Method and system for improving grid supply load prediction accuracy of power grid

Abstract

The invention relates to a method for improving power grid supply load prediction accuracy. The method comprises the following steps: acquiring whole-grid power load historical data, whole-grid power load actual data and meteorological data; calculating load similarity according to a dynamic time warping algorithm; screening similar historical loads from the load similarity; determining whether the small electric loads are consistent or not according to a whole-society load curve in similar historical loads; determining meteorological characteristics influencing the power load according to the historical data and the meteorological data of the power load of the whole network; generating a power load prediction model according to the meteorological characteristics and the whole-network power load historical data corresponding to the meteorological characteristics; and evaluating the accuracy of the power load prediction value of the power load prediction model according to the power load actual value, and obtaining the evaluation value of the power load prediction model. According to the technical scheme, the reference values of the similar influence factors can be searched out through the storage and memory functions of software, and the prediction accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the field of electric power technology, and in particular to a method, system, readable storage medium and computer equipment for improving the accuracy of power grid load prediction. Background Art

[0002] With the standardization of the operation of the power spot market, the requirements for load forecasting accuracy have also increased. The load forecast of the Honghe Power Grid is mainly affected by the following factors: Affected by the output of small power generation: Small hydropower generation is basically run-of-river power generation, and the power generation output varies greatly during the rainy season. And with the development of distributed photovoltaics, residential electricity consumption is greatly affected by temperature changes, resulting in large fluctuations in the grid supply load. Factors such as maintenance and energy efficiency control lead to power outages for users and power plants, and the grid supply load is affected. The changes in the grid supply load affected by the above factors are all judged subjectively by the mode personnel, which are greatly affected by the subjective factors of the forecast personnel, and the prediction accuracy varies greatly. The load data of the Honghe Power Grid has not yet achieved typical data records. It is entirely based on the memory of the forecast personnel to find the load forecast data with similar influencing factors for reference, which will waste a lot of time for the forecast personnel, and because of the long time and large amount of data, it is difficult to remember the influencing factors accurately, resulting in inaccurate reference data, thereby affecting the accuracy of the load forecast. Summary of the invention

[0003] The present invention aims to solve at least one of the technical problems existing in the prior art or related art.

[0004] To this end, the purpose of the present invention is to provide a method for improving the accuracy of power grid load prediction, which can utilize the storage and memory function of the software to search for reference values ​​of similar influencing factors and improve the prediction accuracy.

[0005] To achieve the above-mentioned purpose, the technical solution of the first aspect of the present invention provides a method for improving the accuracy of power grid load prediction, comprising:

[0006] Obtaining the historical data of the power load of the entire network, the actual data of the power load of the entire network corresponding to 96 o'clock every day, and the meteorological data corresponding to the historical data of the power load of the entire network;

[0007] Calculate the load similarity of each day in the historical effective data of the whole network power load and 96 points of the day before the predicted day in the actual data of the whole network power load according to the dynamic time warping algorithm;

[0008] Filtering similar historical loads from the load similarities according to a preset similarity threshold;

[0009] Determine whether the small power load is consistent based on the overall social load curve in similar historical loads;

[0010] If so, the grid supply load is consistent;

[0011] If not, calculate the impact of rainfall and temperature on power generation load and power consumption load to obtain the actual value of power load;

[0012] Determine the meteorological characteristics that affect the power load according to the historical data of the power load of the entire network and the meteorological data;

[0013] Generate a power load forecasting model according to the meteorological characteristics and the historical data of power load of the entire network corresponding to the meteorological characteristics;

[0014] Evaluating the accuracy of the power load prediction value of the power load prediction model according to the actual power load value to obtain an evaluation value of the power load prediction model;

[0015] The power load forecasting model is optimized according to the evaluation value until the evaluation value reaches the evaluation standard.

[0016] In the above technical solution, preferably, calculating the influence of rainfall and temperature on power generation load and power consumption load to obtain the actual value of power load includes:

[0017] Determine the day Y on which the difference between the historical data of the power load of the entire network and the actual data of the power load of the entire network is within ±20MW according to the rainfall threshold, and take the second day Y+1 of the day Y as the basic load of the entire society;

[0018] The similarity between the social base load and the small power load is calculated according to the dynamic time warping algorithm. If the similarity for three hours is within ±30MW, the small power load on the second day Y+1 is the actual small power load, and the grid supply load predicted = the grid supply load on the second day Y+1 is the actual grid supply load;

[0019] According to the small power generation load = the actual load of the whole society - the actual load of the grid supply, according to the basic load of the whole society - the actual load of small power, find the historical small power generation load Z with a difference of ±30MW, Z is a similar rainfall day, calculate the load growth, load growth = the basic load of the whole society - the actual load of the whole society - the actual load of the whole society on the second day of the similar rainfall day Z, predict the grid supply load = the basic load of the whole society - the actual load of the grid supply + the load growth, predict the total social load on the day = the actual value of the small power load + the actual value of the grid supply;

[0020] Calculate the small power load according to the power generation coefficient: the power generation coefficient for rainfall of 15-35 is 0.25; the power generation coefficient for rainfall of 15-45 is 0.6; the power generation coefficient for rainfall above 45 is 1, the small power load increment = 500 (installed load) * power generation coefficient - the basic load of the whole society - the actual load of small power, at this time:

[0021] Grid load forecast value = total social base load - grid supply increment - small power load increment;

[0022] Small power load forecast value = total social base load - grid supply increment + small power load increment;

[0023] The actual value of the total social load = actual grid load + actual small power load.

[0024] In the above technical solution, preferably, calculating the influence of rainfall and temperature on power generation load and power load to obtain the actual value of power load also includes:

[0025] According to the preset minimum and maximum temperatures, the date with a temperature similar to that of day N+1 and its corresponding social load curve T are selected from the historical data of the power load of the entire network;

[0026] From the historical data of the power load of the whole network, select the date with similar temperature to that of day N-1 and its corresponding social load curve, and take the most recent date as the social load curve T-1;

[0027] Calculate the whole society growth load curve according to the whole society load curve T and the whole society load curve T-1;

[0028] Calculate the total social load curve on day N+1 according to the total social growth load curve, that is, the total social load curve on day N+1 = the total social load curve T of similar temperature on day N+1 + growth load curve;

[0029] Assuming that “small electricity” refers to a specific portion of electricity consumption, its predicted amount remains constant and does not change with temperature. Therefore, the total electricity supply demand is calculated based on the total social load curve on day N+1.

[0030] In the above technical solution, preferably, the total power supply demand is calculated according to the whole society load curve on day N+1, including:

[0031] Find the total social load curve T of similar temperature on day N+1, and then find the recent total social load curve of similar temperature and the total social load curve T-1 of similar temperature on day N-1. The growth load curve = the total social load curve on day N-1 - the total social load curve T-1, and the total social load curve on day N+1 = the total social load curve T of similar temperature on day N+1 + the growth load curve; the predicted amount of small electricity on day N+1 remains unchanged, and the grid supply load curve is calculated.

[0032] The technical solution of the second aspect of the present invention provides a system for improving the accuracy of power grid load prediction, comprising:

[0033] An acquisition module is configured to acquire the historical data of the power load of the entire network, the actual data of the power load of the entire network corresponding to 96 o'clock every day, and the meteorological data corresponding to the historical data of the power load of the entire network;

[0034] A load similarity calculation module is configured to calculate the load similarity of 96 points of each day in the historical valid data of the whole network power load and the day before the predicted day in the actual data of the whole network power load according to a dynamic time warping algorithm;

[0035] A similar load calculation module is configured to filter out similar historical loads from the load similarities according to a preset similarity threshold;

[0036] A judgment module is configured to determine whether the small power load is consistent according to the whole society load curve in similar historical loads;

[0037] If so, the grid supply load is consistent;

[0038] If not, calculate the impact of rainfall and temperature on power generation load and power consumption load to obtain the actual value of power load;

[0039] A meteorological feature determination module is configured to determine meteorological features affecting power load based on the historical data of power load of the entire network and the meteorological data;

[0040] A power load model generation module is configured to generate a power load prediction model according to the meteorological characteristics and the historical data of the power load of the entire network corresponding to the meteorological characteristics;

[0041] A model evaluation module is configured to evaluate the accuracy of the power load prediction value of the power load prediction model according to the actual power load value, and obtain an evaluation value of the power load prediction model;

[0042] The model optimization module is configured to optimize the power load prediction model according to the evaluation value until the evaluation value reaches the evaluation standard.

[0043] The technical solution of the third aspect of the present invention provides a readable storage medium having a computer program stored thereon. When the program is executed by a processor, the steps of the method for improving the accuracy of power grid load prediction provided by the technical solution of the first aspect are implemented.

[0044] The technical solution of the fourth aspect of the present invention provides a computer device, including a storage medium and a processor; the storage medium is used to store a computer program; the processor is used to execute the computer program to implement the steps of the method for improving the accuracy of power grid load prediction provided by the technical solution of the first aspect above.

[0045] Compared with the prior art, the advantages of the method and device provided by the present invention for improving the accuracy of load prediction of the power grid are as follows: (1) The factors affecting the load prediction are judged subjectively by the forecasters, which are greatly affected by the subjective factors of the forecasters and result in large differences in the prediction accuracy. The software automatically calculates the 96-point load values ​​of the power grid load, quantifies the influencing factors, and reduces the subjective prediction of the forecasters, resulting in low prediction accuracy. The load data of the Honghe power grid has not yet achieved typical data records, and the forecasters rely entirely on their memory to find the load prediction data of similar influencing factors for reference. Due to the long time and large amount of data, the reference data found is inaccurate, which affects the load prediction accuracy. By using the storage and memory function of the software, the reference values ​​of similar influencing factors can be searched out to improve the prediction accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0047] Figure 1 A flowchart of a method according to an embodiment of the present invention is shown;

[0048] Figure 2 The structure block diagram of the system involved in the embodiment of the present invention is shown. DETAILED DESCRIPTION

[0049] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0050] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0051] Definition

[0052] Total social load: electricity consumption in a certain area, which is affected by temperature, rainfall, etc. and affects users' electricity demand.

[0053] Small power load: power generation of small hydropower stations

[0054] Grid load: The power supply of the power grid equipment of the power supply bureau in the region. The electricity demand of the entire region is first met by small power plants, and the insufficient power supply is provided by the power grid of the power supply bureau. Therefore, the grid load depends on the electricity demand of the whole society and the power generation capacity of small power plants. Small power plants are generated by hydropower stations, and the power generation is affected by temperature and rainfall.

[0055] The following various thresholds are configurable values, and the present invention does not limit the specific values.

[0056] like Figure 1 As shown, according to an embodiment of the present invention, a method for improving the accuracy of power grid load prediction includes:

[0057] S1, obtaining the historical data of the power load of the entire network, the actual data of the power load of the entire network corresponding to 96 points per day (one point every 15 minutes), and the meteorological data corresponding to the historical data of the power load of the entire network;

[0058] In this step, the historical data of power load of the whole network, meteorological data and actual data of power load of the whole network are all cleaned by removing outliers, processing missing data and other data cleaning processes to ensure the integrity and accuracy of the data; and normalization processing is performed: the load data and meteorological data are normalized or standardized to eliminate the impact of different dimensions and unit differences on model training.

[0059] S2, calculating the load similarity of each day in the historical valid data of the whole network power load and 96 points of the day before the predicted day in the actual data of the whole network power load according to the dynamic time warping algorithm;

[0060] In this step, select the similarity index: Common similarity metrics include Euclidean distance, Manhattan distance, cosine similarity, dynamic time warping (DTW), etc. For time series data, DTW is usually a good choice because it can handle the expansion and curvature of data on the time axis. Here, the dynamic time warping (DTW) algorithm is used to find similar historical load curves.

[0061] Dynamic Time Warping (DTW) is an algorithm used to measure the similarity between two (possibly unequal) time series. It is widely used in speech recognition, handwriting recognition, and time series analysis. In power load curve analysis, DTW can be used to compare load curves in different time periods to find the most similar patterns.

[0062] Assume there are two load curves A and B, with lengths n and m respectively. They can be expressed as:

[0063] A=a1,a2,...,a n

[0064] B=b1,b2,...,b m

[0065] Where a and b are the load values ​​of A and B at specific moments respectively.

[0066] (2) Initialize the distance matrix

[0067] Create a (+1)×(+1)(n+1)×(m+1) distance matrix d, where dij represents the distance from the i-th point in A to the j-th point in B. Euclidean distance is usually used:

[0068]

[0069] (3) Calculate the cumulative distance matrix

[0070] Define a cumulative distance matrix D, whose size is (+1)×(+1)(n+1)×(m+1). The initial conditions are:

[0071] D(i,j)=∞for all i∈[1,n],j∈[1,m];

[0072] The boundary conditions are:

[0073] D(0,j)=D(i,0)=∞for all i∈[1,n],j∈[1,m]

[0074] D(0,0)=0;

[0075] (4) Recursive formula

[0076] Use dynamic programming to fill in the cumulative distance matrix:

[0077] D(i,j)=d(a i ,b j )+min{D(i-1,j),D(i,j-1),D(i-1,j-1)};

[0078] (5) Calculate the minimum distance

[0079] The final DTW distance is the value in the lower right corner of the cumulative distance matrix:

[0080] DTW(A,B)=D(n,m).

[0081] S3, filtering out similar historical loads from the load similarities according to a preset similarity threshold;

[0082] S4, determining whether the small power load is consistent based on the whole society load curve in similar historical loads;

[0083] If so, the grid supply load is consistent;

[0084] If not, calculate the impact of rainfall and temperature on power generation load and power consumption load to obtain the actual value of power load;

[0085] S5, determining meteorological characteristics that affect the power load according to the historical data of the power load of the entire network and the meteorological data;

[0086] S6, generating a power load prediction model according to the meteorological characteristics and the historical data of power load of the entire network corresponding to the meteorological characteristics;

[0087] S7, evaluating the accuracy of the power load prediction value of the power load prediction model according to the actual power load value, and obtaining an evaluation value of the power load prediction model;

[0088] S8, optimizing the power load forecasting model according to the evaluation value until the evaluation value reaches the evaluation standard.

[0089] In the above technical solution, preferably, calculating the influence of rainfall and temperature on power generation load and power consumption load to obtain the actual value of power load includes:

[0090] Determine the day Y on which the difference between the historical data of the power load of the entire network and the actual data of the power load of the entire network is within ±20MW according to the rainfall threshold, and take the second day Y+1 of the day Y as the basic load of the entire society;

[0091] The similarity between the social base load and the small power load is calculated according to the dynamic time warping algorithm. If the similarity for three hours is within ±30MW, the small power load on the second day Y+1 is the actual small power load, and the grid supply load predicted = the grid supply load on the second day Y+1 is the actual grid supply load;

[0092] According to the small power generation load = the actual load of the whole society - the actual load of the grid supply, according to the basic load of the whole society - the actual load of small power, find the historical small power generation load Z with a difference of ±30MW, Z is a similar rainfall day, calculate the load growth, load growth = the basic load of the whole society - the actual load of the whole society - the actual load of the whole society on the second day of the similar rainfall day Z, predict the grid supply load = the basic load of the whole society - the actual load of the grid supply + the load growth, predict the total social load on the day = the actual value of the small power load + the actual value of the grid supply;

[0093] Calculate the small power load according to the power generation coefficient: the power generation coefficient for rainfall of 15-35 is 0.25; the power generation coefficient for rainfall of 15-45 is 0.6; the power generation coefficient for rainfall above 45 is 1, the small power load increment = 500 (installed load) * power generation coefficient - the basic load of the whole society - the actual load of small power, at this time:

[0094] Grid load forecast value = total social base load - grid supply increment - small power load increment;

[0095] Small power load forecast value = total social base load - grid supply increment + small power load increment;

[0096] The actual value of the total social load = actual grid load + actual small power load.

[0097] In the above technical solution, preferably, calculating the influence of rainfall and temperature on power generation load and power load to obtain the actual value of power load also includes:

[0098] According to the preset minimum and maximum temperatures, the date with a temperature similar to that of day N+1 and its corresponding social load curve T are selected from the historical data of the power load of the entire network;

[0099] From the historical data of the power load of the whole network, select the date with similar temperature to that of day N-1 and its corresponding social load curve, and take the most recent date as the social load curve T-1;

[0100] Calculate the whole society growth load curve according to the whole society load curve T and the whole society load curve T-1;

[0101] Calculate the total social load curve on day N+1 according to the total social growth load curve, that is, the total social load curve on day N+1 = the total social load curve T of similar temperature on day N+1 + growth load curve;

[0102] Assuming that “small electricity” refers to a specific portion of electricity consumption, its predicted amount remains constant and does not change with temperature. Therefore, the total electricity supply demand is calculated based on the total social load curve on day N+1.

[0103] In the above technical solution, preferably, the total power supply demand is calculated according to the whole society load curve on day N+1, including:

[0104] Find the total social load curve T of similar temperature on day N+1, and then find the recent total social load curve of similar temperature and the total social load curve T-1 of similar temperature on day N-1. The growth load curve = the total social load curve on day N-1 - the total social load curve T-1, and the total social load curve on day N+1 = the total social load curve T of similar temperature on day N+1 + the growth load curve; the predicted amount of small electricity on day N+1 remains unchanged, and the grid supply load curve is calculated.

[0105] In this embodiment, the computing network provides:

[0106] Based on the above calculated social load curve for day N+1, combined with the unchanged small power forecast, the total power supply demand is calculated.

[0107] 3.2 Maximum temperature below 20℃ below 15℃ below 10℃

[0108] When today's maximum temperature is lower than 20℃ by more than 10 degrees, the curve from 0:00 to 9:30 remains unchanged from the N-day curve.

[0109] When the temperature drops by 5-8℃ (today's highest temperature - tomorrow's highest temperature), the load at 12:00 increases by 150MW, the load at 14:30 increases by 90MW, the load from 17:15 to 22:30 increases by 180MW, and the curve load from 22:30 to 24:00 increases by 90MW

[0110] When the temperature drops by 8℃-12℃, the curve from 0 o'clock to 9:30 remains unchanged from the curve of day N, the load increases by 230MW at 12 o'clock, 160MW at 14:30, 260MW from 17:15 to 22:30, and 160MW from 22:30 to 24:00.

[0111] The load curve for the day can be derived using the values ​​at 0 o'clock, 5 o'clock, 9:30, 12 o'clock, 14:30, 17:15, 20 o'clock, and 24 o'clock.

[0112] When the maximum temperature on days N-1, N, and N+1 is lower than 10℃, the load curve on day N+1 is consistent with that on days N-1 and N.

[0113] In order to realize this temperature and load curve adjustment system, we need to design a program that can receive the daily maximum temperature data and adjust the load curve according to the temperature changes and set conditions. The following are the detailed steps and principles for implementing this process:

[0114] (1) Data preparation

[0115] Historical data: stores the daily maximum temperatures on day N-1, day N, and in the future (such as day N+1).

[0116] Basic load curve: load data at each time point in N days (for example, 0 o'clock, 5 o'clock, 9:30, etc.).

[0117] (2) Temperature condition judgment

[0118] Maximum temperature is lower than 20℃ and higher than 10℃: If the maximum temperature on day N+1 is lower than 20℃ and higher than 10℃, perform the following steps.

[0119] Determination of the temperature drop: Calculate the temperature difference between day N and day N+1 (ΔT=TN-TN+1), and adjust the load according to the temperature difference.

[0120] (3) Load adjustment logic

[0121] Cool down by 5-8℃, 0:00 to 9:30: the load remains unchanged at the N-day level.

[0122] 12 o'clock: Load increased by 150MW.

[0123] 14:30: Load increased by 90MW.

[0124] 17:15 to 22:30: Load increased by 180MW.

[0125] 22:30 to 24:00: Load increased by 90MW.

[0126] Temperature drops by 8-12℃, 0:00 to 9:30: The load remains unchanged at the N-day level.

[0127] 12 o'clock: Load increased by 230MW.

[0128] 14:30: Load increased by 160MW.

[0129] 17:15 to 22:30: Load increased by 260MW.

[0130] 22:30 to 24:00: Load increased by 160MW.

[0131] The temperature is below 10℃. If the maximum temperatures on day N-1, day N and day N+1 are all below 10℃, the load curve on day N+1 will directly adopt the load curve on day N-1 or day N.

[0132] like Figure 2 As shown, according to another embodiment of the present invention, a system 100 for improving the accuracy of power grid load prediction includes:

[0133] The acquisition module 10 is configured to acquire the historical data of the power load of the entire network, the actual data of the power load of the entire network corresponding to 96 o'clock every day, and the meteorological data corresponding to the historical data of the power load of the entire network;

[0134] The load similarity calculation module 20 is configured to calculate the load similarity of 96 points of each day in the historical valid data of the whole network power load and the day before the predicted day in the actual data of the whole network power load according to the dynamic time warping algorithm;

[0135] A similar load calculation module 30 is configured to filter out similar historical loads from the load similarities according to a preset similarity threshold;

[0136] The judgment module 40 is configured to determine whether the small electric load is consistent according to the whole society load curve in the similar historical load;

[0137] If so, the grid supply load is consistent;

[0138] If not, calculate the impact of rainfall and temperature on power generation load and power consumption load to obtain the actual value of power load;

[0139] A meteorological characteristic determination module 50 is configured to determine meteorological characteristics affecting the power load according to the historical data of the power load of the entire network and the meteorological data;

[0140] The power load model generation module 60 is configured to generate a power load prediction model according to the meteorological characteristics and the historical data of the power load of the entire network corresponding to the meteorological characteristics;

[0141] A model evaluation module 70 is configured to evaluate the accuracy of the power load prediction value of the power load prediction model according to the actual power load value, and obtain an evaluation value of the power load prediction model;

[0142] The model optimization module 80 is configured to optimize the power load prediction model according to the evaluation value until the evaluation value reaches the evaluation standard.

[0143] Based on the above Figure 1 The method shown, accordingly, an embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps of the method for improving the accuracy of power grid load prediction in any of the above embodiments are implemented.

[0144] Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, USB flash drive, mobile hard disk, etc.), including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various implementation scenarios of the present application.

[0145] Based on the above Figure 1 The method shown, and Figure 2 The virtual device embodiment shown, in order to achieve the above-mentioned purpose, the embodiment of the present application also provides a computer device, including a storage medium and a processor; the storage medium is used to store computer programs; the processor is used to execute the computer program to implement the steps of any of the above-mentioned embodiments of the method for improving the accuracy of power grid load prediction.

[0146] Optionally, the computer device may further include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a WI-FI module, etc. The user interface may include a display, an input unit such as a keyboard, etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (such as a Bluetooth interface, a WI-FI interface), etc.

[0147] Those skilled in the art will appreciate that the computer device structure provided in this embodiment does not limit the computer device, and may include more or fewer components, or a combination of certain components, or different component arrangements.

[0148] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages and saves the hardware and software resources of the computer device, and supports the operation of information processing programs and other software and / or programs. The network communication module is used to realize communication between the components inside the storage medium, and communication with other hardware and software in the physical device.

[0149] In the present invention, the terms "first", "second", and "third" are used for descriptive purposes only and should not be understood as indicating or implying relative importance; the term "plurality" refers to two or more, unless otherwise clearly defined. The terms "installed", "connected", "connected", "fixed", etc. should be understood in a broad sense. For example, "connected" can be a fixed connection, a detachable connection, or an integral connection; "connected" can be a direct connection or an indirect connection through an intermediate medium. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0150] In the description of the present invention, it is necessary to understand that the directions or positional relationships indicated by terms such as “upper”, “lower”, “left”, “right”, “front” and “back” are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or unit referred to must have a specific direction, be constructed and operated in a specific orientation, and therefore, cannot be understood as a limitation on the present invention.

[0151] In the description of this specification, the description of the terms "one embodiment", "some embodiments", "specific embodiments", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0152] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for improving the accuracy of power grid load prediction, characterized in that: include: Obtaining the historical data of the power load of the entire network, the actual data of the power load of the entire network corresponding to 96 o'clock every day, and the meteorological data corresponding to the historical data of the power load of the entire network; Calculate the load similarity of each day in the historical effective data of the whole network power load and 96 points of the day before the predicted day in the actual data of the whole network power load according to the dynamic time warping algorithm; Filtering similar historical loads from the load similarities according to a preset similarity threshold; Determine whether the small power load is consistent based on the overall social load curve in similar historical loads; If so, the grid supply load is consistent; If not, calculate the impact of rainfall and temperature on power generation load and power consumption load to obtain the actual value of power load; Determine the meteorological characteristics that affect the power load according to the historical data of the power load of the entire network and the meteorological data; Generate a power load forecasting model according to the meteorological characteristics and the historical data of power load of the entire network corresponding to the meteorological characteristics; Evaluating the accuracy of the power load prediction value of the power load prediction model according to the actual power load value to obtain an evaluation value of the power load prediction model; The power load forecasting model is optimized according to the evaluation value until the evaluation value reaches the evaluation standard.

2. The method for improving the accuracy of power grid load prediction according to claim 1, characterized in that: Calculate the impact of rainfall and temperature on power generation load and power consumption load, and obtain the actual value of power load, including: Determine the day Y on which the difference between the historical data of the power load of the entire network and the actual data of the power load of the entire network is within ±20MW according to the rainfall threshold, and take the second day Y+1 of the day Y as the basic load of the entire society; The similarity between the basic load of the whole society and the small power load is calculated according to the dynamic time warping algorithm. If the similarity of three hours is within ±30MW, the small power load on the second day Y+1 is the actual small power load, and the grid supply load on the second day Y+1 is predicted to be the actual grid supply load; According to the small power generation load = the actual load of the whole society - the actual load of the grid supply, according to the basic load of the whole society - the actual load of small power, find the historical small power generation load Z with a difference of ±30MW, Z is a similar rainfall day, calculate the load growth, load growth = basic load of the whole society - actual load of the whole society - actual load of the whole society on the second day of similar rainfall day Z, predict the grid supply load = basic load of the whole society - actual load of the grid supply + load growth, predict the total social load on the day = actual value of small power load + actual value of grid supply; Calculate the small power load according to the power generation coefficient: the power generation coefficient for rainfall of 15-35 is 0.25; the power generation coefficient for rainfall of 15-45 is 0.6; the power generation coefficient for rainfall above 45 is 1, the small power load increment = 500 (installed load) * power generation coefficient - the basic load of the whole society - the actual load of small power, at this time: Grid load forecast value = total social base load - grid supply increment - small power load increment; Small power load forecast value = total social base load - grid supply increment + small power load increment; The actual value of the total social load = actual grid load + actual small power load.

3. The method for improving the accuracy of power grid load prediction according to claim 2, characterized in that: Calculate the impact of rainfall and temperature on power generation load and power consumption load, obtain the actual value of power load, and also include: According to the current date N, the preset minimum and maximum temperatures are selected from the historical data of the power load of the entire network to find the date with a temperature similar to that of N+1 (the second day of the current date) and its corresponding social load curve T; From the historical data of the power load of the whole network, select the date with similar temperature to N-1 (the day before the current date) and its corresponding social load curve, and take the most recent date as the social load curve T-1; Calculate the whole society growth load curve according to the whole society load curve T and the whole society load curve T-1; Calculate the total social load curve of day N+1 according to the total social growth load curve, that is, the total social load curve of day N+1 = the total social load curve T of similar temperature on day N+1 (the second day of the current date) + the growth load curve; Assuming that "small electricity" refers to a specific part of electricity consumption, its predicted amount remains unchanged and does not change with temperature. Therefore, the total electricity supply demand is calculated based on the total social load curve on day N+1.

4. The method for improving the accuracy of power grid load prediction according to claim 2, characterized in that: The total electricity supply demand is calculated based on the social load curve on day N+1, the day after the current date, including: Find the total social load curve T of similar temperature on day N+1, and then find the recent total social load curve of similar temperature and the total social load curve T-1 of similar temperature on day N-1. The growth load curve = the total social load curve on day N-1 - the total social load curve T-1, and the total social load curve on day N+1 = the total social load curve T of similar temperature on day N+1 + the growth load curve; the predicted amount of small electricity on day N+1 remains unchanged, and the grid supply load curve is calculated.

5. A system for improving the accuracy of power grid load prediction, characterized in that: include: An acquisition module is configured to acquire the historical data of the power load of the entire network, the actual data of the power load of the entire network corresponding to 96 o'clock every day, and the meteorological data corresponding to the historical data of the power load of the entire network; A load similarity calculation module is configured to calculate the load similarity of 96 points of each day in the historical valid data of the whole network power load and the day before the predicted day in the actual data of the whole network power load according to a dynamic time warping algorithm; A similar load calculation module is configured to filter out similar historical loads from the load similarities according to a preset similarity threshold; A judgment module is configured to determine whether the small power load is consistent according to the whole society load curve in similar historical loads; If so, the grid supply load is consistent; If not, calculate the impact of rainfall and temperature on power generation load and power consumption load to obtain the actual value of power load; A meteorological characteristic determination module is configured to determine meteorological characteristics affecting power load based on the historical data of power load of the entire network and the meteorological data; The power load model generation module is configured to generate a power load prediction model based on the meteorological characteristics and the historical data of the power load of the entire network corresponding to the meteorological characteristics.

6. The system for improving the accuracy of grid load prediction according to claim 5, characterized in that: It also includes a model evaluation module, which is configured to evaluate the accuracy of the power load prediction value of the power load prediction model according to the actual value of the power load, and obtain the evaluation value of the power load prediction model.

7. The system for improving the accuracy of power grid load prediction according to claim 5, characterized in that: It also includes a model optimization module, which is configured to optimize the power load prediction model according to the evaluation value until the evaluation value reaches the evaluation standard.

8. A readable storage medium, characterized in that: A computer program is stored thereon, and when the program is executed by a processor, the steps of the method for improving the accuracy of power grid load prediction as claimed in any one of claims 1 to 4 are implemented.

9. A computer device, characterized in that: It comprises a storage medium and a processor; the storage medium is used to store a computer program; the processor is used to execute the computer program to implement the steps of the method for improving the accuracy of power grid load prediction as described in any one of claims 1 to 4.

Images