Novel probabilistic load flow calculation method considering wind temperature long-range correlation characteristics

Abstract

A novel probabilistic load flow calculation method considering wind temperature long-range correlation characteristics comprises the following steps: firstly, based on actually measured wind temperature data, generating a wind temperature long-range trend sequence by using a moving average method, and determining a sample scale by using a Pearson correlation coefficient method; secondly, adopting a Latin hypercube sampling Monte Carlo simulation method to generate a random wind temperature sequence, and performing correlation processing through Nataf transformation to obtain wind temperature data under multiple scenes; secondly, initializing a probabilistic load flow model, inputting wind speed and line parameters of the wind power plant, and performing conventional probabilistic load flow calculation to obtain initial current of each branch; and then, calculating the line temperature in different scenes by using a heat balance equation. And finally, calculating an impedance correction factor according to the line temperature, correcting an impedance parameter of the power transmission line, performing probabilistic load flow calculation again, and analyzing and evaluating a result. According to the method, long-range correlation characteristics of wind temperature are effectively considered, and the accuracy and reliability of probabilistic load flow calculation are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field related to power system optimization analysis, and in particular relates to a novel probabilistic power flow calculation method taking into account the long-range correlation characteristics of wind temperature. Background Art

[0002] The access capacity of clean energy such as wind and solar energy continues to grow rapidly, which has gradually transformed the power system into a "double-high" power system with "high proportion of new energy" grid connection and "high proportion of power electronic equipment" feed-in. This change has significantly increased the uncertainty factors in the operation planning of the new power system, bringing challenges to the economy, safety and reliability of the system. The probabilistic flow calculation can reflect the impact of random changes of various factors in the power system on the operation of the system, and comprehensively consider the uncertainty of changes in variables such as the power system network topology, component parameters, and node loads. It is an important tool for power system planning and safety and reliability analysis.

[0003] The document of application publication number CN111162537B discloses a Latin hypercube sampling method for probability power flow calculation based on a combined Copula function, and the specific steps include: selecting multiple Copula functions to form a combined Copula function according to the correlation of distributed energy power generation variables and the tail symmetry characteristics of distributed energy output; then using the combined Copula function to generate a random number matrix; using the Latin hypercube sampling method to sample the random number matrix and establish a sample matrix after Latin hypercube sampling; using the cubic spline interpolation method to obtain the inverse function of the cumulative distribution function of the distributed energy power generation variable; for the distributed energy power generation variable , its cumulative distribution function is , and the inverse function of the cumulative distribution function, according to the sample matrix, establish the sample matrix of the distributed energy generation power variable; use the sample matrix as the input to calculate the flow, obtain the discrete result of the output variable, and finally use the kernel density estimation to fit the discrete structure of the output variable to obtain the probability density function; the document with application publication number CN111668845B discloses a probability flow calculation method taking into account photovoltaic correlation, and the specific steps are: use the K-means clustering algorithm to cluster the light intensity data samples generated by the Monte Carlo simulation method, and use the fuzzy clustering algorithm based on the maximum deviation similarity criterion to determine the k value of the K-means clustering algorithm to solve the problem of the difficulty in selecting the k value. Secondly, use the improved particle swarm algorithm to optimize the K-means clustering algorithm to improve the accuracy of the clustering algorithm, obtain the clustering center of the K-means clustering algorithm and use it as the uncertain input of the flow calculation, which greatly reduces the calculation amount of the probability flow. Finally, in view of the situation that the correlation coefficient matrix of the input variables may not be positive definite in the semi-invariant power flow calculation, which makes the commonly used Cholesky decomposition method no longer applicable, the singular value decomposition method is used to decompose the correlation coefficient matrix; the document with application publication number CN111984906B discloses a method for fast calculation of probabilistic power flow considering the time series correlation between photovoltaic and load, and its steps are as follows: first, a day is divided into 24 time periods, and the probability density distribution models of photovoltaic output and load are established respectively by using the adaptive diffusion kernel density estimation method to improve the local adaptability of the probability model, and the correlation between the two is described by the Copula theory; then the fuzzy C-means clustering method is used to divide the photovoltaic output and load scenarios, and the scenario clustering center and the scenario occurrence probability are used instead of the Monte Carlo simulation process to perform probabilistic power flow calculation.

[0004] In summary, environmental factors such as wind speed and light have complex internal relationships and strong volatility and uncertainty, which will have a great impact on the operating status of the power system. In order to further solve this problem and accurately describe the actual operating conditions of the power system, it is urgent to propose a new probabilistic power flow calculation model that takes into account the long-range correlation characteristics of wind and temperature. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a new probabilistic power flow calculation method taking into account the long-range correlation characteristics of wind and temperature, so as to solve the problems pointed out in the background technology, such as insufficient consideration of the influence of complex environmental factors on the power system and low accuracy of power flow calculation results. The thermal balance equation is introduced into the field of probabilistic power flow calculation with line temperature and impedance parameters as the link, and the influence of the correlation characteristics of wind speed and air temperature around the transmission line on the line impedance parameters is considered to correct the power flow results. The proposed electrothermal coupling probabilistic power flow calculation method taking into account the long-range correlation characteristics of wind and temperature is proposed.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: A new probabilistic power flow calculation method taking into account the long-range correlation characteristics of wind temperature, the steps are: Step 1: Based on the measured wind temperature time series data, the initial long-term trend sequence of wind temperature data is generated by the Simple Moving Average (SMA) method. The moving average number parameter of the SMA model is modified according to the correlation of the long-term trend sequence of wind temperature, and the sample scale of the input wind temperature sequence is determined by the Pearson correlation coefficient method.

[0007] Step 2: The Monte Carlo simulation method based on Latin hypercube sampling is used to generate random wind temperature trend series data and residual series data with the same distribution as the initial input. Then, correlation processing is performed through the Nataf transform based on singular value decomposition to obtain the wind temperature series data under three different scenarios.

[0008] Step 3: Initialize the probabilistic power flow model, input the wind farm wind speed and line parameters, perform conventional probabilistic power flow calculations, and obtain the initial current of each branch. Initialize the thermal balance equation, input the wind temperature data and branch current under different scenarios to complete the thermal balance equation calculation, and obtain the line temperature parameters respectively.

[0009] Step 4: Receive the line temperature, calculate the impedance correction factor, correct the impedance parameters of the transmission line, perform the probabilistic power flow calculation again, and analyze and evaluate the output results.

[0010] Preferably, in step 1, a moving average model (SMA model) is introduced to generate an initial long-range trend sequence of wind temperature data, and the sample scale of the input wind temperature data is determined by the Pearson correlation coefficient method, which specifically includes the following steps: Step 1-1: Input the measured wind temperature series data, use the SMA model to obtain the long-term trend series of wind temperature, calculate the Pearson correlation coefficient of the series, establish the relationship between the number of moving average items of the SMA model and the Pearson correlation coefficient of the wind temperature series, and set the number of moving average items while retaining the data information to the greatest extent.

[0011] Step 1-2: Use the Pearson correlation coefficient method to establish the relationship between the wind temperature Pearson correlation coefficient and the sample scale, and select the sample scale with the strongest correlation as the model input parameter.

[0012] Preferably, in step 1-1, it is necessary to take into account both the original information contained in the input wind temperature data and the correlation between the two, wherein the SMA model calculation formula is as follows: (1); In the formula, For time point The moving average of is the size of the number of moving average items; It's time point The original data.

[0013] Preferably, in step 2, a Monte Carlo simulation method based on Latin hypercube sampling is introduced to generate a random sequence with the same distribution as the initial input, and then the random sequence is correlated using Nataf transformation based on singular value decomposition to obtain input wind temperature data sequences under multiple scenarios, which specifically includes the following steps: Step 2-1: Use the kernel density estimation method suitable for dealing with complex nonlinear relationships to model the uncertainty of wind speed and temperature and obtain the cumulative probability distribution curve of wind temperature data.

[0014] Step 2-2: Use the Latin hypercube sampling method to sample based on the wind temperature probability distribution curve to obtain a random sequence with the same distribution, which is merged with the residual sequence as the input of scenario 2.

[0015] Step 2-3: Introduce the inverse transform of Nataf transform to process the random wind temperature sequence, give the random sequence the original correlation characteristics, and merge the sequence obtained with the residual sequence as the input of scenario three. Finally, the actually measured wind temperature data is used as the input of scenario one.

[0016] Preferably, in step 2-1, a kernel density estimation method based on a Gaussian kernel function is selected to model the wind temperature data. Taking wind speed data as an example, assume It's from somewhere measured wind speed samples, then the probability density function of the wind speed sequence is The kernel density estimate of is defined as: (2); In the formula, The wind speed series Kernel density estimation function of ; is the number of wind speed samples; is the window width; For the wind speed samples; is the Gaussian kernel function, and the formula is: (3); Finally, the wind speed kernel density estimation function based on Gaussian kernel function is obtained as follows: (4); Preferably, in step 2-3, the Pearson correlation coefficient matrix between the sequences is first calculated by the Pearson correlation coefficient method, and then the random wind temperature data is subjected to correlation processing by the improved Nataf transformation method.

[0017] Preferably, in step 3, it is necessary to calculate the temperature data of the line through a conventional probabilistic power flow calculation model and a heat balance equation, and the specific steps are as follows: Step 3-1: Initialize the probability flow model, then input the wind farm wind speed, line impedance and other parameters, perform conventional probability flow calculation, and obtain the initial current of each branch. The power system probability flow calculation is based on the linear model Newton-Raphson iteration, and the node power equation in the polar coordinate system is: (5); In the formula, Node The active power and reactive power of Node The voltage amplitude, For Node The voltage phase angle With Node The voltage phase angle difference, They are the node conductance matrix and node susceptance matrix respectively. Written in matrix form: (6); In the formula, Represents the injected active power of all nodes except the balancing node and reactive power , The state variables representing the voltage magnitude and phase angle of all nodes except the equilibrium node, Represents the output quantity, including the active power and reactive power flowing through each branch. Represent the relevant equations for calculating node voltage and branch power respectively.

[0018] Step 3-2: Initialize the heat balance equation. Input the wind temperature data and initial branch current under different scenarios obtained in step 2 to complete the heat balance equation calculation and obtain the line temperature parameters. This paper selects the IEEE standard 738 heat balance equation to consider the impact of environmental factors on line temperature. The main equation is established as follows: (7); In the formula, Produce heat for the current flowing through the wire conductor; The heat generated by sunlight on the transmission line is the only heat that is not affected by the temperature of the transmission line. The amount of impact; The heat dissipation is caused by air convection; Radiate heat for the circuit.

[0019] Preferably, in step 3-2, the IEEE standard 738 heat balance equation is introduced to calculate the line temperature, wherein the detailed steps of solving the heat balance equation are: (1) Calculate the solar heat absorption power of the line based on parameters such as solar radiation.

[0020] (2) Set the initial temperature of each line .

[0021] (3) Calculate the impedance correction factor of the branch .

[0022] (4) Initial current calculated based on conventional power flow Calculate the heating power of the conductor .

[0023] (5) Input the calculated parameters and solve the equation to obtain the line temperature .

[0024] (6) The iteration accuracy is set to ,like , the line temperature converges, then jump to (7). Otherwise, let , jump back to (3) to recalculate, and iterate until the iteration accuracy requirement is met.

[0025] (7) Iteration stops and the circuit temperature is output .

[0026] Preferably, in step 4, the impedance correction factor is calculated by receiving the line temperature, so as to correct the impedance and recalculate the probability power flow. The relevant process is as follows: Step 4-1: Calculate the impedance correction factor of the line according to the temperature calculated in step 3, and correct the impedance parameters of the transmission line. The impedance correction factor calculation formula is: (8); In the formula, is the line skin effect coefficient; The reference temperature of the conductor is The corresponding resistance temperature coefficient is usually taken as .

[0027] Step 4-2: Perform the probabilistic power flow calculation again based on the corrected line impedance parameters, and analyze and evaluate the output results.

[0028] Preferably, in step 4-2, an analysis and evaluation is performed based on the probabilistic power flow results calculated under different wind temperature scenarios to verify the accuracy and effectiveness of the proposed model.

[0029] The present invention can achieve the following beneficial effects: 1. Use the moving average method (SMA) to process the measured wind temperature data to obtain the initial long-term trend sequence, eliminate short-term fluctuations, reveal long-term trends, and determine the model input sample scale by calculating the Pearson correlation coefficient.

[0030] 2. The random sequence of wind temperature is obtained by Latin hypercube sampling, and the long-range correlation characteristics between wind speed and air temperature are taken into account. The correlation of the random sequence of wind temperature is processed by the improved Nataf transform based on singular value decomposition to obtain sequence data taking into account the correlation. This can effectively deal with the situation where the correlation coefficient matrix between random variables is non-positive definite, and improve the accuracy of the power flow calculation results.

[0031] 3. The present invention directly calculates the initial circuit of each branch through conventional probabilistic power flow, and then introduces the thermal balance equation to consider the influence of environmental factors such as wind temperature on power system parameters, and outputs the line temperature according to the wind temperature data in each scenario.

[0032] 4. Calculate the impedance correction factor of each branch to correct the line impedance parameters, and finally perform probabilistic power flow calculation based on the corrected parameters to improve the accuracy and reliability of the power flow results. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The present invention will be further described below in conjunction with the accompanying drawings and embodiments: Figure 1 It is a flow chart of the electric-thermal coupling probabilistic power flow calculation model considering wind-temperature correlation proposed by the present invention; Figure 2 It is the relationship curve between the Pearson correlation coefficient and the number of moving average items between the initial wind temperature long sequence obtained by the moving average method SMA; Figure 3 It is the curve of the Pearson correlation coefficient of the wind temperature long-term series changing with the sample size; Figure 4 It is a series curve chart of wind temperature measured data and its long-term trend; Figure 5 The probability density function curves of the probability power flow calculation results before and after the introduction of the heat balance equation in winter and summer periods; Figure 6 It is a probability density function curve of the probability flow calculation results output by the model before and after taking into account the long-range correlation of the wind temperature series. DETAILED DESCRIPTION

[0034] The preferred solution is Figures 1 to 6 As shown, a new probabilistic power flow calculation method taking into account the long-range correlation characteristics of wind and temperature includes the following steps: Step 1: Based on the measured wind temperature time series data, the initial long-range trend sequence of wind temperature data is generated by the Simple Moving Average (SMA) method. The number of moving average items of the SMA model is modified according to the correlation of the long-range trend sequence of wind temperature, and the sample scale of the input wind temperature sequence is determined by the Pearson correlation coefficient method. Step 2: The Monte Carlo simulation method based on Latin hypercube sampling is used to generate random wind temperature trend series data and residual series data with the same distribution as the initial input. Then, correlation processing is performed through the Nataf transformation based on singular value decomposition to obtain wind temperature series data under three different scenarios; Step 3: Initialize the probability power flow model, input the wind farm wind speed and line parameters, perform conventional probability power flow calculations, and obtain the initial current of each branch. Initialize the thermal balance equation, input the wind temperature data and branch current under different scenarios to complete the thermal balance equation calculation, and obtain the line temperature parameters respectively; Step 4: Receive the line temperature, calculate the impedance correction factor, correct the impedance parameters of the transmission line, perform the probabilistic power flow calculation again, and analyze and evaluate the output results; In step 1: the moving average model (SMA model) is introduced to generate the initial long-range trend sequence of wind temperature data, and the sample scale of the input wind temperature data is determined by the Pearson correlation coefficient method. The specific steps include: Step 1-1: Input the measured wind temperature series data, use the SMA model to obtain the long-term trend series of wind temperature, calculate the Pearson correlation coefficient of the series, establish the relationship between the number of moving average items of the SMA model and the Pearson correlation coefficient of the wind temperature series, and set the number of moving average items while retaining the data information to a large extent.

[0035] Step 1-2: Use the Pearson correlation coefficient method to establish the relationship between the wind temperature Pearson correlation coefficient and the sample scale, and select the sample scale with the strongest correlation as the model input parameter.

[0036] In step 2: the Monte Carlo simulation method based on Latin hypercube sampling is introduced to generate a random sequence with the same distribution as the initial input. Then, the Nataf transformation based on singular value decomposition is used to process the correlation of the random sequence, and then the input wind temperature data sequence under multiple scenarios is obtained, which specifically includes the following steps: Step 2-1: Use the kernel density estimation method suitable for dealing with complex nonlinear relationships to model the uncertainty of wind speed and temperature, and obtain the cumulative probability distribution curve of wind temperature data based on the Gaussian kernel function. for a certain place measured wind speed samples, then the probability density function of the wind speed sequence is The kernel density estimate of is defined as: (4); In the formula, The wind speed series Kernel density estimation function of ; is the number of wind speed samples; is the window width; For the wind speed samples; Step 2-2: Use the Latin hypercube sampling method to sample based on the wind temperature probability distribution curve to obtain a random sequence with the same distribution, which is merged with the residual sequence as the input of scenario 2.

[0037] Step 2-3: Introduce the inverse transform of Nataf transform based on singular value decomposition to process the random wind temperature sequence, give the random sequence the original correlation characteristics, and merge the sequence obtained with the residual sequence as the input of scenario three. Finally, the actually measured wind temperature data is used as the input of scenario one.

[0038] The improved Nataf transform can handle the case where the correlation coefficient matrix between random variables is not positive definite, which expands the scope of application of the Nataf transform. is an m-dimensional random variable, is a random variable The cumulative distribution function of is as follows: (1) Using the inverse transformation of the cumulative distribution function (CDF) dimensional random variable The marginal distribution of is converted to a standard normal distribution random variable ; (9); In the formula, For the Input random variables The corresponding standard normal random variable; is the inverse function of the standard normal distribution cumulative probability distribution function; The input random variable The corresponding cumulative probability distribution function; is the dimension of the random variable.

[0039] (2) Obtained by correlation coefficient method The correlation coefficient matrix Then perform singular value decomposition on it.

[0040] (10); In the formula, is a unitary matrix; is a diagonal matrix composed of the matrix singular values ​​arranged from large to small; is an orthogonal matrix. have: (11); In the formula is a random variable and The correlation coefficient between .

[0041] In step 3: the temperature data of the line needs to be calculated through the conventional probabilistic power flow model and the heat balance equation. The specific steps are as follows: Step 3-1: Initialize the probability flow model, then input the wind farm wind speed, line impedance and other parameters, perform conventional probability flow calculation, and obtain the initial current of each branch. The power system probability flow calculation is based on the linear model Newton-Raphson iteration, and the node power equation in the polar coordinate system is: (12); In the formula, Node The active power and reactive power of Node The voltage amplitude, For Node The voltage phase angle With Node The voltage phase angle difference, They are the node conductance matrix and node susceptance matrix respectively. Written in matrix form: (13); In the formula, Represents the injected active power of all nodes except the balancing node and reactive power , The state variables representing the voltage magnitude and phase angle of all nodes except the equilibrium node, Represents the output quantity, including the active power and reactive power flowing through each branch. Represent the relevant equations for calculating node voltage and branch power respectively.

[0042] Step 3-2: Initialize the heat balance equation. Input the wind temperature data and initial branch current under different scenarios obtained in step 2 to complete the heat balance equation calculation and obtain the line temperature parameters. This paper selects the IEEE standard 738 heat balance equation to consider the impact of environmental factors on line temperature. The main equation is established as follows: (14); In the formula, Produce heat for the current flowing through the wire conductor; The heat generated by sunlight on the transmission line is the only heat that is not affected by the temperature of the transmission line. The amount of impact; The heat dissipation is caused by air convection; Radiate heat for the circuit.

[0043] In step 4: the line temperature is received to calculate the impedance correction factor, and then the impedance is corrected and the probability power flow calculation is re-calculated. The relevant process is as follows: Step 4-1: Calculate the impedance correction factor of the line according to the temperature calculated in step 3, and correct the impedance parameters of the transmission line. The impedance correction factor calculation formula is: (15); In the formula, is the line skin effect coefficient; The reference temperature of the conductor is The corresponding resistance temperature coefficient is usually taken as .

[0044] Step 4-2: Perform the probabilistic power flow calculation again based on the corrected line impedance parameters, analyze and evaluate the output results, and verify the accuracy and effectiveness of the proposed model.

[0045] Example: In order to verify the effectiveness of the present invention, the present invention selects the IEEE30 node test system in the Matpower toolkit on the MATLAB platform for simulation experiments, and connects 20 additional wind turbines with a rated power of 2.5MW at the 20 busbars, with a cut-in wind speed of 2.5m / s, a rated wind speed of 11m / s, a cut-out wind speed of 22m / s, and a power factor of 0.95. A wind speed sequence with shape parameters and scale parameters of c=4 and k=9 is selected as the input of the wind turbine. The wind speed and temperature data around the transmission line input in the experiment are taken from the data recorded by the Measurement and Instrumentation Data Center (MIDC) of the National Renewable Energy Laboratory (NREL) at Elizabeth City State University. The data duration is 2 years (January 2012 to December 2013), the time interval is 1h, and there are 17,544 data points in total, as shown in Figure 2. Figure 4 shown.

[0046] Figure 2 Pearson correlation coefficient representing the long-range trend series of wind temperature With the moving average model window width The change of window width can be easily obtained from the figure. When the temperature is close to the wind speed, it can retain the information contained in the wind temperature series as much as possible while ensuring a strong correlation.

[0047] from Figure 3 It can be seen that the correlation of the long-range trend series of wind temperature will change with the change of the sample scale of the input wind temperature data. The figure shows that the wind temperature series is positively correlated at the short scale, and the negative correlation increases sharply at the medium and long scales, and finally tends to be relatively stable, and at the sample scale The minimum Pearson correlation coefficient obtained nearby is -0.72, showing a significant negative correlation characteristic.

[0048] Figure 4 The graphs are the measured series and long-range trend series of wind speed and temperature. It can be seen from the graph that wind speed and temperature have strong complementarity.

[0049] This paper uses MAE (Mean Absolute Error) and RMSE (Root Mean Square Error) to evaluate the flow results before and after considering the correlation to verify the effectiveness of the model. Figure 5 This is the probability density distribution of the probability flow results before and after considering thermal balance in winter and summer. From the figure analysis, it can be seen that the node voltage amplitude level in winter increases after considering the thermal balance correction, and the expected value increases. pu, the voltage amplitude level is reduced in summer, and the expectation is reduced pu, and the branch power is expected to increase in winter pu, which is reduced in summer pu.

[0050] Figure 6 The comparison chart of the probability flow results before and after considering the long-range correlation of wind temperature series and the probability flow results under the measured wind temperature data is shown. Taking the measured wind temperature data of scene one as the benchmark, the MAE and RMSE of the probability density curves of branch flow power and voltage amplitude are significantly reduced before and after considering the long-range correlation of wind speed and air temperature. The branch power is reduced by 26.02% and 29.06% respectively, and the voltage amplitude is reduced by 48.66% and 50.13% respectively. That is, the accuracy of the probability flow calculation output is significantly improved after considering the long-range correlation, making it more in line with the actual situation.

[0051] From the test results, it can be seen that the new probabilistic power flow calculation model proposed in the present invention which takes into account the long-range correlation characteristics of wind temperature can effectively take into account the influence of environmental factors such as air temperature and wind speed on the probabilistic power flow of the power system, and further takes into account the long-range correlation between wind speed and air temperature on the basis of considering the thermal balance correction line impedance parameters, so as to more accurately describe the operating status of the power system.

[0052] The novel probabilistic power flow calculation model proposed in the present invention taking into account the long-range correlation characteristics of wind temperature is of great value in accurately reflecting the operating status of the power system and ensuring the safe, economical and stable operation of the power system.

[0053] The above embodiments are only preferred technical solutions of the present invention and should not be regarded as limiting the present invention. The protection scope of the present invention shall be the technical solutions recorded in the claims, including equivalent replacement solutions of the technical features in the technical solutions recorded in the claims. That is, equivalent replacement improvements within this scope are also within the protection scope of the present invention.

Claims

1. A new probabilistic power flow calculation method taking into account the long-range correlation characteristics of wind and temperature, characterized by The following steps are involved: Step 1: Based on the measured wind temperature time series data, the initial long-range trend sequence of wind temperature data is generated using the moving average method model, wherein the moving average number parameter of the moving average method model is modified according to the correlation of the wind temperature long-range trend sequence, and the sample scale of the input wind temperature sequence is determined by the Pearson correlation coefficient method; Step 2: The Monte Carlo simulation method based on Latin hypercube sampling is used to generate random wind temperature trend series data and residual series data with the same distribution as the initial input. Then, correlation processing is performed through the Nataf transformation based on singular value decomposition to obtain wind temperature series data under three different scenarios; Step 3: Initialize the probability power flow model, input the wind farm wind speed and line parameters, perform conventional probability power flow calculations, and obtain the initial current of each branch; Initialize the heat balance equation, input wind temperature data and branch current in different scenarios to complete the heat balance equation calculation, and obtain the line temperature parameters respectively; Step 4: Receive the line temperature, calculate the impedance correction factor, correct the impedance parameters of the transmission line, perform the probabilistic power flow calculation again, and analyze and evaluate the output results.

2. According to claim 1, a novel probabilistic power flow calculation method taking into account the long-range correlation characteristics of wind temperature is characterized by: In step 1, a moving average model is introduced to generate the initial long-range trend sequence of wind temperature data, and the sample scale of the input wind temperature data is determined by the Pearson correlation coefficient method, which includes the following steps: Step 1-1: Input the measured wind temperature series data, use the moving average method model to obtain the long-term trend series of wind temperature, calculate the Pearson correlation coefficient of the series, establish the relationship between the number of moving average items of the moving average method model and the Pearson correlation coefficient of the wind temperature series, and set the number of moving average items while taking into account the maximum retention of data information; Step 1-2: Use the Pearson correlation coefficient method to establish the relationship between the wind temperature Pearson correlation coefficient and the sample scale, and select the sample scale with the strongest correlation as the model input parameter.

3. According to claim 2, a novel probabilistic power flow calculation method taking into account the long-range correlation characteristics of wind temperature is characterized by: In step 1-1, it is necessary to consider both the original information contained in the input wind temperature data and the correlation between the two. The calculation formula of the moving average method model is as follows: (1)； In the formula For time point The moving average of is the size of the number of moving average items; It's time point The original data.

4. According to claim 1, a novel probabilistic power flow calculation method taking into account the long-range correlation characteristics of wind temperature is characterized by: In step 2, the Monte Carlo simulation method based on Latin hypercube sampling is introduced to generate a random sequence with the same distribution as the initial input. Then, the Nataf transform based on singular value decomposition is used to process the correlation of the random sequence, and then the input wind temperature data sequence under multiple scenarios is obtained, which includes the following steps: Step 2-1: Use the kernel density estimation method suitable for dealing with complex nonlinear relationships to model the uncertainty of wind speed and temperature, and obtain the cumulative probability distribution curve of wind temperature data; Step 2-2: Use the Latin hypercube sampling method to sample based on the wind temperature probability distribution curve to obtain a random sequence with the same distribution, which is combined with the residual sequence as the input of scenario 2; Step 2-3: Introduce the inverse transform of Nataf transform to process the random wind temperature sequence, give the random sequence the original correlation characteristics, and merge the sequence obtained with the residual sequence as the input of scenario three. Finally, the actually measured wind temperature data is used as the input of scenario one.

5. According to claim 4, a novel probabilistic power flow calculation method taking into account the long-range correlation characteristics of wind temperature is characterized by: In step 2-1, the kernel density estimation method based on Gaussian kernel function is selected to model the wind temperature data. Taking wind speed data as an example, assume It's from somewhere measured wind speed samples, then the probability density function of the wind speed sequence is The kernel density estimate of is defined as: (2)； In the formula, The wind speed series Kernel density estimation function of ; is the number of wind speed samples; is the window width; For the wind speed samples; is the Gaussian kernel function, and the formula is: (3)； Finally, the wind speed kernel density estimation function based on Gaussian kernel function is obtained as follows: (4)。 6. A novel probabilistic power flow calculation method taking into account the long-range correlation characteristics of wind temperature according to claim 4, characterized in that: In step 2-3, the Pearson correlation coefficient matrix between sequences is first calculated using the Pearson correlation coefficient method, and then the random wind temperature data is subjected to correlation processing using the improved Nataf transformation method.

7. A novel probabilistic power flow calculation method taking into account the long-range correlation characteristics of wind temperature according to claim 1, characterized in that: In step 3, the temperature data of the line needs to be calculated through the conventional probabilistic power flow calculation model and the heat balance equation. The steps are as follows: Step 3-1: Initialize the probabilistic power flow model, then input parameters such as wind farm wind speed and line impedance, perform conventional probabilistic power flow calculation, and obtain the initial current of each branch; The power system probabilistic flow calculation is based on the linear model Newton-Raphson iteration, and the node power equation in the polar coordinate system is: (5)； In the formula, Node The active power and reactive power of Node The voltage amplitude, For Node The voltage phase angle With Node The voltage phase angle difference, They are the node conductance matrix and the node susceptance matrix respectively; written in matrix form: (6)； In the formula, Represents the injected active power of all nodes except the balancing node and reactive power , The state variables representing the voltage magnitude and phase angle of all nodes except the equilibrium node, Represents the output quantity, including the active power and reactive power flowing in each branch; Represent the relevant equations for calculating node voltage and branch power respectively; Step 3-2: Initialize the heat balance equation. Input the wind temperature data and initial branch current under different scenarios obtained in step 2 to complete the heat balance equation calculation and obtain the line temperature parameters. This paper selects the IEEE standard 738 heat balance equation to consider the impact of environmental factors on line temperature. The main equation is established as follows: (7)； In the formula, Produce heat for the current flowing through the wire conductor; The heat generated by sunlight on the transmission line is the only heat that is not affected by the temperature of the transmission line. The amount of impact; The heat dissipation is caused by air convection; Radiate heat for the circuit.

8. A novel probabilistic power flow calculation method taking into account the long-range correlation characteristics of wind temperature according to claim 7, characterized in that: In step 3-2, the IEEE standard 738 heat balance equation is introduced to calculate the line temperature, where the detailed steps for solving the heat balance equation are: (1) Calculate the solar heat absorption power of the line based on parameters such as solar radiation; (2) Set the initial temperature of each line ; (3) Calculate the impedance correction factor of the branch ; (4) Initial current calculated based on conventional power flow Calculate the heating power of the conductor ; (5) Input the calculated parameters and solve the equation to obtain the line temperature ; (6) The iteration accuracy is set to ,like , if the line temperature converges, jump to (7); otherwise, let , jump back to (3) to recalculate, and iterate until the iteration accuracy requirement is met; (7) Iteration stops and the circuit temperature is output .

9. A novel probabilistic power flow calculation method taking into account the long-range correlation characteristics of wind temperature according to claim 1, characterized in that: In step 4, the impedance correction factor is calculated by receiving the line temperature, so as to correct the impedance and recalculate the probability power flow. The process is as follows: Step 4-1: Calculate the impedance correction factor of the line according to the temperature calculated in step 3, and correct the impedance parameters of the transmission line. The impedance correction factor calculation formula is: (8)； In the formula, is the line skin effect coefficient; The reference temperature of the conductor is The corresponding resistance temperature coefficient when Step 4-2: Perform the probabilistic power flow calculation again based on the corrected line impedance parameters, and analyze and evaluate the output results.

10. A novel probabilistic power flow calculation method taking into account the long-range correlation characteristics of wind temperature according to claim 9, characterized in that: In step 4-2, the probabilistic power flow results calculated under different wind temperature scenarios are analyzed and evaluated to verify the accuracy and effectiveness of the proposed model.

Images