribution model of the net load deviation with the second distribution parameter.

[0030] The requirement generation module overlays the deterministic ramp-up requirements and uncertain ramp-up requirements for this period to generate the system's up / down ramp-up requirements for that period, using the following formula: System ramp demand = max(0, deterministic ramp demand + 97.5th percentile of net load forecast deviation); System ramp demand = min(0, deterministic ramp demand + 2.5% quantile of net load forecast deviation); Where max represents the maximum value and min represents the minimum value.

Claims

1. A method for calculating the ramp demand of a power system that takes into account the uncertainty of new energy sources, characterized in that, include: S1. Based on the ramp-up market data of the period to be calculated in history, establish a probability distribution model of the single prediction error of load, wind power and photovoltaic, and establish a probability distribution model of net load deviation based on the output of the probability distribution model of the single prediction error. S2. Obtain the ramp-up market data for the period to be calculated during the calculation day, and calculate the deterministic ramp-up demand for that period. S3. Input the ramp market data for the period to be calculated during the day into the probability distribution model of net load deviation, and calculate the uncertain ramp demand for that period. S4. Overlay the deterministic ramping demand and the uncertain ramping demand for this period to generate the system ramping / down ramping demand for this period.

2. The method for calculating power system ramp demand taking into account the uncertainty of new energy sources according to claim 1, characterized in that, The probability distribution models for single forecast errors of load, wind power, and photovoltaic power, as described in S1 based on historical ramp-up market data, include: S1-1. Obtain the historical scheduled number of days for the period to be calculated, including system load forecast data, wind power forecast total data, photovoltaic forecast total data, and system load measured data, wind power measured total data, and photovoltaic measured total data. Subtract the measured data from the forecast data to obtain system load forecast deviation data, wind power forecast deviation data, and photovoltaic forecast deviation data, respectively. S1-2. Normalize the system load forecast data, wind power forecast total data, and photovoltaic forecast total data respectively. S1-3. Based on the system load forecast deviation data, wind power forecast deviation data, and photovoltaic forecast deviation data, obtain the upper and lower bounds of the predetermined confidence intervals for historical load forecast deviation, historical wind power forecast deviation, and historical photovoltaic forecast deviation. Using normalized system load forecast data, total wind power forecast data, and total photovoltaic forecast data as inputs, and the upper and lower bounds of the predetermined confidence intervals for historical load forecast deviations, historical wind power forecast deviations, and historical photovoltaic forecast deviations as outputs, a probability distribution model for a single forecast error is established. The first distribution parameter that fits the inputs and outputs is then solved. The probability distribution model for the single forecast error adopts a quadratic function form.

3. The method for calculating power system ramping demand taking into account the uncertainty of new energy sources according to claim 2, characterized in that, The probability distribution model for establishing the net load deviation based on the output of the probability distribution model of a single prediction error described in S1 includes: S1-4. Input the historical ramp-up market data of the period to be calculated into the probability distribution model of the single prediction error using the first distribution parameter, and calculate the upper and lower bounds of the predetermined confidence interval of the load prediction deviation, the upper and lower bounds of the predetermined confidence interval of the wind power prediction deviation, and the upper and lower bounds of the predetermined confidence interval of the photovoltaic prediction deviation. The net load forecast deviation data is calculated using the following formula: Net load forecast deviation data = Load forecast deviation data - Wind power forecast deviation data - Photovoltaic forecast deviation data; Calculate the upper and lower bounds of the predetermined confidence interval for the net load forecast deviation based on the net load forecast deviation data; Using the upper and lower bounds of the predetermined confidence intervals for load forecast deviation, wind power forecast deviation, and photovoltaic forecast deviation as inputs, and the upper and lower bounds of the predetermined confidence intervals for net load forecast deviation as outputs, a probability distribution model for net load deviation is established. The second distribution parameter that fits the inputs and outputs is then solved. The probability distribution model for net load deviation adopts a multiple linear regression structure.

4. The method for calculating power system ramp demand taking into account the uncertainty of new energy sources according to claim 3, characterized in that, S2 calculates the revised value of the new energy forecast for this period, including: S2-1. Obtain the ramp-up market data for the target time period t to be calculated and the next time period t+1 during the calculation day, including system load forecast data, new energy forecast data, tie line plan data, fixed output data, and the actual measured new energy data for the calculation time period a and the previous time period a-1. S2-2. Calculate the dynamic deviation correction and the new energy forecast correction value for the target time period t. The formula is: ； New energy forecast correction value [t] = New energy forecast data [t] - Dynamic deviation correction amount [t] - Initial correction deviation; in, Measured change = Measured data of new energy sources in period a - Measured data of new energy sources in period a-1; Initial correction deviation = New energy forecast data for period t - New energy measured data for period a; The predicted change [t] = the new energy prediction data for time period t+1 - the new energy prediction data for time period t.

5. The method for calculating power system ramp demand taking into account the uncertainty of new energy sources according to claim 4, characterized in that, The deterministic ramp requirement for this period is calculated in S2, including: S2-3. Based on system load forecast data, renewable energy forecast correction values, tie-line plan data, and fixed output data, calculate the deterministic ramp-up demand for time period t. The formula is: The deterministic ramp-up demand for time period t = the net system load for time period t+1 - the net system load for time period t, where the net system load = system load forecast data - new energy forecast correction value - tie line plan data - fixed output data.

6. The method for calculating power system ramp demand taking into account the uncertainty of new energy sources according to claim 5, characterized in that, In S3, the uncertain ramp-up demand for this period includes the upper and lower bounds of the predetermined confidence interval of the net load forecast deviation, which is obtained by inputting the load forecast data, wind power forecast total data, and photovoltaic forecast total data from the ramp-up market data of the period to be calculated into the probability distribution model of the net load deviation using the second distribution parameter.

7. The method for calculating power system ramp demand taking into account the uncertainty of new energy sources according to claim 6, characterized in that, In S4, the deterministic ramping demand and the uncertain ramping demand for this period are superimposed to generate the system's up / down ramping demand for this period, as shown in the formula: System ramp demand = max(0, deterministic ramp demand + upper bound of the predetermined confidence interval for net load forecast deviation); System ramp demand = min(0, deterministic ramp demand + lower bound of the predetermined confidence interval for net load forecast deviation); Where max represents the maximum value and min represents the minimum value.

8. A power system ramping demand calculation system that takes into account the uncertainty of new energy sources, characterized in that, include: The probability distribution model building module is used to build probability distribution models of single prediction errors for load, wind power, and photovoltaic based on the ramp-up market data of the period to be calculated in history, and to build probability distribution models of net load deviation based on the output of the probability distribution models of single prediction errors. The deterministic ramp-up demand calculation module is used to obtain ramp-up market data for the period to be calculated during the calculation day and calculate the deterministic ramp-up demand for that period. The Uncertainty Ramp Demand Calculation Module is used to input ramp market data for the period to be calculated during the calculation day into the probability distribution model of net load deviation, and calculate the uncertain ramp demand for that period. The demand generation module is used to overlay the deterministic ramping demand and the uncertain ramping demand for the current period to generate the system ramping / down ramping demand for that period.

9. An apparatus comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images