Power transmission line dynamic capacity increasing method based on bayesian optimization random forest online training

Abstract

The application relates to a power transmission line dynamic capacity increasing method based on Bayesian optimization random forest online training, and relates to the field of power transmission line detection. The application is aimed at solving the problem that the existing power transmission line dynamic capacity increasing technology has a fixed power transmission line heat balance equation model and lacks pertinence, thereby affecting the accuracy of the power transmission line dynamic capacity increasing. In the application, meteorological forecast data and conductor current values are substituted into a random forest model to calculate the conductor temperature, until the temperature is higher than 70 DEG C, an I-T curve is constructed, otherwise the conductor current value is updated, if the updated value is smaller than the maximum value, the conductor temperature is calculated by using the random forest model, otherwise the conductor temperature is calculated by using the power transmission line heat balance model; until the temperature is higher than 70 DEG C, an I-T curve is constructed, otherwise the conductor current value is updated, and the conductor temperature is calculated by using the power transmission line heat balance model; the minimum value of the current-carrying capacity limit is calculated based on the I-T curve to calculate the capacity prediction result, and the capacity prediction result is used to realize the power transmission line dynamic capacity increasing.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of power transmission line detection. BACKGROUND

[0002] With the continuous increase of power load and new energy penetration rate, the problem of insufficient capacity of power transmission lines is becoming increasingly serious. The capacity of power transmission lines is related to meteorological conditions. In engineering, the capacity of power transmission lines is often calculated by assuming extreme adverse meteorological conditions, which causes waste of the capacity of power transmission lines. In order to fully tap the transmission potential of power transmission lines, the existing dynamic capacity increasing technology of power transmission lines often calculates the capacity of power transmission lines based on the heat balance equation according to meteorological information. However, the heat balance equation model of power transmission lines is fixed and lacks pertinence, and it is difficult to accurately reflect the relationship between meteorological conditions, conductor current and conductor temperature under actual operating conditions, which affects the accuracy of the dynamic capacity increasing of power transmission lines. SUMMARY

[0003] The application is to solve the problem that the existing dynamic capacity increasing technology of power transmission lines often calculates the capacity of power transmission lines based on the heat balance equation according to meteorological information. However, the heat balance equation model of power transmission lines is fixed and lacks pertinence, and it is difficult to accurately reflect the relationship between meteorological conditions, conductor current and conductor temperature under actual operating conditions, which affects the accuracy of the dynamic capacity increasing of power transmission lines. The present application provides a dynamic capacity increasing method of power transmission lines based on Bayesian optimization random forest online training.

[0004] The dynamic capacity increasing method of power transmission lines based on Bayesian optimization random forest online training comprises:

[0005] The meteorological forecast data and the conductor current value are substituted into the random forest model to calculate the conductor temperature. The conductor current value is the minimum conductor current I tmin used when training the random forest model. The meteorological forecast data includes wind speed, wind direction, solar irradiance and environmental temperature.

[0006] If the conductor temperature calculated by the random forest model is higher than 70℃, an I-T curve is constructed using the conductor current value and the corresponding conductor temperature, otherwise the conductor current value is updated by a first step, wherein the first step is (I tmax -I tmin ), I tmax is the maximum conductor current used when training the random forest model, and the input of the random forest model is meteorological forecast data and conductor current, and the output is conductor temperature.

[0007] If the conductor current value updated by the first step is less than the maximum conductor current, the conductor temperature is recalculated using the random forest model based on the conductor current value updated by the first step, otherwise the conductor temperature is calculated according to the heat balance model of power transmission lines.

[0008] If the conductor temperature calculated by the power transmission line thermal equilibrium model is higher than 70℃, an I-T curve is constructed by using the conductor current value and the corresponding conductor temperature, otherwise the conductor current value is updated by a second step, and the conductor temperature is recalculated by the power transmission line thermal equilibrium model based on the conductor current value updated by the second step, the second step being 50A;

[0009] Based on the I-T curve, the ampacity limit corresponding to each conductor current value is calculated, the minimum value of the ampacity limit is selected to calculate the capacity prediction result, and the capacity prediction result is used to realize dynamic capacity increase of the power transmission line.

[0010] Further, the conductor temperature calculated according to the power transmission line thermal equilibrium model includes:

[0011] T=g(v,θ,I s ,T A ,I)+T k -g(v,θ,I s ,T A ,I tmax ),

[0012] Wherein, T is the conductor temperature, g() is the power transmission line thermal equilibrium model, v is the wind speed, θ is the wind direction, I s is the solar irradiance, T A is the ambient temperature, I is the conductor current, T k is the conductor temperature calculated by the random forest model when the conductor current is I tmax .

[0013] Further, the expression of the power transmission line thermal equilibrium model g() is:

[0014] a4T 4 +a3T 3 +a2T 2 +a1T+a0=0,

[0015] a4=πeSD,

[0016] a3=πeSD*(4*273),

[0017] a2=πeSD*(6*273 2 ),

[0018] a1=πeSD*(4*273 3 )+9.92(VD) 0.485 -I 2 Rα,

[0019] a0=πeSD[273 4 -(T A +273) 4 ]-I2 R(1-20α)-a s I s D-9.92T A (VD) 0.485 ,

[0020] Where e is the heat dissipation coefficient, S is the Stefan-Boltzmann constant, D is the outer diameter of the conductor, R is the AC resistance per unit length of the conductor, α is the temperature coefficient of resistance, and a s heat absorption coefficient,

[0021] The expression for the intermediate variable V is:

[0022]

[0023] in, For power factor, It is the minimum angle between the wind direction and the lines on both sides of the temperature measurement point.

[0024] Furthermore, the above-mentioned construction of the IT curve using conductor current values ​​and corresponding conductor temperatures includes:

[0025] Plot the conductor current and conductor temperature as a curve with current on the x-axis and temperature on the y-axis, and then shift the curve upwards by ΔT. α One unit, to obtain the IT curve;

[0026] The ΔT α This is the lower 99th percentile of the conductor temperature error.

[0027] Furthermore, the lower 99th percentile of the aforementioned conductor temperature error is obtained by the following formula:

[0028] P(T E <ΔT α ) = 0.99,

[0029] Among them, T E To test the temperature error of the conductor, which is the difference between the actual conductor temperature and the predicted value from the random forest model, P(T) E <ΔT α ) represents T E <ΔT α The probability of.

[0030] Furthermore, the calculation of the current-carrying limit corresponding to the current value of each conductor based on the IT curve includes:

[0031] The current carrying capacity limit is calculated using the following formula:

[0032]

[0033] Among them, I DTRI i and I i-1 are conductor current values at the i-th and i-1-th iterations, respectively, T i and T i-1 are conductor temperatures at the i-th and i-1-th iterations, respectively.

[0034] Further, the minimum value of the conductor current limit is calculated to obtain the capacity prediction result, including:

[0035] The capacity prediction result is calculated according to the following formula:

[0036]

[0037] wherein, P max is the capacity prediction result, U is the rated voltage of the power transmission line, I DTR_min is the minimum value of the conductor current limit, is the power factor, is the minimum value of the angle between the wind direction and the line on both sides of the temperature measurement point.

[0038] Further, the method for constructing the random forest model is:

[0039] Read historical meteorological forecast data, conductor current of the power transmission line and conductor temperature of the power transmission line, filter abnormal data and supplement missing data to obtain a data set;

[0040] Group the data in the data set according to the season, meteorological forecast time and the number of temperature measurement towers;

[0041] Find the optimal hyperparameters based on the data set using the Bayesian optimization algorithm, and construct the random forest model based on the optimal hyperparameters.

[0042] Further, the filtering of abnormal data includes:

[0043] Determine whether each element data in the meteorological forecast data is out of its value range, and if so, it is considered abnormal and deleted;

[0044] The value range of each element data is as follows:

[0045] The value range of the wind speed v is [0, v 5max ], v 5max is the maximum wind speed of the temperature measurement tower in the past 5 years;

[0046] The value range of the wind direction θ is [0, π), the wind direction is 0° with north as the positive direction, clockwise as the positive direction, and converted to radian, when θ ≥ π, θ minus π;

[0047] The value range of the solar irradiance I s is [0, I s5max ], Is5max the maximum value of solar radiation of the temperature measuring tower in the past 5 years;

[0048] ambient temperature T A , the value range of T A5min is [T A5max , T A5min ], T A5max and T s are the minimum and maximum values of ambient temperature of the temperature measuring tower in the past 5 years.

[0049] Further, the above meteorological forecast data is the hourly forecast value of the next 6 hours of the current hour, and the conductor current and conductor temperature of the transmission line are the measured average values of the current hour.

[0050] The transmission line dynamic capacity increasing method based on Bayesian optimization random forest online training provided by the application, based on the operation history data of the transmission line, constructs a Bayesian optimization random forest model, and combines the model with the heat balance equation of the transmission line, to obtain a model of transmission line dynamic capacity increasing that can more truly reflect the field operation data, and improve the accuracy of the transmission line dynamic capacity increasing result. BRIEF DESCRIPTION OF DRAWINGS

[0051] Figure 1 The flowchart of the transmission line dynamic capacity increasing method based on Bayesian optimization random forest online training. DETAILED DESCRIPTION

[0052] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the application. It should be noted that, in the case of no conflict, the embodiments in the application and the features in the embodiments can be combined with each other.

[0053] Referring to Figure 1 The transmission line dynamic capacity increasing method based on Bayesian optimization random forest online training in the embodiment includes the following steps.

[0054] I. Data processing

[0055] 1. Real-time reading of meteorological forecast data, conductor current of the transmission line and conductor temperature of the transmission line. The meteorological forecast data is the hourly forecast value of the next 6 hours, and the meteorological forecast data includes: wind speed v, wind direction θ, solar radiation I s and ambient temperature T A . The conductor current and conductor temperature of the transmission line are the measured average values of the current hour.

[0056] 2. Determine whether any element of the meteorological forecast data of each temperature measuring pole tower is outside the allowed value range. If so, the meteorological data is considered abnormal. The specific judgment criteria are as follows:

[0057] The wind speed v value range is [0, v 5max ], v 5max is the maximum wind speed of the temperature measuring pole tower in the past 5 years;

[0058] The wind direction θ value range is [0, π), with north as 0° and clockwise as the positive direction, converted to radian. When θ ≥ π, θ is reduced by π;

[0059] The solar irradiance I s value range is [0, I s5max ], I s5max is the maximum solar irradiance of the temperature measuring pole tower in the past 5 years;

[0060] The ambient temperature T A value range is [T A5min , T A5max ], T A5min and T A5max are the minimum and maximum ambient temperatures of the temperature measuring pole tower in the past 5 years, respectively.

[0061] Missing data refers to any missing data as missing.

[0062] Filter abnormal data and supplement missing data.

[0063] 3. Group all processed data according to season, meteorological advance prediction time (referred to as advance time) and number of temperature measuring pole towers. Assuming there are 4 temperature measuring pole towers, there are 24 data in total because it is future 6-hour weather forecast data; Since there are four seasons, there are 4x6x4 = 96 groups. Determine whether the number of data groups in the data set is greater than 5000. If it is greater than 5000, delete the oldest group of data.

[0064] II. Model construction

[0065] Divide the data set into training set, validation set and test set according to the ratio of 3:1:1. Use Bayesian optimization algorithm to find the optimal hyperparameters. Among them, set the parameters of Bayesian optimization: leaf node minimum sample size search range [1, 10], decision tree number search range [1, 100], upper limit of iteration times: 50, prohibit parallel computing. Take the root mean square error of the validation set as the objective function, execute the Bayesian optimization process, and record the optimal hyperparameters.

[0066] Based on the optimal hyperparameters, construct a random forest model, with meteorological forecast data and conductor current as input variables and conductor temperature as output variable.

[0067] Based on the built random forest model, the lower 99th percentile of the conductor temperature error in the test set is calculated by the following formula α :

[0068] P(T E <ΔT α )=0.99,

[0069] Wherein, T E is the conductor temperature error in the test set, that is, the true value of the conductor temperature minus the predicted value of the random forest model. P(T E <ΔT α ) represents the probability of T E <ΔT α .

[0070] Three, capacity prediction

[0071] Initialization: read the minimum value I tmin and the maximum value I tmax of the conductor current in each group of data set in the training set, and take the minimum value I tmin of the conductor current as the initial value of the conductor current;

[0072] S1: substitute the weather forecast data and the conductor current value I i in the i-th iteration into the random forest model to calculate the conductor temperature T i in the i-th iteration;

[0073] S2: determine whether the conductor temperature T i obtained in S1 is higher than 70℃, if yes, execute S8, otherwise execute S3:

[0074] S3: update I tmax =I tmin +[(I tmax -I tmin ) / 10] with (I i+1 -I i ) / 10 as the step size, and then execute S4;

[0075] S4: determine whether the conductor current value I i+1 is less than the maximum value I tmax of the conductor current, if yes, make i=i+1, and then return to S1, otherwise make i=i+1, and then execute S5;

[0076] S5: calculate the conductor temperature T i according to the following formula, and then execute S6:

[0077] T i =g(v,θ,I s ,T A ,Ii )+T k -g(v,θ,I s ,T A ,I tmax ),

[0078] where T k is the conductor temperature calculated by the random forest model when I i = I tmax ;

[0079] g() is the thermal equilibrium model of the transmission line, which is a monomial quartic equation about the conductor temperature T i , and the expression is:

[0080] a4T i 4 +a3T i 3 +a2T i 2 +a1T i +a0=0,

[0081] where a4=πeSD,

[0082] a3=πeSD*(4*273),

[0083] a2=πeSD*(6*273 2 ),

[0084] a1=πeSD*(4*273 3 )+9.92(VD) 0.485 -I 2 Rα,

[0085] a0=πeSD[273 4 -(T A +273) 4 ]-I 2 R(1-20α)-a s I s D-9.92T A (VD) 0.485 ,

[0086] e is the heat dissipation coefficient, S is the Stefan-Boltzmann constant, D is the conductor outer diameter, I is the conductor current, R is the conductor AC resistance per unit length, α is the resistance temperature coefficient, a s heat absorption coefficient,

[0087] The intermediate variable V is:

[0088] is the power factor, The minimum value of the angle between the wind direction and the line on both sides of the temperature measuring point.

[0089] S6: judging the conductor temperature T i whether it is higher than 70℃, if yes, executing S8, otherwise executing S7;

[0090] S7: updating I i+1 = I i + 50, making i=i+1 and returning to S5;

[0091] S8: drawing the conductor current and the conductor temperature under each iteration into a curve with the current as the horizontal coordinate and the temperature as the vertical coordinate, then shifting the curve upward by ΔT α units to obtain the I-T curve, and then executing S9;

[0092] S9: calculating the current-carrying capacity limit according to the I-T curve obtained in S8 based on the following formula, and then executing S10,

[0093]

[0094] wherein, I i-1 is the conductor current value under the i-1th iteration, T i-1 is the conductor temperature under the i-1th iteration;

[0095] S10: calculating the capacity prediction result P max according to the following formula:

[0096]

[0097] wherein, U is the rated voltage of the transmission line, I DTR_min is the minimum value of the current-carrying capacity limit.

[0098] IV. Error evaluation

[0099] The real-time conductor current and the conductor temperature of each temperature measuring rod tower are read, the real-time conductor current is substituted into each group of I-T curve respectively, the conductor temperature is calculated based on the linear interpolation method, and the actual conductor temperature is subtracted to calculate the conductor temperature error.

[0100] Although the present application is described herein with reference to particular embodiments, it is to be understood that these examples are merely illustrative of principles and applications of the present application. It should therefore be understood that numerous modifications can be made to the illustrative embodiments and that other arrangements can be devised without departing from the spirit and scope of the present application as defined by the appended claims. It should be understood that the features described in connection with one embodiment can be used in conjunction with other embodiments described herein. It should be understood that the features described in connection with one embodiment can be used in conjunction with other embodiments described herein.

Claims

1. A method for dynamic capacity expansion of transmission lines based on online training of Bayesian optimized random forest, characterized in that, include: The meteorological forecast data and conductor current values ​​are substituted into the random forest model to calculate the conductor temperature. The initial value of the conductor current is the minimum conductor current used when training the random forest model. The meteorological forecast data includes: wind speed, wind direction, solar irradiance, and ambient temperature; If the conductor temperature calculated by the random forest model is higher than 70℃, then an IT curve is constructed using the conductor current value and the corresponding conductor temperature; otherwise, the conductor current value is updated with a first-step length of [length missing]. , The maximum conductor current used when training the random forest model is the maximum value of the conductor current. The input of the random forest model is weather forecast data and conductor current, and the output is conductor temperature. If the conductor current value updated in the first step is less than the maximum conductor current value, the conductor temperature is recalculated using the random forest model based on the conductor current value updated in the first step; otherwise, the conductor temperature is calculated based on the transmission line thermal balance model. If the conductor temperature calculated by the transmission line thermal balance model is higher than 70°C, then the IT curve is constructed using the conductor current value and the corresponding conductor temperature; otherwise, the conductor current value is updated with a second step size, and the conductor temperature is recalculated using the transmission line thermal balance model based on the updated conductor current value with the second step size of 50A. Based on the IT curve, the current carrying capacity limit corresponding to the current value of each conductor is calculated. The minimum value of the current carrying capacity limit is selected to calculate the capacity prediction result, and the capacity prediction result is used to realize the dynamic capacity expansion of the transmission line. The calculation of conductor temperature based on the transmission line thermal balance model includes: ， in, For the temperature of the conductor, For the thermal balance model of transmission lines, For wind speed, For wind direction, Solar irradiance, For ambient temperature, For the current in the conductor, When the current in the conductor is taken The conductor temperature calculated by the random forest model; The thermal balance model of the transmission line The expression is: ， ， ， ， ， ， in, For heat dissipation coefficient, This is the Stefan Boltzmann constant. The outer diameter of the conductor. The AC resistance per unit length of the conductor. Temperature coefficient of resistance heat absorption coefficient, intermediate variables The expression is: ， in, For power factor, It is the minimum angle between the wind direction and the lines on both sides of the temperature measurement point.

2. The method for dynamic capacity expansion of transmission lines based on online training of Bayesian optimized random forest according to claim 1, characterized in that, The method of constructing the IT curve using conductor current value and corresponding conductor temperature includes: Plot the conductor current and conductor temperature as a curve with current on the x-axis and temperature on the y-axis, and then shift the curve upwards. One unit, to obtain the IT curve; The This is the lower 99th percentile of the conductor temperature error.

3. The method for dynamic capacity expansion of transmission lines based on online training of Bayesian optimized random forest according to claim 2, characterized in that, The lower 99th percentile of the conductor temperature error is obtained by the following formula: ， in, To test the temperature error of the conductor, which is the difference between the actual conductor temperature and the predicted value from the random forest model, express The probability of.

4. The method for dynamic capacity expansion of transmission lines based on online training of Bayesian optimized random forest according to claim 1, characterized in that, The calculation of the current carrying capacity limit corresponding to the current value of each conductor based on the IT curve includes: The current carrying capacity limit is calculated using the following formula: ， in, For the capacity limit, and The first and The conductor current value in the next iteration. and The first and The conductor temperature in the next iteration.

5. The method for dynamic capacity expansion of transmission lines based on online training of Bayesian optimized random forest according to claim 1, characterized in that, The calculation of the capacity prediction result by selecting the minimum value of the current carrying capacity limit includes: The capacity prediction result is calculated based on the following formula: ， in, For capacity prediction results, This is the rated voltage of the transmission line. This is the minimum value of the carrying capacity limit. For power factor, It is the minimum angle between the wind direction and the lines on both sides of the temperature measurement point.

6. The method for dynamic capacity expansion of transmission lines based on online training of Bayesian optimized random forest according to claim 1, characterized in that, The method for constructing the random forest model is as follows: Read historical weather forecast data, conductor current and conductor temperature of transmission lines, filter out abnormal data and fill in missing data to obtain a dataset; The data in the dataset are grouped according to season, weather forecast time, and number of temperature monitoring towers; Based on the dataset, the Bayesian optimization algorithm is used to find the optimal hyperparameters, and a random forest model is constructed based on the optimal hyperparameters.

7. The method for dynamic capacity expansion of transmission lines based on online training of Bayesian optimized random forest according to claim 6, characterized in that, The filtered abnormal data includes: Determine whether the data of each element in the weather forecast data exceeds its value range. If it does, it is considered abnormal and deleted. The value ranges for each element are as follows: wind speed The range of values ​​is , This is the highest wind speed recorded at the temperature monitoring tower in the past 5 years. wind direction The range of values ​​is The wind direction is due north. Clockwise is the positive direction, turning to radians, when hour, minus ; Solar irradiance The range of values ​​is , This represents the highest solar irradiance of the temperature measuring tower in the past 5 years. Ambient temperature The range of values ​​is , and These represent the minimum and maximum ambient temperatures of the temperature monitoring towers over the past five years.

8. The method for dynamic capacity expansion of transmission lines based on online training of Bayesian optimized random forest according to claim 1, characterized in that, The meteorological forecast data are hourly forecasts for the next 6 hours from the current hour, and the conductor current and conductor temperature of the transmission line are average measurements taken in the current hour.

Images