Power prediction method and system under extreme weather condition

Abstract

The invention provides a power prediction method and system under an extreme weather condition, relates to the technical field of power prediction, and aims to realize accurate power prediction under a freezing condition. Respectively acquiring a freezing influence weight of each type of freezing element data, wherein the greater the freezing influence weight is, the greater the influence of the corresponding freezing element data on freezing is; obtaining a freezing type prediction result according to the freezing influence weight; and performing power prediction according to the freezing type prediction result. The method has the advantages that the model prediction capability is improved, the freezing data is accurately acquired, and accurate extreme weather power prediction is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of power prediction, and in particular to a method and system for power prediction under extreme weather conditions. Background Art

[0002] Power forecasting is highly dependent on weather conditions. The current improvement in extreme weather forecasting capabilities mainly relies on the improvement of computer processing capabilities and the establishment and operation of a short-term numerical weather forecasting business model.

[0003] Based on the continuous establishment of new computer systems, the processing capacity of meteorological information is constantly improving, so that new numerical weather forecast models with shorter spatial resolution and time step can be put into operation, and atmospheric monitoring information has also increased rapidly. However, there is still a lack of power algorithm models for extreme weather backgrounds, and there is no sufficient amount of historical extreme weather event data to establish targeted algorithm models, and it is impossible to accurately warn of extreme weather events in advance and call the above-established power algorithm models.

[0004] Therefore, it is urgent to establish a forecasting method that can cope with extreme weather and realize accurate power forecasting under freezing conditions. Summary of the invention

[0005] The purpose of the present invention is to provide a power prediction method and system under extreme weather conditions, which can achieve accurate power prediction under freezing conditions.

[0006] The present invention is achieved through the following technical solutions:

[0007] The present invention first provides a power prediction method under extreme weather conditions, comprising the following steps:

[0008] Collect data on various freezing elements;

[0009] respectively obtaining a freezing influence weight of each freezing element data, wherein a larger freezing influence weight represents a greater freezing influence of the corresponding freezing element data;

[0010] Obtaining a freezing type prediction result according to the freezing influence weight;

[0011] Power prediction is performed according to the freezing type prediction result.

[0012] Preferably, the freezing element data includes humidity, pressure, wind speed and precipitation.

[0013] Preferably, the method for respectively obtaining the freezing influence weight of each type of the freezing element data is:

[0014] Determine the extent of the impact of each freezing element data on freezing based on historical data;

[0015] Assign freezing effects a weight based on how much they affect the ice.

[0016] Preferably, an empirical weight threshold is set, and if the freezing influence weight of the freezing element data is less than the empirical weight threshold, the freezing element data is eliminated, otherwise the freezing element data is used as the freezing element data for prediction.

[0017] Preferably, the freezing element data is periodically and continuously acquired through high-resolution numerical model forecasting.

[0018] Preferably, the method for obtaining the freezing type prediction result according to the freezing influence weight is:

[0019] Collecting a frozen element data group, wherein the frozen element data group includes frozen element data at multiple moments;

[0020] Inputting the freezing impact weight and the freezing element data group into a freezing classification prediction model;

[0021] The freezing classification prediction model outputs freezing type information.

[0022] Preferably, the freezing classification prediction models with different parameters are trained for different terrains, and each freezing classification prediction model is trained using historical data of the corresponding region.

[0023] Preferably, the method for performing power prediction according to the freezing type prediction result is:

[0024] Obtain conventional power prediction results P based on power curve reg ;

[0025] Obtain the power reduction rate prediction result p based on the power reduction rate prediction model d ;

[0026] Get the predicted power P rea :

[0027] P rea =P reg -p d .

[0028] Preferably, the method for obtaining the power reduction rate prediction result based on the power reduction rate prediction model is:

[0029] Collecting a freezing element data group and a freezing type prediction result group, wherein the freezing element data group includes freezing element data at multiple moments, and the freezing type prediction result group includes the freezing type prediction results at multiple moments, and the moments in the freezing element data group and the freezing type prediction result group correspond to each other one by one;

[0030] Inputting the freezing element data group and the freezing type prediction result group into the power reduction rate prediction model;

[0031] The power reduction rate prediction model outputs a predicted power reduction rate.

[0032] The present invention also provides a power prediction method system under extreme weather conditions, which is applied to any of the above power prediction methods under extreme weather conditions, including:

[0033] Data collection module, used to collect data of various freezing elements;

[0034] An importance judgment module is used to obtain the freezing influence weight of each freezing element data respectively, wherein the larger the freezing influence weight is, the greater the freezing influence of the corresponding freezing element data is;

[0035] A freezing prediction module, used for obtaining a freezing type prediction result according to the freezing influence weight;

[0036] A power prediction module is used to perform power prediction according to the freezing type prediction result.

[0037] The technical solution of the present invention has at least the following advantages and beneficial effects:

[0038] The present invention can predict freezing types based on freezing element data in all aspects, and accurately and timely obtain environmental freezing conditions;

[0039] The present invention can judge the influence degree of each type of freezing element data on freezing, and incorporate the influence degree of different freezing elements into model learning and consideration when making freezing predictions, thereby improving the prediction accuracy;

[0040] The present invention predicts the power drop based on the freezing data and calculates the actual predicted power in combination with the standard power obtained through the power curve, and has high reliability and accuracy.

[0041] When performing power forecasting, the present invention performs different processing for different regions, thereby improving the effectiveness and adaptability of the forecasting method for different scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 A schematic flow chart of a method for power prediction under extreme weather conditions provided in Example 1 of the present invention. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0044] Example 1

[0045] This embodiment provides a power prediction method under extreme weather conditions. Figure 1 , including the following steps:

[0046] Collect data on various freezing elements;

[0047] respectively obtaining a freezing influence weight of each freezing element data, wherein a larger freezing influence weight represents a greater freezing influence of the corresponding freezing element data;

[0048] Obtaining a freezing type prediction result according to the freezing influence weight;

[0049] Power prediction is performed according to the freezing type prediction result.

[0050] In this embodiment, the freezing element data preferably includes humidity, pressure, wind speed and precipitation.

[0051] As a preferred solution, the method for respectively obtaining the freezing influence weight of each freezing element data is:

[0052] Determine the extent of the impact of each freezing element data on freezing based on historical data;

[0053] Assign freezing effects a weight based on how much they affect the ice.

[0054] Furthermore, an empirical weight threshold may be set. If the freezing influence weight of the freezing element data is less than the empirical weight threshold, the freezing element data is eliminated. Otherwise, the freezing element data is used as the freezing element data for prediction.

[0055] On the other hand, it is preferred to obtain the freezing element data periodically and continuously through high-resolution numerical model forecasting.

[0056] In addition, the method for obtaining the freezing type prediction result according to the freezing influence weight is preferably:

[0057] Collecting a frozen element data group, wherein the frozen element data group includes frozen element data at multiple moments;

[0058] Inputting the freezing impact weight and the freezing element data group into a freezing classification prediction model;

[0059] The freezing classification prediction model outputs freezing type information.

[0060] As a preferred solution, it is necessary to train the freezing classification prediction models with different parameters for different terrains, and each freezing classification prediction model is trained using historical data of the corresponding region.

[0061] Finally, the method for performing power prediction based on the freezing type prediction result is:

[0062] Obtain conventional power prediction results P based on power curve reg ;

[0063] Obtain the power reduction rate prediction result p based on the power reduction rate prediction model d ;

[0064] Get the predicted power P rea :

[0065] P rea =P reg -p d .

[0066] Specifically, the method for obtaining the power reduction rate prediction result based on the power reduction rate prediction model is preferably:

[0067] Collecting a freezing element data group and a freezing type prediction result group, wherein the freezing element data group includes freezing element data at multiple moments, and the freezing type prediction result group includes the freezing type prediction results at multiple moments, and the moments in the freezing element data group and the freezing type prediction result group correspond to each other one by one;

[0068] Inputting the freezing element data group and the freezing type prediction result group into the power reduction rate prediction model;

[0069] The power reduction rate prediction model outputs a predicted power reduction rate.

[0070] By collecting data on a variety of freezing elements, this embodiment can fully understand all aspects of the freezing phenomenon and provide rich basic data support; then the influence weights of each freezing element are obtained respectively, so that the prediction model can focus on the factors that have the greatest impact on freezing, thereby improving the accuracy and reliability of the prediction, and then based on the evaluation of the influence weights, the freezing type can be more accurately predicted, the decision-making process can be optimized, and a more targeted basis can be provided for subsequent power predictions; guiding power predictions through freezing type prediction results can improve the accurate assessment of the impact on the power supply system, help take measures in advance, and reduce risks; and the scheme of this embodiment can flexibly adjust the prediction model according to the data changes of different freezing elements, enhance adaptability, and dynamically monitor and evaluate the freezing impact under different environmental conditions. Based on the power prediction of this embodiment, it is convenient to perform forward-looking operations, such as normal operation or shutdown protection. And the prediction models of this embodiment can all be trained based on existing models, such as recurrent neural networks and long short-term memory neural networks.

[0071] In particular, the following solutions can be adopted when obtaining weights:

[0072] The freezing degree data of multiple historical moments are acquired in time series and divided into multiple groups of freezing degree arrays, each group of the freezing degree arrays includes freezing degree data of n consecutive moments;

[0073] Respectively obtain the average value of each group of the freezing degree array;

[0074] Acquire a plurality of difference reference array pairs, wherein the difference reference array pairs include the freezing degree arrays at two consecutive moments, and the freezing degree difference of the freezing degree arrays at two consecutive moments is greater than a preset difference;

[0075] Respectively obtaining the freezing element data at the time corresponding to each freezing degree data in the difference reference array pair;

[0076] The freezing influence weight is obtained according to the freezing element data and the freezing degree difference.

[0077] The method for obtaining multiple difference reference array pairs is as follows:

[0078] Calculate the average value of the freezing degree array at the i-th moment and the average value of the freezing degree array at the i-1-th moment, 1<i≤n;

[0079] Obtaining a first difference absolute value, where the first difference absolute value is the absolute value of the difference between the average value of the freezing degree array at the i-th moment and the average value of the freezing degree array at the i-1-th moment;

[0080] It is determined whether the absolute value of the first difference is greater than a preset difference threshold. If so, the average value of the freezing degree array at the i-th moment and the freezing degree array at the i-1-th moment are used as the difference reference array pair.

[0081] Specifically, the method for obtaining the freezing influence weight according to the freezing element data and the freezing degree difference is:

[0082] Calculate the average value of the climate data of the frozen element data at the jth moment and the average value of the climate data of the frozen element data at the j-1th moment, j∈[2,n], j is the number of the difference reference array pair;

[0083] Obtaining a second difference absolute value, where the second difference absolute value is an absolute value of a difference between an average value of the climate data of the frozen element data at the j-th moment and an average value of the climate data of the frozen element data at the j-1-th moment;

[0084] Respectively acquiring a first sequence formed by the absolute values ​​of the first differences and a second sequence formed by the absolute values ​​of the second differences arranged in time series;

[0085] Get the correlation between the first sequence and the second sequence:

[0086]

[0087] Among them, ρ m is the correlation between the first sequence and the second sequence of the mth type of the frozen element data, M is the total number of types of the frozen element data, a 1，q is the qth absolute value of the first difference in the first sequence, is the average value of the first sequence, a 2，m，q is the qth absolute value of the second difference in the second sequence of the mth type of frozen element data, is the average value of the second sequence of the mth type of frozen element data, and Q is the total number of elements in the first sequence;

[0088] Get the freezing influence weight:

[0089]

[0090] Among them, α m is the freezing influence weight of the t-th type of the freezing element data, and t is the correlation between the first sequence and the second sequence of the t-th type of the freezing element data.

[0091] On this basis, unimportant factors can be eliminated by presetting the freezing influence weights.

[0092] Example 2

[0093] This embodiment provides a power prediction method system under extreme weather conditions, which is applied to any of the above power prediction methods under extreme weather conditions, including:

[0094] Data collection module, used to collect data of various freezing elements;

[0095] An importance judgment module is used to obtain the freezing influence weight of each freezing element data respectively, wherein the larger the freezing influence weight is, the greater the freezing influence of the corresponding freezing element data is;

[0096] A freezing prediction module, used for obtaining a freezing type prediction result according to the freezing influence weight;

[0097] A power prediction module is used to perform power prediction according to the freezing type prediction result.

[0098] The above are only preferred embodiments of the present invention and are 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 power prediction method under extreme weather conditions, characterized in that: The following steps are involved: Collect data on various freezing elements; respectively obtaining a freezing influence weight of each freezing element data, wherein a larger freezing influence weight represents a greater freezing influence of the corresponding freezing element data; Obtaining a freezing type prediction result according to the freezing influence weight; Power prediction is performed according to the freezing type prediction result.

2. The power prediction method under extreme weather conditions according to claim 1, characterized in that: The freezing element data includes humidity, pressure, wind speed and precipitation.

3. The power prediction method under extreme weather conditions according to claim 1, characterized in that: The method for respectively obtaining the freezing influence weight of each freezing element data is as follows: Determine the extent of the impact of each freezing element data on freezing based on historical data; Assign freezing effects a weight based on how much they affect the ice.

4. The power prediction method under extreme weather conditions according to claim 3, characterized in that: An empirical weight threshold is set. If the freezing influence weight of the frozen element data is less than the empirical weight threshold, the frozen element data is eliminated. Otherwise, the frozen element data is used as the frozen element data for prediction.

5. The power prediction method under extreme weather conditions according to claim 1, characterized in that: The freezing element data are obtained periodically and continuously through high-resolution numerical model forecasting.

6. The power prediction method under extreme weather conditions according to claim 1, characterized in that: The method for obtaining the freezing type prediction result according to the freezing influence weight is: Collecting a frozen element data group, wherein the frozen element data group includes frozen element data at multiple moments; Inputting the freezing impact weight and the freezing element data group into a freezing classification prediction model; The freezing classification prediction model outputs freezing type information.

7. The power prediction method under extreme weather conditions according to claim 6, characterized in that: The freezing classification prediction models with different parameters are trained for different terrains, and each freezing classification prediction model is trained using historical data of the corresponding region.

8. The power prediction method under extreme weather conditions according to claim 6, characterized in that: The method for performing power prediction according to the freezing type prediction result is: Obtain conventional power prediction results P based on power curve reg ; Obtain the power reduction rate prediction result p based on the power reduction rate prediction model d ; Get the predicted power P rea : P rea =P reg -p d 。 9. The power prediction method under extreme weather conditions according to claim 8, characterized in that: The method for obtaining the power reduction rate prediction result based on the power reduction rate prediction model is: Collecting a freezing element data group and a freezing type prediction result group, wherein the freezing element data group includes freezing element data at multiple moments, and the freezing type prediction result group includes the freezing type prediction results at multiple moments, and the moments in the freezing element data group and the freezing type prediction result group correspond to each other one by one; Inputting the freezing element data group and the freezing type prediction result group into the power reduction rate prediction model; The power reduction rate prediction model outputs a predicted power reduction rate.

10. A power prediction method system under extreme weather conditions, applied to the power prediction method under extreme weather conditions as claimed in any one of claims 1 to 9, characterized in that: include: Data collection module, used to collect data of various freezing elements; An importance judgment module is used to obtain the freezing influence weight of each freezing element data respectively, wherein the larger the freezing influence weight is, the greater the freezing influence of the corresponding freezing element data is; A freezing prediction module, used for obtaining a freezing type prediction result according to the freezing influence weight; A power prediction module is used to perform power prediction according to the freezing type prediction result.

Images