Electric power intelligent adaptive compensation method and system

Abstract

The invention belongs to the technical field of power systems, and discloses a power intelligent adaptive compensation method and system. The method comprises the steps that the weather state, the environment temperature, the environment humidity, the cloud cover and the time interval in unit time in the future are collected; collecting the longitude and latitude of the area where the photovoltaic power station is located; collecting power key data of the power utilization area in unit time; preprocessing the time interval, the longitude and the latitude to obtain a season value, a solar altitude mean value, a workday label and a time period number; whether the workday and the time interval are considered comprehensively, the prediction accuracy of the reactive power demand is improved, the weather state, the temperature, the humidity, the cloud cover, the solar altitude and the predicted illumination intensity are considered comprehensively, the prediction accuracy of the direct current generating capacity is improved, and reactive power self-adaptive compensation meets the complex and changeable actual demand. And the flexibility and adaptability of reactive power compensation are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and more specifically, to a method and system for intelligent self-adaptive power compensation. Background Art

[0002] Reactive power has a direct impact on the operation of the power grid. Reactive power compensation can improve the quality of power, meet user needs, reduce economic losses caused by power quality problems, and make the power system operate efficiently, stably and economically.

[0003] It is difficult for traditional reactive power compensation equipment to make a quick response in a short time for reactive power compensation. In order to improve the reactive power compensation effect, dispatch resources, and realize the optimal dispatch and matching between user needs and resources, the existing Chinese patent with authorization announcement number CN116722561B proposes a reactive power compensation system, including: a data acquisition device, an intelligent control terminal and a reactive power compensation device, which are connected in sequence; the data acquisition device is used to collect the current load, historical load, current photovoltaic output and historical photovoltaic output of each node in the power grid; the intelligent control terminal is used to predict the load and photovoltaic output of each node in the next time period according to the current load, historical load, current photovoltaic output and historical photovoltaic output of each node; the reactive power optimization model is solved according to the load and photovoltaic output of each node in the next time period to generate the reactive power demand curve of each node; the reactive power demand curve is decomposed to obtain the reactive power demand curve of the reactive compensation device with different adjustable periodic response levels in each node, and the reactive power of the corresponding reactive compensation device is adjusted in the next time period based on the reactive power.

[0004] Although the above method can meet most scenarios, research and practical application of the above method and existing technology have found that the above method and existing technology have at least the following defects:

[0005] The existing technology only relies on the current load and historical load to predict the future load, without considering various influencing factors in the future, such as whether it is a working day and the time interval. The prediction of future photovoltaic output, that is, photovoltaic power generation, also only relies on the current photovoltaic output and historical photovoltaic output, without considering various influencing factors in the future, such as weather conditions, temperature, solar altitude angle, cloud cover and predicted light intensity, etc., resulting in low prediction accuracy of future load and future photovoltaic output, and thus making it impossible for reactive power adaptive compensation to meet the complex and changeable actual needs.

[0006] In view of this, the present invention proposes an electric power intelligent adaptive compensation method and system to solve the above problems. Summary of the invention

[0007] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solution: an intelligent adaptive power compensation method, the method comprising:

[0008] Collect environmental data and time intervals within future unit time;

[0009] Collect the longitude and latitude of the area where the photovoltaic power station is located;

[0010] Collect key power data of power consumption areas per unit time;

[0011] Preprocess the time interval, longitude and latitude to obtain seasonal values, mean solar altitude angle, weekday labels and time period numbers;

[0012] Analyze and process environmental data separately to obtain key environmental data;

[0013] Input seasonal values, key environmental data and the average solar altitude angle into the illumination prediction model to obtain the predicted illumination intensity;

[0014] Input seasonal values, key environmental data, average solar altitude angle and predicted light intensity into the power generation prediction model to obtain the predicted DC power generation;

[0015] The working day label, time period number and power key data are input into the power prediction model to obtain the predicted reactive power of the power consumption area;

[0016] According to the DC power generation and predicted reactive power, the photovoltaic reactive power is calculated and the power factor of the photovoltaic inverter is adjusted. The grid reactive power is calculated and the reactive compensation equipment of the grid is adjusted to meet the adaptive compensation of reactive power.

[0017] Furthermore, the method for obtaining the seasonal value includes:

[0018] Step 1: Get the year, month and day of the time interval and mark it as the predicted date;

[0019] Step 2: Define the period from the Beginning of Spring to the day before the Beginning of Summer as spring, the period from the Beginning of Summer to the day before the Beginning of Autumn as summer, the period from the Beginning of Autumn to the day before the Beginning of Winter as autumn, and the period from the Beginning of Winter to the day before the Beginning of Spring as winter.

[0020] Step 3: numerically label spring, summer, autumn and winter respectively to obtain spring value, summer value, autumn value and winter value;

[0021] Step 4: Match the predicted date with spring, summer, autumn and winter to obtain the seasonal value corresponding to the predicted date.

[0022] Furthermore, the method for obtaining the mean solar altitude angle includes:

[0023] S101: Calculate the Julian day according to the year, month and day of the time interval;

[0024] S102: Calculate the Julian century from the Julian day;

[0025] S103: Calculate the ecliptic inclination, solar mean anomaly and solar mean ecliptic longitude according to the Julian century number;

[0026] S104: Calculate the solar center error based on the Julian century number and the solar mean anomaly;

[0027] S105: Calculating the solar ecliptic longitude according to the solar mean ecliptic longitude and the solar center difference;

[0028] S106: Calculate the solar declination angle according to the ecliptic inclination and the solar apparent ecliptic longitude;

[0029] S107: Calculate the time equation according to the ecliptic inclination, the solar mean ecliptic longitude and the solar mean anomaly;

[0030] S108: The initial value of i is preset to be 1, and the value range of i is , Q means there are Q time points in the future unit time;

[0031] S109: Calculate the local mean solar time at the i-th time point according to the time equation and the longitude;

[0032] S110: Calculate the hour angle at the i-th time point according to the local mean solar time;

[0033] S111: Calculate the solar altitude angle at the i-th time point according to the hour angle, and add the solar altitude angle to the solar altitude angle set;

[0034] S112: Let i=i+1; if i is less than or equal to Q, execute S109 to S111; if i is greater than Q, execute S113;

[0035] S113: Calculate the average of the solar altitude angles in the solar altitude angle set, that is, obtain the average solar altitude angle.

[0036] Furthermore, the Julian date The calculation method is:

[0037] ;

[0038] The Julian century number The calculation method is:

[0039] ;

[0040] The ecliptic inclination The calculation method is:

[0041] .

[0042] The mean solar anomaly The calculation method is:

[0043] ;

[0044] The mean ecliptic longitude of the sun The calculation method is:

[0045] ;

[0046] The sun center difference The calculation method is:

[0047] ;

[0048] ;

[0049] ;

[0050] The solar optic meridian The calculation method is:

[0051] ;

[0052] For the year, is the month, D is the date, Represents the modulo operation, that is, the remainder operation.

[0053] Furthermore, the method for calculating the solar declination angle includes:

[0054] ;

[0055] is the solar declination angle;

[0056] The calculation method of the time equation is:

[0057] ;

[0058] ;

[0059] ;

[0060] ;

[0061] is the time equation, is the eccentricity of the Earth’s orbit;

[0062] The calculation method of the local mean solar time is:

[0063] ;

[0064] is the local mean solar time at the ith time point, is the i-th time point in the future unit time; is the longitude of the area where the photovoltaic power station is located, The time zone of the region where the photovoltaic power station is located;

[0065] The calculation method of the hour angle is:

[0066] ;

[0067] is the hour angle at the i-th time point;

[0068] The calculation method of the solar altitude angle is:

[0069] ;

[0070] is the solar altitude angle at the i-th time point, is the latitude of the area where the photovoltaic power station is located.

[0071] Furthermore, the method for obtaining the time period number includes:

[0072] Keep only the hours and minutes in the time interval to get the time period of the future unit time;

[0073] Pre-constructing a time period number matching table; the time period number matching table includes time periods and time period numbers corresponding to time periods, and the time periods only include hours and minutes; the time period length in the time period number matching table is greater than or equal to the time period length of the future unit time;

[0074] If the time period of the future unit time is completely included in the time period in the time period number matching table, the corresponding number in the time period number matching table is used as the time period number of the future unit time;

[0075] If the time period of the future unit time is not completely included in any time period in the time period number matching table, the proportion of the time period of the future unit time in each time period in the time period number matching table is calculated, and the time period number corresponding to the time period with the largest proportion is taken as the time period number of the future unit time.

[0076] Further, the environmental data includes weather conditions, ambient temperature, ambient humidity and cloud cover; the key environmental data includes weather condition average, weather condition standard deviation, temperature average, temperature score, humidity average and cloud cover average;

[0077] The method for obtaining the weather state average value and the weather state standard deviation includes:

[0078] Numerical annotation of weather conditions;

[0079] Establish a weather analysis set based on the weather conditions in the future unit time, wherein the weather analysis set includes the weather conditions at Q time points, and calculate the average value and standard deviation of the weather conditions in the weather analysis set;

[0080] The method for obtaining the average temperature, average humidity and average cloud cover includes:

[0081] A temperature analysis set is established for the temperature in the future unit time, the temperature analysis set includes the temperature at Q time points, and the temperature average value in the temperature analysis set is calculated; a humidity analysis set is established for the humidity in the future unit time, the humidity analysis set includes the humidity at Q time points, and the humidity average value in the humidity analysis set is calculated; a cloud amount analysis set is established for the cloud amount in the future unit time, the cloud amount analysis set includes the cloud amount at Q time points, and the cloud amount average value in the cloud amount analysis set is calculated;

[0082] Furthermore, the method for obtaining the temperature score includes:

[0083] A temperature scoring table is pre-constructed; the temperature scoring table includes temperature intervals and scores corresponding to the temperature intervals; the temperature average value is matched with the temperature interval in the temperature scoring table, and the score corresponding to the temperature interval where the temperature average value is located is the temperature score.

[0084] Furthermore, the training method of the illumination prediction model includes:

[0085] Pre-collecting a light prediction data set, wherein the light prediction data set includes light prediction data and predicted light intensity corresponding to the light prediction data, wherein the light prediction data includes seasonal values, key environmental data, and a mean value of the solar altitude angle; dividing the light prediction data set into a training set and a test set, constructing a classifier, using the light prediction data in the training set as input of the light prediction model, using the predicted light intensity in the training set as output of the light prediction model, training the classifier to obtain an initial classifier, testing the initial classifier using the test set, and outputting a classifier that meets a preset accuracy as a light prediction model, wherein the light prediction model is a naive Bayes model or a support vector machine model;

[0086] The training method of the power generation prediction model includes:

[0087] A power generation prediction data set is collected in advance, wherein the power generation prediction data set includes power generation prediction data and DC power generation corresponding to the power generation prediction data, wherein the power generation prediction data includes seasonal values, environmental key data, solar altitude angle mean and predicted light intensity; the power generation prediction data set is divided into a training set and a test set, the power generation prediction data in the training set is used as the input of a power generation prediction model, and the DC power generation in the training set is used as the output of the power generation prediction model; when the prediction accuracy on the test set reaches a preset threshold, it is considered that the power generation prediction model has converged, and the training is stopped immediately; the power generation prediction model is a gradient boosting tree model.

[0088] Further, the key power data includes active power, reactive power and apparent power;

[0089] The training method of the power prediction model includes:

[0090] A power prediction data set is collected in advance, wherein the power prediction data set includes power prediction data and predicted reactive power corresponding to the power prediction data, wherein the power prediction data includes a workday label, a time period number and power key data; the power prediction data set is divided into a training set and a test set, the power prediction data in the training set is used as the input of a power prediction model, and the predicted reactive power in the training set is used as the output of the power prediction model; when the prediction accuracy on the test set reaches a preset threshold, the power prediction model is considered to have converged, and the training is stopped immediately; the power prediction model is a long short-term memory network model.

[0091] Furthermore, the method for adaptively compensating reactive power includes:

[0092] Preset the photovoltaic reactive power ratio of the photovoltaic power station; the photovoltaic reactive power ratio is the percentage of the direct current power generation used to compensate the reactive power to the power grid to the total direct current power generation of the photovoltaic power station;

[0093] The predicted DC power generation is multiplied by the photovoltaic reactive power ratio to obtain the DC power generation used by the photovoltaic power station to compensate the reactive power to the power grid, which is marked as the first DC power generation;

[0094] The first DC power generation is multiplied by the conversion efficiency of the photovoltaic inverter to obtain the reactive power compensated by the photovoltaic power station to the power grid, which is marked as photovoltaic reactive power;

[0095] The predicted DC power generation is subtracted from the first DC power generation to obtain the DC power generation of the photovoltaic power station for providing active power to the power grid, which is marked as the second DC power generation;

[0096] The active power provided by the photovoltaic power station to the grid is obtained by multiplying the second DC power generation by the conversion efficiency of the photovoltaic inverter, which is marked as photovoltaic active power;

[0097] The inverter power factor is calculated based on the photovoltaic reactive power and photovoltaic active power;

[0098] The calculation method of the inverter power factor is:

[0099] ;

[0100] is the inverter power factor, is the photovoltaic active power, is the photovoltaic reactive power;

[0101] In the future unit time, the power factor of the photovoltaic inverter is adjusted to the calculated inverter power factor;

[0102] The reactive power of the grid is obtained by subtracting the predicted reactive power from the photovoltaic reactive power;

[0103] In the future unit time, the reactive power compensation equipment of the power grid is adjusted so that the reactive power compensated by the reactive power compensation equipment is the calculated reactive power of the power grid.

[0104] The electric power intelligent adaptive compensation system implements the electric power intelligent adaptive compensation method, including:

[0105] The first acquisition module is used to collect environmental data and time intervals within a future unit time;

[0106] The second acquisition module is used to collect the longitude and latitude of the area where the photovoltaic power station is located;

[0107] The third acquisition module is used to collect key power data of the power consumption area within a unit time;

[0108] The first processing module is used to pre-process the time interval, longitude and latitude to obtain the seasonal value, the average value of the solar altitude angle, the working day label and the time period number;

[0109] The second processing module is used to analyze and process the environmental data to obtain key environmental data;

[0110] The light prediction module is used to input seasonal values, key environmental data and the average value of the solar altitude angle into the light prediction model to obtain the predicted light intensity;

[0111] The power generation prediction module is used to input seasonal values, key environmental data, average solar altitude angle and predicted light intensity into the power generation prediction model to obtain the predicted DC power generation;

[0112] The power prediction module is used to input the working day label, time period number and power key data into the power prediction model to obtain the predicted reactive power of the power consumption area;

[0113] The power compensation module is used to calculate the photovoltaic reactive power and adjust the power factor of the photovoltaic inverter according to the DC power generation and predicted reactive power, calculate the grid reactive power, and adjust the reactive compensation equipment of the grid to meet the adaptive compensation of reactive power.

[0114] The technical effects and advantages of the intelligent adaptive power compensation method and system of the present invention are as follows: by collecting weather conditions, ambient temperature, ambient humidity, cloud cover and time interval in the future unit time, as well as collecting longitude and latitude, and preprocessing this information, more meaningful features such as seasonal values, average solar altitude angles, working day labels and time period numbers are obtained. The periodic pattern of power demand is identified through working day labels and time period numbers, thereby improving the prediction accuracy of reactive power demand. The weather conditions are numerically processed, and the average and standard deviation are calculated to better reflect the impact of weather on photovoltaic power generation. The temperature score table can be used to more accurately evaluate the impact of temperature on photovoltaic power generation efficiency. Taking into account the changes in the position of the sun, the solar altitude angle is obtained through a complex calculation process to more accurately reflect the changes in light intensity. By predicting light intensity, DC power generation can be more accurately predicted. Comprehensive consideration of whether it is a working day and time interval improves the prediction accuracy of reactive power demand. Comprehensive consideration of weather conditions, temperature, humidity, cloud cover, solar altitude angle and predicted light intensity improves the prediction accuracy of DC power generation, so that reactive power adaptive compensation meets complex and changeable actual needs and improves the flexibility and adaptability of reactive power compensation. BRIEF DESCRIPTION OF THE DRAWINGS

[0115] Figure 1 This is a schematic diagram of an intelligent adaptive power compensation system according to Embodiment 1 of the present invention;

[0116] Figure 2 This is a flow chart of the power intelligent adaptive compensation method according to Embodiment 2 of the present invention;

[0117] Figure 3 A flow chart of a method for adaptive compensation of reactive power;

[0118] Figure 4 This is a flow chart of the method for obtaining the mean solar altitude angle. DETAILED DESCRIPTION

[0119] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0120] Example 1

[0121] See also Figure 1 As shown, the electric power intelligent adaptive compensation system described in this embodiment includes a first acquisition module, a second acquisition module, a third acquisition module, a first processing module, a second processing module, a light prediction module, a power generation prediction module, a power prediction module, and a power compensation module; each module realizes data transmission through wired and / or wireless connections.

[0122] The first acquisition module is used to collect environmental data and time intervals within a future unit time, wherein the environmental data includes weather conditions, ambient temperature, ambient humidity and cloud cover.

[0123] The temperature and weather conditions in the future unit time are obtained through a meteorological station; the weather conditions include sunny, cloudy, overcast, light rain, sleet, light snow, moderate rain, moderate snow, heavy snow, heavy rain and hail;

[0124] The time interval is a continuous time period from the future unit time starting point to the future unit time ending point, and the time interval includes year, month, day, hour, minute, and second. For example, the future unit time starting point is 12:01:10 on May 20, 2024, the future unit time ending point is 13:00:30 on May 20, 2024, and the time interval is from 12:01:10 on May 20, 2024 to 13:00:30 on May 20, 2024.

[0125] The second acquisition module is used to acquire the longitude and latitude of the area where the photovoltaic power station is located.

[0126] The longitude and latitude are obtained through map software.

[0127] It should be noted that longitude and latitude information has an important impact on the prediction of light intensity, because the position of the sun and the light intensity received by the earth's surface change with time and location; for example, longitude, latitude and time determine the solar altitude angle. The higher the solar altitude angle, the shorter the path of sunlight and the greater the light intensity; conversely, the lower the solar altitude angle, the lower the light intensity. Regions at different latitudes have different sunshine hours in a year. Regions near the equator have long sunshine hours almost all year round, while high-latitude regions have long sunshine hours in summer and short sunshine hours in winter, which directly affects the light intensity received in the area where the photovoltaic power station is located.

[0128] The third acquisition module is used to collect key power data of the power consumption area within a unit time, and the key power data includes active power, reactive power and apparent power; the active power, reactive power and apparent power can be collected by a smart meter.

[0129] The first processing module is used to pre-process the time interval, longitude and latitude to obtain seasonal values, average solar altitude angles, weekday labels and time period numbers.

[0130] The method for obtaining the seasonal value includes:

[0131] Step 1: Get the year, month, and day of the time interval and mark it as the predicted date.

[0132] Step 2: Define the period from the Beginning of Spring to the day before the Beginning of Summer as spring, the period from the Beginning of Summer to the day before the Beginning of Autumn as summer, the period from the Beginning of Autumn to the day before the Beginning of Winter as autumn, and the period from the Beginning of Winter to the day before the Beginning of Spring as winter.

[0133] Step 3: Label spring, summer, autumn and winter with numerical values ​​respectively to obtain spring value, summer value, autumn value and winter value. For example, label spring as 1, summer as 2, autumn as 3 and winter as 4.

[0134] Step 4: Match the predicted date with spring, summer, autumn and winter to get the season value corresponding to the predicted date. For example, February 4, 2024 is the beginning of spring, May 5, 2024 is the beginning of summer, February 4, 2024 to May 4, 2024 is spring, and the spring value is 1. If the predicted date is March 2, 2024, the season value corresponding to the predicted date is 1.

[0135] It should be noted that due to the revolution of the earth around the sun and the existence of the obliquity of the ecliptic, the sun's direct point moves between 23°26′ north and south latitude of the earth, causing changes in the altitude of the sun at noon and the length of day and night, forming the four seasons. In different seasons, the position of the sun's direct point is different, which directly affects the light intensity at each latitude. For example, in summer, the light intensity received by high latitudes in the northern hemisphere is higher than in spring and winter.

[0136] like Figure 4 As shown, the method for obtaining the mean solar altitude angle includes:

[0137] S101: Calculate the Julian day according to the year, month and day of the time interval;

[0138] The calculation method of the Julian day is:

[0139] ;

[0140] in, Julian Day, For the year, is the month, D is the day, Indicates rounding down.

[0141] S102: Calculate the Julian century from the Julian day;

[0142] The Julian century number is calculated as follows:

[0143] ;

[0144] in, The Julian century.

[0145] S103: Calculate the ecliptic inclination, solar mean anomaly and solar mean ecliptic longitude according to the Julian century number;

[0146] The calculation method of the ecliptic inclination is:

[0147] .

[0148] in, is the inclination of the ecliptic.

[0149] The calculation method of the solar average anomaly angle is:

[0150] ;

[0151] in, is the mean anomaly angle of the sun.

[0152] The calculation method of the solar mean ecliptic longitude is:

[0153] ;

[0154] in, It is the mean ecliptic longitude of the sun.

[0155] S104: Calculate the solar center error based on the Julian century number and the solar mean anomaly;

[0156] The calculation method of the solar center difference is:

[0157] ;

[0158] ;

[0159] ;

[0160] in, is the solar center difference, is the first part of the solar center difference, is the second part of the solar center difference.

[0161] It should be noted that is the main perturbation term caused by the elliptical shape of the Earth's orbit, The harmonic correction term is used to correct the influence of double and triple mean anomaly to improve the calculation accuracy.

[0162] S105: Calculating the solar ecliptic longitude according to the solar mean ecliptic longitude and the solar center difference;

[0163] The calculation method of the solar ecliptic longitude is:

[0164] ;

[0165] in, It is the sun's ecliptic meridian. Represents the modulo operation, that is, the remainder operation.

[0166] S106: Calculate the solar declination angle according to the ecliptic inclination and the solar apparent ecliptic longitude;

[0167] The calculation method of the solar declination angle includes:

[0168] ;

[0169] in, is the solar declination angle.

[0170] S107: Calculate the time equation according to the ecliptic inclination, the solar mean ecliptic longitude and the solar mean anomaly;

[0171] The calculation method of the time equation is:

[0172] ;

[0173] ;

[0174] ;

[0175] ;

[0176] in, is the time equation, is the eccentricity of the Earth’s orbit, which is 0.016708634. is the first part of the time equation, is the second part of the time equation, is the third part of the time equation.

[0177] S108: The initial value of i is preset to be 1, and the value range of i is , Q means there are Q time points in the future unit time;

[0178] S109: Calculate the local mean solar time at the i-th time point according to the time equation and the longitude;

[0179] The calculation method of the local mean solar time is:

[0180] ;

[0181] in, is the local mean solar time at the ith time point, is the i-th time point in the future unit time; is the longitude of the area where the photovoltaic power station is located, is the time zone of the area where the photovoltaic power station is located. The value is a positive number in the Eastern Time Zone. The whole country uses the Beijing Time Zone, which is Eastern Time Zone 8. Therefore, the unified Eastern Time Zone 8 is used here. The value of is 8.

[0182] S110: Calculate the hour angle at the i-th time point according to the local mean solar time;

[0183] The calculation method of the hour angle is:

[0184] ;

[0185] in, is the hour angle at the i-th time point. The hour angle represents the angle of the sun relative to the local meridian.

[0186] S111: Calculate the solar altitude angle at the i-th time point according to the hour angle, and add the solar altitude angle to the solar altitude angle set; the solar altitude angle is calculated as follows:

[0187] ;

[0188] in, is the solar altitude angle at the i-th time point, is the latitude of the area where the photovoltaic power station is located.

[0189] S112: Let i=i+1; if i is less than or equal to Q, execute S109 to S111; if i is greater than Q, execute S113;

[0190] S113: Calculate the mean of the solar altitude angles in the solar altitude angle set, that is, obtain the mean solar altitude angle. The solar altitude angle mean calculation method is:

[0191] ;

[0192] in, is the mean solar altitude angle, is the qth solar altitude angle in the solar altitude angle set, and q is the index variable for the summation operation.

[0193] The method for obtaining the working day label includes:

[0194] Keep only the year, month, and day in the time interval and mark it as the label date; if the label date is a weekday, mark the weekday label as , if the label date is a non-working day, the working day label is marked as .

[0195] The method for obtaining the time period number includes:

[0196] Keep only the hours and minutes in the time interval to get the time period of the future unit time;

[0197] Pre-constructing a time period number matching table; the time period number matching table includes time periods and time period numbers corresponding to time periods, and the time periods only include hours and minutes; the time period length in the time period number matching table is greater than or equal to the time period length of the future unit time;

[0198] If the time period of the future unit time is completely included in the time period in the time period number matching table, the corresponding number in the time period number matching table is used as the time period number of the future unit time;

[0199] If the time period of the future unit time is not completely included in any time period in the time period number matching table, the proportion of the time period of the future unit time in each time period in the time period number matching table is calculated, and the time period number corresponding to the time period with the largest proportion is taken as the time period number of the future unit time.

[0200] For example, if the time period of the future unit time is 00:01 to 01:00, the time period 00:01 to 01:00 exists in the time period number matching table, and the time period number mark is 1, then the time period number of the future unit time is 1.

[0201] If the time period of the future unit time is 00:41 to 01:40, and the time period number matching table contains the time period 00:01 to 01:00, with the time period number marked as 1, and the time period 01:01 to 02:00, with the time period number marked as 2, then 20 minutes is in the time period 00:01 to 01:00, and the proportion is , 40 minutes in the time period 01:01 to 02:00, accounting for , that is, the time period of the future unit time is numbered 2.

[0202] It should be noted that specific activities on weekdays (such as office and industrial production) and specific activities on non-working days (such as family entertainment and rest) lead to differences in electricity demand, which have obvious periodicity. These periodic patterns can be effectively identified through weekday labels and time period numbers, thereby improving the prediction accuracy of reactive power.

[0203] The second processing module is used to analyze and process the environmental data respectively to obtain key environmental data, which include weather condition average, weather condition standard deviation, temperature average, temperature score, humidity average and cloud cover average.

[0204] The method for obtaining the weather state average value and the weather state standard deviation includes:

[0205] Numerical annotation of weather conditions;

[0206] For example, sunny days are marked as 11, cloudy days are marked as 10, overcast days are marked as 9, light rain is marked as 8, sleet is marked as 7, light snow is marked as 6, moderate rain is marked as 5, moderate snow is marked as 4, heavy snow is marked as 3, heavy rain is marked as 2, and hail is marked as 1.

[0207] The weather conditions in the future unit time are used to establish a weather analysis set, wherein the weather analysis set includes the weather conditions at Q time points, and the average weather condition value and the standard deviation of the weather condition in the weather analysis set are calculated; the specific method is as follows:

[0208] The calculation method of the weather state average value includes:

[0209] ;

[0210] The calculation method of the weather state standard deviation includes:

[0211] ;

[0212] in, is the average weather condition; is the weather status at the qth time point in the weather analysis set, and q is the index variable of the summation operation; is the standard deviation of weather conditions; the standard deviation of weather conditions reflects the concentration of changes in weather condition values. The larger the average value of weather conditions and the smaller the standard deviation of weather conditions, the better the weather conditions are, which is more conducive to photovoltaic power generation; conversely, the worse the weather conditions are, the less conducive to photovoltaic power generation.

[0213] The method for obtaining the temperature average value includes:

[0214] The temperature in the future unit time is used to establish a temperature analysis set, which includes the temperature at Q time points, and the average temperature in the temperature analysis set is calculated. The specific method is as follows:

[0215] The method for calculating the temperature average value includes:

[0216] ;

[0217] in, is the average temperature, is the temperature at the qth time point in the temperature analysis set. It should be noted that the increase in temperature is accompanied by an increase in light intensity, and the decrease in temperature is accompanied by a decrease in light intensity. Therefore, temperature can be used as one of the key factors in predicting light intensity, and q is the index variable of the summation operation.

[0218] The method for obtaining the temperature score includes:

[0219] A temperature scoring table is pre-constructed; the temperature scoring table includes temperature intervals and scores corresponding to the temperature intervals.

[0220] Match the temperature average with the temperature range in the temperature score table, and the score corresponding to the temperature range where the temperature average falls is the temperature score.

[0221] The temperature score table is shown below:

[0222]

[0223] For example, if the average temperature is 22°C, the temperature score is 10 points; if the average temperature is 28°C, the temperature score is 9 points. It should be noted that the temperature score can reflect the impact of temperature on photovoltaic power generation. The higher the temperature score, the more favorable the temperature is for photovoltaic power generation. On the contrary, the lower the temperature score, the less favorable the temperature is for photovoltaic power generation. The temperature score can improve the prediction accuracy of photovoltaic power generation.

[0224] The method for obtaining the average humidity value includes:

[0225] The humidity in the future unit time is used to establish a humidity analysis set, which includes the humidity at Q time points, and the average humidity in the humidity analysis set is calculated. The specific method is as follows:

[0226] The calculation method of the humidity average value includes:

[0227] ;

[0228] in, is the average humidity value, is the humidity at the qth time point in the humidity analysis set, and q is the index variable for the summation operation.

[0229] The method for obtaining the average cloud amount includes:

[0230] Establish a cloud amount analysis set based on the cloud amount in the future unit time, where the cloud amount analysis set includes the cloud amounts at Q time points, and calculate the average cloud amount in the cloud amount analysis set;

[0231] The calculation method of the cloud amount average is as follows:

[0232] ;

[0233] in, is the average cloud cover, is the cloud amount at the qth time point in the cloud amount analysis set, and q is the index variable for the summation operation.

[0234] It should be noted that cloud cover has a significant impact on light intensity because clouds can reflect, absorb and scatter solar radiation. For example, cloudless or lightly cloudy (0-20% cloud cover): sunlight directly hits the ground, and the light intensity is the highest; partly cloudy (20-50% cloud cover): clouds will reflect and scatter part of the sunlight, and the light intensity is weakened, but there is still strong direct light and scattered light; cloudy (50-80% cloud cover): more sunlight is reflected and absorbed by the clouds, and the light intensity is significantly weakened; the ground mainly receives scattered light, and less direct light; cloudy (80-100% cloud cover): almost all direct light is blocked by clouds, the light intensity drops significantly, and the ground mainly receives scattered light.

[0235] The light prediction module is used to input seasonal values, key environmental data and the average value of the solar altitude angle into the light prediction model to obtain the predicted light intensity.

[0236] The training method of the illumination prediction model includes:

[0237] A lighting prediction data set is collected in advance, and the lighting prediction data set includes lighting prediction data and predicted lighting intensity corresponding to the lighting prediction data. The lighting prediction data set is divided into a training set and a test set, a classifier is constructed, the lighting prediction data in the training set is used as the input of the lighting prediction model, and the predicted lighting intensity in the training set is used as the output of the lighting prediction model, the classifier is trained to obtain an initial classifier, the initial classifier is tested using the test set, and a classifier that meets a preset accuracy is output as a lighting prediction model, and the lighting prediction model is a naive Bayes model or a support vector machine model.

[0238] The power generation prediction module is used to input seasonal values, key environmental data, average solar altitude angle and predicted light intensity into the power generation prediction model to obtain the predicted DC power generation.

[0239] The training method of the power generation prediction model includes:

[0240] A power generation prediction data set is collected in advance, wherein the power generation prediction data set includes power generation prediction data and DC power generation corresponding to the power generation prediction data; the power generation prediction data set is divided into a training set and a test set, the power generation prediction data in the training set is used as the input of a power generation prediction model, and the DC power generation in the training set is used as the output of the power generation prediction model; when the prediction accuracy on the test set reaches a preset threshold, it is considered that the power generation prediction model has converged and the training is stopped immediately; the power generation prediction model is a gradient boosting tree model.

[0241] The calculation formula for prediction accuracy is: ,in, For prediction accuracy, For the The predicted DC power generation corresponding to the group power generation prediction data, For the The actual DC power generation corresponding to the group power generation forecast data.

[0242] The power prediction module is used to input the working day label, time period number and power key data into the power prediction model to obtain the predicted reactive power of the power consumption area.

[0243] The training method of the power prediction model includes:

[0244] A power prediction data set is collected in advance, wherein the power prediction data set includes power prediction data and predicted reactive power corresponding to the power prediction data; the power prediction data set is divided into a training set and a test set, the power prediction data in the training set is used as the input of a power prediction model, and the predicted reactive power in the training set is used as the output of the power prediction model; when the prediction accuracy on the test set reaches a preset threshold, it is considered that the power prediction model has converged and the training is stopped immediately; the power prediction model is a long short-term memory network model.

[0245] The power compensation module is used to calculate the photovoltaic reactive power and adjust the power factor of the photovoltaic inverter according to the DC power generation and predicted reactive power, calculate the grid reactive power, and adjust the reactive compensation equipment of the grid to meet the adaptive compensation of reactive power.

[0246] like Figure 3 As shown, the method for adaptively compensating reactive power includes:

[0247] The photovoltaic reactive power ratio of the photovoltaic power station is preset; the photovoltaic reactive power ratio is the percentage of the direct current power generation used to compensate reactive power to the power grid to the total direct current power generation of the photovoltaic power station.

[0248] For example, the DC power generation of a photovoltaic power station is 100kW, and the proportion of photovoltaic reactive power is 10%, that is, 10kW of DC power generation is used to compensate reactive power to the grid, and 90kW of DC power generation is active power. It should be noted that the DC power generated by the photovoltaic power station can be converted into reactive power for compensation through an inverter, but not all DC power can be used for reactive power compensation. This is because the main purpose of photovoltaic power generation is to supply active power to the grid, and reactive power compensation is an auxiliary function based on the needs of the grid. Therefore, the proportion of photovoltaic reactive power is preset to reasonably calculate the reactive power that needs to be compensated by the photovoltaic power station.

[0249] The predicted DC power generation is multiplied by the photovoltaic reactive power ratio to obtain the DC power generation used by the photovoltaic power station to compensate the reactive power to the power grid, which is marked as the first DC power generation;

[0250] The calculation method of the first DC power generation is:

[0251] ;

[0252] in, is the first DC power generation, is the predicted DC power generation, is the proportion of photovoltaic reactive power.

[0253] The first DC power generation is multiplied by the conversion efficiency of the photovoltaic inverter to obtain the reactive power compensated by the photovoltaic power station to the power grid, which is marked as photovoltaic reactive power;

[0254] The calculation method of the photovoltaic reactive power is:

[0255] ;

[0256] in, is the photovoltaic reactive power, is the conversion efficiency of the photovoltaic inverter. It should be noted that the function of the photovoltaic inverter is to convert direct current into reactive power and active power, such as a conversion efficiency of about 95%.

[0257] The predicted DC power generation is subtracted from the first DC power generation to obtain the DC power generation used by the photovoltaic power station to provide active power to the power grid, which is marked as the second DC power generation.

[0258] The calculation method of the second DC power generation is:

[0259]

[0260] in, is the second DC power generation.

[0261] The active power provided by the photovoltaic power station to the grid is obtained by multiplying the second DC power generation by the conversion efficiency of the photovoltaic inverter, which is marked as photovoltaic active power;

[0262] The calculation method of the photovoltaic active power is:

[0263] ;

[0264] in, is the photovoltaic active power.

[0265] The inverter power factor is calculated based on the photovoltaic reactive power and photovoltaic active power;

[0266] The calculation method of the inverter power factor is:

[0267] ;

[0268] in, is the inverter power factor. It should be noted that the relationship between photovoltaic active power, photovoltaic reactive power and apparent power is , SZ is the apparent power, is the photovoltaic active power, is the photovoltaic reactive power; according to the power factor formula Transformation ;Will In use Substitute and simplify to get .

[0269] In the future unit time, the power factor of the photovoltaic inverter is adjusted to the calculated inverter power factor.

[0270] It should be noted that in the process of converting direct current into alternating current, the photovoltaic inverter can adjust the power factor of the inverter to generate the required reactive power. The reactive power demand of the power grid changes dynamically. Therefore, the present invention provides appropriate reactive power compensation to the power grid by predicting the reactive power demand in the future unit time, thereby realizing adaptive reactive power compensation.

[0271] The reactive power of the grid is obtained by subtracting the predicted reactive power from the photovoltaic reactive power.

[0272] The calculation method of the reactive power of the power grid is:

[0273] ;

[0274] in, is the grid reactive power, To predict reactive power.

[0275] In the future unit time, the reactive power compensation equipment of the power grid is adjusted so that the reactive power compensated by the reactive power compensation equipment is the calculated reactive power of the power grid.

[0276] Example 2

[0277] See also Figure 2 As shown, this embodiment provides an intelligent adaptive power compensation method, which also includes:

[0278] Collect environmental data and time intervals within future unit time;

[0279] Collect the longitude and latitude of the area where the photovoltaic power station is located;

[0280] Collect key power data of power consumption areas per unit time;

[0281] Preprocess the time interval, longitude and latitude to obtain seasonal values, mean solar altitude angle, weekday labels and time period numbers;

[0282] Analyze and process environmental data separately to obtain key environmental data;

[0283] Input seasonal values, key environmental data and the average solar altitude angle into the illumination prediction model to obtain the predicted illumination intensity;

[0284] Input seasonal values, key environmental data, average solar altitude angle and predicted light intensity into the power generation prediction model to obtain the predicted DC power generation;

[0285] The working day label, time period number and power key data are input into the power prediction model to obtain the predicted reactive power of the power consumption area;

[0286] According to the DC power generation and predicted reactive power, the photovoltaic reactive power is calculated and the power factor of the photovoltaic inverter is adjusted. The grid reactive power is calculated and the reactive compensation equipment of the grid is adjusted to meet the adaptive compensation of reactive power.

[0287] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

[0288] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. The power intelligent adaptive compensation method is characterized by: include: Collect environmental data and time intervals within future unit time; Collect the longitude and latitude of the area where the photovoltaic power station is located; Collect key power data of power consumption areas per unit time; Preprocess the time interval, longitude and latitude to obtain seasonal values, mean solar altitude angle, weekday labels and time period numbers; Analyze and process environmental data separately to obtain key environmental data; Input seasonal values, key environmental data and the average solar altitude angle into the illumination prediction model to obtain the predicted illumination intensity; Input seasonal values, key environmental data, average solar altitude angle and predicted light intensity into the power generation prediction model to obtain the predicted DC power generation; The working day label, time period number and power key data are input into the power prediction model to obtain the predicted reactive power of the power consumption area; According to the DC power generation and predicted reactive power, the photovoltaic reactive power is calculated and the power factor of the photovoltaic inverter is adjusted. The grid reactive power is calculated and the reactive compensation equipment of the grid is adjusted to meet the adaptive compensation of reactive power.

2. The power intelligent adaptive compensation method according to claim 1 is characterized in that: The method for adaptively compensating reactive power includes: Preset the photovoltaic reactive power ratio of the photovoltaic power station; the photovoltaic reactive power ratio is the percentage of the direct current power generation used to compensate the reactive power to the power grid to the total direct current power generation of the photovoltaic power station; The predicted DC power generation is multiplied by the photovoltaic reactive power ratio to obtain the DC power generation used by the photovoltaic power station to compensate the reactive power to the power grid, which is marked as the first DC power generation; The first DC power generation is multiplied by the conversion efficiency of the photovoltaic inverter to obtain the reactive power compensated by the photovoltaic power station to the power grid, which is marked as photovoltaic reactive power; The predicted DC power generation is subtracted from the first DC power generation to obtain the DC power generation of the photovoltaic power station for providing active power to the power grid, which is marked as the second DC power generation; The active power provided by the photovoltaic power station to the grid is obtained by multiplying the second DC power generation by the conversion efficiency of the photovoltaic inverter, which is marked as photovoltaic active power; The inverter power factor is calculated based on the photovoltaic reactive power and photovoltaic active power; In the future unit time, the power factor of the photovoltaic inverter is adjusted to the calculated inverter power factor; The reactive power of the grid is obtained by subtracting the predicted reactive power from the photovoltaic reactive power; In the future unit time, the reactive power compensation equipment of the power grid is adjusted so that the reactive power compensated by the reactive power compensation equipment is the calculated reactive power of the power grid.

3. The electric power intelligent adaptive compensation method according to claim 2 is characterized in that: The method for obtaining the mean value of the solar altitude angle comprises: S101: Calculate the Julian day according to the year, month and day of the time interval; S102: Calculate the Julian century from the Julian day; S103: Calculate the ecliptic inclination, solar mean anomaly and solar mean ecliptic longitude according to the Julian century number; S104: Calculate the solar center error based on the Julian century number and the solar mean anomaly; S105: Calculating the solar ecliptic longitude according to the solar mean ecliptic longitude and the solar center difference; S106: Calculate the solar declination angle according to the ecliptic inclination and the solar apparent ecliptic longitude; S107: Calculate the time equation according to the ecliptic inclination, the solar mean ecliptic longitude and the solar mean anomaly; S108: The initial value of i is preset to be 1, and the value range of i is , Q means there are Q time points in the future unit time; S109: Calculate the local mean solar time at the i-th time point according to the time equation and the longitude; S110: Calculate the hour angle at the i-th time point according to the local mean solar time; S111: Calculate the solar altitude angle at the i-th time point according to the hour angle, and add the solar altitude angle to the solar altitude angle set; S112: Let i=i+1; if i is less than or equal to Q, execute S109 to S111; if i is greater than Q, execute S113; S113: Calculate the average of the solar altitude angles in the solar altitude angle set, that is, obtain the average solar altitude angle.

4. The electric power intelligent adaptive compensation method according to claim 3 is characterized in that: The method for obtaining the time period number includes: Keep only the hours and minutes in the time interval to get the time period of the future unit time; Pre-constructing a time period number matching table; the time period number matching table includes time periods and time period numbers corresponding to time periods, and the time periods only include hours and minutes; the time period length in the time period number matching table is greater than or equal to the time period length of the future unit time; If the time period of the future unit time is completely included in the time period in the time period number matching table, the corresponding number in the time period number matching table is used as the time period number of the future unit time; If the time period of the future unit time is not completely included in any time period in the time period number matching table, the proportion of the time period of the future unit time in each time period in the time period number matching table is calculated, and the time period number corresponding to the time period with the largest proportion is taken as the time period number of the future unit time.

5. The electric power intelligent adaptive compensation method according to claim 4 is characterized in that: The environmental data include weather conditions, ambient temperature, ambient humidity and cloud cover; the key environmental data include weather condition average, weather condition standard deviation, temperature average, temperature score, humidity average and cloud cover average; The method for obtaining the weather state average value and the weather state standard deviation includes: Numerical annotation of weather conditions; Establish a weather analysis set based on the weather conditions in the future unit time, wherein the weather analysis set includes the weather conditions at Q time points, and calculate the average value and standard deviation of the weather conditions in the weather analysis set; The method for obtaining the average temperature, average humidity and average cloud cover includes: A temperature analysis set is established for the temperature in the future unit time, and the temperature analysis set includes the temperatures at Q time points, and the temperature average value in the temperature analysis set is calculated; a humidity analysis set is established for the humidity in the future unit time, and the humidity analysis set includes the humidity at Q time points, and the humidity average value in the humidity analysis set is calculated; a cloud cover analysis set is established for the cloud cover in the future unit time, and the cloud cover analysis set includes the cloud cover at Q time points, and the cloud cover average value in the cloud cover analysis set is calculated.

6. The electric power intelligent adaptive compensation method according to claim 5 is characterized in that: The method for obtaining the temperature score includes: A temperature scoring table is pre-constructed; the temperature scoring table includes temperature intervals and scores corresponding to the temperature intervals; the temperature average value is matched with the temperature interval in the temperature scoring table, and the score corresponding to the temperature interval where the temperature average value is located is the temperature score.

7. The electric power intelligent adaptive compensation method according to claim 6 is characterized in that: The training method of the illumination prediction model includes: Pre-collecting a light prediction data set, wherein the light prediction data set includes light prediction data and predicted light intensity corresponding to the light prediction data, wherein the light prediction data includes seasonal values, key environmental data, and a mean value of the solar altitude angle; dividing the light prediction data set into a training set and a test set, constructing a classifier, using the light prediction data in the training set as input of the light prediction model, using the predicted light intensity in the training set as output of the light prediction model, training the classifier to obtain an initial classifier, testing the initial classifier using the test set, and outputting a classifier that meets a preset accuracy as a light prediction model, wherein the light prediction model is a naive Bayes model or a support vector machine model; The power generation prediction model is obtained by using the training method of the illumination prediction model.

8. The electric power intelligent adaptive compensation method according to claim 7 is characterized in that: The key power data include active power, reactive power and apparent power; The training method of the power prediction model includes: A power prediction data set is collected in advance, wherein the power prediction data set includes power prediction data and predicted reactive power corresponding to the power prediction data, wherein the power prediction data includes a workday label, a time period number and power key data; the power prediction data set is divided into a training set and a test set, the power prediction data in the training set is used as the input of a power prediction model, and the predicted reactive power in the training set is used as the output of the power prediction model; when the prediction accuracy on the test set reaches a preset threshold, the power prediction model is considered to have converged, and the training is stopped immediately; the power prediction model is a long short-term memory network model.

9. The electric power intelligent adaptive compensation method according to claim 1, characterized in that: The key power data include active power, reactive power and apparent power; the key environmental data include weather condition average, weather condition standard deviation, temperature average, temperature score, humidity average and cloud cover average; The method for obtaining the seasonal value includes: Step 1: Get the year, month and day of the time interval and mark it as the predicted date; Step 2: Define the period from the Beginning of Spring to the day before the Beginning of Summer as spring, the period from the Beginning of Summer to the day before the Beginning of Autumn as summer, the period from the Beginning of Autumn to the day before the Beginning of Winter as autumn, and the period from the Beginning of Winter to the day before the Beginning of Spring as winter. Step 3: numerically label spring, summer, autumn and winter respectively to obtain spring value, summer value, autumn value and winter value; Step 4: Match the predicted date with spring, summer, autumn and winter to obtain the seasonal value corresponding to the predicted date.

10. An intelligent power adaptive compensation system, implementing the intelligent power adaptive compensation method according to any one of claims 1 to 9, characterized in that: include: The first acquisition module is used to collect environmental data and time intervals within a future unit time; The second acquisition module is used to collect the longitude and latitude of the area where the photovoltaic power station is located; The third acquisition module is used to collect key power data of the power consumption area per unit time; The first processing module is used to pre-process the time interval, longitude and latitude to obtain the seasonal value, the average value of the solar altitude angle, the working day label and the time period number; The second processing module is used to analyze and process the environmental data to obtain key environmental data; The light prediction module is used to input seasonal values, key environmental data and the average value of the solar altitude angle into the light prediction model to obtain the predicted light intensity; The power generation prediction module is used to input seasonal values, key environmental data, average solar altitude angle and predicted light intensity into the power generation prediction model to obtain the predicted DC power generation; The power prediction module is used to input the working day label, time period number and power key data into the power prediction model to obtain the predicted reactive power of the power consumption area; The power compensation module is used to calculate the photovoltaic reactive power and adjust the power factor of the photovoltaic inverter according to the DC power generation and predicted reactive power, calculate the grid reactive power, and adjust the reactive compensation equipment of the grid to meet the adaptive compensation of reactive power.

Images