Power demand prediction method and system based on big data analysis

Abstract

The application discloses a power demand prediction method and system based on big data analysis, relates to the technical field of power systems, and mainly aims to solve the problems of poor prediction accuracy and effectiveness of existing power demand, and save power transmission effectiveness. The method comprises the following steps: determining load information to be demanded and time period average use data corresponding to the load information, wherein the time period average use data comprises average values of power scheduling use in multiple time periods; based on the load information, a corresponding demand prediction model is called, and the time period average use data is processed based on the demand prediction model to determine power demand reference data of a target time period, wherein the demand prediction model is obtained by training a deep learning network based on multi-time period use sample data; a reference time period of the target time period is determined, and the power demand reference data is adjusted based on the power demand extreme value and the power demand average value of the reference time period to obtain a power demand prediction result.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power systems, in particular to a power demand prediction method and system based on big data analysis. BACKGROUND

[0002] In the power system, power demand is an index for power supply and power transmission, and accurate prediction can avoid waste of power energy. At present, the existing power demand prediction is usually based on historical use to obtain single operation, but the historical use cannot meet the change of sudden power demand, nor can it meet the flexibility demand of power demand in different scenarios, thereby greatly reducing the prediction accuracy and effectiveness of power demand. SUMMARY

[0003] Therefore, the present application provides a power demand prediction method and system based on big data analysis, which mainly aims to solve the problem of poor prediction accuracy and effectiveness of existing power demand prediction to save power transmission effectiveness.

[0004] According to one aspect of the present application, a power demand prediction method based on big data analysis is provided, comprising:

[0005] determining the load information to be demanded, and the time period average use data corresponding to the load information, the time period average use data including the average value of power dispatch use in multiple time periods;

[0006] based on the load information, the corresponding demand prediction model is called, and the time period average use data is processed based on the demand prediction model to determine the power demand reference data of the target time period, the demand prediction model being obtained by model training of a deep learning network based on multi-time period use sample data;

[0007] determining the reference time period in the target time period, and adjusting the power demand reference data based on the power demand extreme value and the power demand average value of the reference time period to obtain the power demand prediction result.

[0008] Further, before the power demand prediction model corresponding to the load information is called based on the load information, and the time period average use data is processed based on the demand prediction model to determine the power demand reference data of the target time period, the method further comprises:

[0009] creating a deep learning network containing multiple neural network layers, and obtaining multi-time period use sample data;

[0010] According to different load information, matched sample data is extracted from the multi-period use sample data, and a deep learning network is trained based on the sample data to obtain a demand prediction model matched with different load information and completing model training, wherein the load information includes load voltage, load current, load power and load use time length.

[0011] Further, the extraction of matched sample data from the multi-period use sample data according to different load information includes:

[0012] According to multiple extreme values in the load voltage, the load current, the load power and the load use time length, corresponding target periods are queried from the multi-period use sample data;

[0013] If the target periods overlap, sample data is extracted according to adjacent periods of the target periods;

[0014] If the target periods do not overlap, and the time length between multiple target periods does not exceed a preset time length, sample data is extracted according to the target periods;

[0015] If the target periods do not overlap, and the time length between multiple target periods exceeds a preset time length, the target periods are adjusted according to the preset time length, and sample data is extracted according to the adjusted target periods.

[0016] Further, the training of the deep learning network based on the sample data to obtain a demand prediction model matched with different load information and completing model training includes:

[0017] The sample data is extracted according to a first proportion, and the deep learning network is trained using the extracted first sample data, wherein the weight values of multiple neural network layers in the deep learning network are configured based on the number of periods of the extracted first sample data;

[0018] The sample data is extracted according to a second proportion, and the model loss value of the deep learning network after model training is determined using the extracted second sample data;

[0019] If the model loss value is less than a preset loss threshold, the model training of the deep learning network is completed;

[0020] If the model loss value is greater than or equal to the preset loss threshold, the step of extracting the sample data according to the first proportion and training the deep learning network using the extracted first sample data is re-executed.

[0021] Further, the method further comprises:

[0022] In response to the time period adjustment request, determining a reference time period of the target time period according to time period adjustment requirement information carried in the time period adjustment request;

[0023] If the reference extreme value in the power demand reference data is greater than the power demand extreme value, and the reference average value of the reference time period is greater than the power demand average value, the power demand reference data is adjusted downward;

[0024] If the reference extreme value in the power demand reference data is less than or equal to the power demand extreme value, and the reference average value of the reference time period is less than or equal to the power demand average value, the power demand reference data is adjusted upward.

[0025] Further, the method further comprises:

[0026] Output the power demand prediction result;

[0027] In response to the power demand prediction confirmation instruction, driving the power storage device or the power generation device to transmit power resources according to the power demand prediction result.

[0028] Further, after the power demand reference data is adjusted based on the power demand extreme value and the power demand average value of the reference time period to obtain the power demand prediction result, the method further comprises:

[0029] Obtain power transmission loss information of a power system, and calculate transmission power data of different power transmission paths in the power system based on the power demand prediction result and the power transmission loss information, to generate a power transmission instruction request transmission power based on the transmission power data.

[0030] According to another aspect of the present application, a power demand prediction system based on big data analysis is provided, comprising:

[0031] A determination module is configured to determine load information to be demanded, and time period average use data corresponding to the load information, wherein the time period average use data comprises average values of power scheduling use in multiple time periods;

[0032] a processing module configured to determine a corresponding demand prediction model based on the load information, and process the period average use data based on the demand prediction model to determine power demand reference data of a target period, the demand prediction model being obtained by model training of a deep learning network based on multi-period use sample data;

[0033] an adjusting module configured to determine a reference period in the target period, and adjust the power demand reference data based on a power demand extreme value and a power demand average value of the reference period to obtain a power demand prediction result.

[0034] According to another aspect of the present application, a storage medium is provided, which stores at least one executable instruction, and the executable instruction causes a processor to perform operations corresponding to the power demand prediction method based on big data analysis.

[0035] According to still another aspect of the present application, a terminal is provided, which comprises a processor, a memory, a communication interface and a communication bus, and the processor, the memory and the communication interface complete communication with each other through the communication bus.

[0036] The memory is configured to store at least one executable instruction, and the executable instruction causes the processor to perform operations corresponding to the power demand prediction method based on big data analysis.

[0037] By means of the above technical solutions, the technical solutions provided by the embodiments of the present application have at least the following advantages:

[0038] The present application provides a power demand prediction method and system based on big data analysis. Compared with the prior art, the embodiments of the present application determine load information to be demanded and period average use data corresponding to the load information, the period average use data comprising average values of power dispatch use in multiple time periods; determine a corresponding demand prediction model based on the load information, and process the period average use data based on the demand prediction model to determine power demand reference data of a target period, the demand prediction model being obtained by model training of a deep learning network based on multi-period use sample data; determine a reference period in the target period, and adjust the power demand reference data based on a power demand extreme value and a power demand average value of the reference period to obtain a power demand prediction result, thereby realizing prediction analysis based on power in different periods, using power in multiple periods as a prediction basis, greatly improving power demand prediction accuracy of different loads, and ensuring accuracy and effectiveness of power transmission.

[0039] The above description is only a summary of the technical solutions of the present application. In order to enable a more thorough understanding of the technical means of the present application, the content of the specification can be implemented, and in order to enable the above and other purposes, features and advantages of the present application to be more obvious and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0040] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of the preferred embodiments and are not meant to limit the present application. Moreover, the same reference numerals in the attached drawings refer to the same or similar components. In the drawings:

[0041] Figure 1 A flow chart of a power demand prediction method based on big data analysis provided by an embodiment of the present application is shown;

[0042] Figure 2 A deep learning network structure schematic diagram provided by an embodiment of the present application is shown;

[0043] Figure 3 A power demand display interface schematic diagram provided by an embodiment of the present application is shown;

[0044] Figure 4 A block diagram of a power demand prediction system based on big data analysis provided by an embodiment of the present application is shown;

[0045] Figure 5 A structure schematic diagram of a terminal provided by an embodiment of the present application is shown. DETAILED DESCRIPTION

[0046] Exemplary embodiments of the present disclosure will be described more fully hereinafter with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it is understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that the present disclosure can be more thoroughly understood, and the scope of the present disclosure can be accurately conveyed to those skilled in the art.

[0047] An embodiment of the present application provides a power demand prediction method based on big data analysis, as shown in the figure, the method comprises: Figure 1

[0048] 101, determine the load information to be demanded, and the time period average use data corresponding to the load information.

[0049] ​In the embodiment of the present application, the current execution subject can be a processing server of the power system or a processor terminal of the power system as an execution end for predicting the power demand, so as to obtain the load information to be demanded in real time. At this time, the load information is the specific information of the terminal device of the demanded power, including but not limited to different forms of power consumption devices, such as residential power consumption devices and industrial power consumption devices. The load information can be the time, voltage, current and total power of residential power consumption, such as the load voltage, load current, load power and load use time. The embodiment of the present application does not make specific limitation. At the same time, the current execution end obtains the time period average use data corresponding to the load information. At this time, the time period average use data includes the average value of power dispatching use in multiple time periods, that is, the average value obtained by dividing each terminal as a load according to the historical time period and dividing according to the time period, such as the average power used on the 16th day calculated every 2 hours, that is, 12 average power values are obtained as the time period average use data of the 16th day.

[0050] It should be noted that when obtaining the load information, the power operator can input based on the power system, or the default value calculated in the power system can be obtained. The embodiment of the present application does not make specific limitation.

[0051] 102, based on the load information, the corresponding demand prediction model is called, and the time period average use data is processed based on the demand prediction model to determine the power demand reference data of the target time period.

[0052] In the embodiment of the present application, the demand prediction model used by different load terminals is pre-trained in the current execution end, so as to realize the purpose of accurate prediction. Therefore, the current execution end can call the corresponding demand prediction model according to the load information, process the time period average use data based on the demand prediction model, and determine the power demand reference data of the target time period. The demand prediction model is obtained by training the deep learning network based on the multi-time period use sample data, so as to learn the reference demand condition of each time period from the multi-time period use sample data by using the deep learning network, and predict the power demand data of the target time period.

[0053] 103, determine the reference time period in the target time period, and adjust the power demand reference data based on the power demand extreme value and the power demand average value of the reference time period to obtain the power demand prediction result.

[0054] In the embodiment of the present application, after the current execution end obtains the power demand reference data of the target period, the reference period is determined based on the target period, at this time, the parameter period is the basis period for adjusting the power demand reference data, for example, the target period can be 1:00 to 2:00 in the morning of each day, and the corresponding reference period can be 0:00 to 3:00 in the morning of a day, or 2:00 to 3:00 in the morning of each day, etc., which is not limited in the embodiment of the present application. Further, the power demand extreme value and the power demand average value of the reference period are retrieved to adjust the power demand reference data, and the final power demand prediction result is obtained. At this time, the power demand extreme value is the maximum and minimum value of the power transmission recorded in the reference period, and the power demand average value is calculated based on the completed power transmission data recorded in the reference period, which can be retrieved from the historical database in the power system, which is not limited in the embodiment of the present application.

[0055] It should be noted that, in the embodiment of the present application, in order to make the power demand prediction more accurate, when adjusting the power demand reference data by using the power demand extreme value and the power demand average value, the reference data can be adjusted up or down according to the extreme value, and the reference data can also be adjusted up or down according to the average value, which is not limited in the embodiment of the present application.

[0056] In another embodiment of the present application, in order to further limit and illustrate, before the step of retrieving the corresponding demand prediction model based on the load information and processing the period average use data based on the demand prediction model to determine the power demand reference data of the target period, the method further comprises:

[0057] Creating a deep learning network comprising a plurality of neural network layers, and obtaining multi-period use sample data;

[0058] According to different load information, the matching sample data is extracted from the multi-period use sample data, and the deep learning network is trained based on the sample data to obtain the demand prediction model matched with different load information and completed model training.

[0059] In order to realize effective prediction of power demand based on artificial intelligence, the current execution end creates a deep learning network that needs to be learned in advance to learn based on multi-period use sample data. When creating the deep learning network, it is constructed based on a plurality of neural network layers, specifically a 3-layer neural network, such as Figure 2The deep learning network is shown, wherein the input can be the use of electricity extracted in multiple target periods, and the output is the demand electricity of the expected period (the expected period at this time can be an entire period), and w1, w2, and w3 are weights of each network level. At the same time, the current execution end obtains the data of the use of electricity in different periods in the historical record of the power system as multi-period use sample data, for example, the use of electricity (which can also be understood as the electricity transmitted by the power system) in the periods of 1 o'clock to 2 o'clock, 2 o'clock to 3 o'clock, and 3 o'clock to 4 o'clock is a, which is not limited in the embodiment of the application. Further, in order to adapt to different load terminals, the current execution end first extracts matching sample data from the multi-period use sample data based on different load information, so as to train the deep learning network based on the sample data and obtain a demand prediction model matched with different load information.

[0060] It should be noted that, since the load information is the use of electricity recorded by different load terminals, the load information includes load voltage, load current, load power, and load use time length, the load voltage is the voltage value when the load terminal uses electricity, the load current is the current value when the load terminal uses electricity, the load power is the rated power when the load terminal uses electricity, and the load use time length is the total use time length of the load terminal, which can be obtained based on the historical record in the power system, and the embodiment of the application is not limited.

[0061] In another embodiment of the application, in order to further limit and illustrate, the step of extracting matching sample data from the multi-period use sample data according to different load information includes:

[0062] According to multiple extreme values in the load voltage, the load current, the load power, and the load use time length, the corresponding target period is queried from the multi-period use sample data;

[0063] If the target periods overlap, sample data is extracted according to adjacent periods of the target periods;

[0064] If the target periods do not overlap, and the time length between the multiple target periods does not exceed a preset time length, sample data is extracted according to the target periods;

[0065] If the target periods do not overlap, and the time length between the multiple target periods exceeds the preset time length, the target periods are adjusted according to the preset time length, and sample data is extracted according to the adjusted target periods.

[0066] In order to achieve the purpose of extracting samples from multi-period use sample data to improve the learning accuracy of artificial intelligence algorithm, thereby improving the accuracy of power demand prediction, the current execution end first determines multiple extreme values in the load voltage, load current, load power and load use time length in the load information, so as to find the corresponding target time from the multi-period use sample data according to the multiple extreme values. At this time, the target period is the period corresponding to the determined extreme value, which can be one or multiple. For example, the extreme value of the load voltage appears at 2:10, and the extreme value of the load current appears at 3:25. The corresponding target period is 2, which can be from 2:00 to 3:00 and from 3:00 to 4:00, and the embodiment of the present application is not limited.

[0067] In some embodiments, since the period has a time length, after multiple target periods are queried from the multi-period use sample data, it is first determined whether the multiple target periods overlap. If the target periods overlap, sample data is extracted according to the target period and the corresponding adjacent period, that is, sample data is not only extracted from the target period, but also extracted from the adjacent period. At this time, the adjacent period includes the previous period and the next period of the target period. For example, the target period is from 4:00 to 5:00, and the adjacent period includes the period from 2:00 to 3:00 and the period from 5:00 to 6:00, and the embodiment of the present application is not limited. If there is no overlap between the multiple target periods, it is further determined whether the time length between the multiple target periods exceeds a preset time length. If the time length between the multiple target periods does not exceed the preset time length, sample data is extracted according to the target period. At this time, the preset time length is used to limit the period of extracted sample data to avoid too long period, and therefore can be set to 1 day or 2 days, etc., and the embodiment of the present application is not limited. In addition, if there is no overlap between the multiple target periods and the time length between the multiple target periods exceeds the preset time length, in order to avoid redundancy of sample extraction, the target period is adjusted according to the preset time length, and sample data is extracted according to the adjusted target period. At this time, when the target period is adjusted using the preset time length, the multiple target periods can be directly truncated according to the preset time length to obtain the period of extracted sample data. For example, the preset time length is 1 day, and the target period with 1 day time length can be truncated from the first target period to obtain the target period of the final extracted sample data, and the embodiment of the present application is not limited.

[0068] It should be noted that the length of the period in the embodiment of the present application can be configured based on different prediction scenarios of power demand, including but not limited to 1 hour, 2 hours or 1 day, 2 days, etc., and the embodiment of the present application is not limited. In addition, when extracting sample data from the target period, random extraction or proportional extraction can be performed, and the embodiment of the present application is not limited.

[0069] In another embodiment of the present application, in order to further define and illustrate, the step of performing model training on the deep learning network based on the sample data comprises:

[0070] The sample data is extracted according to a first proportion, and the deep learning network is trained using the extracted first sample data, wherein the weight values of the plurality of neural network layers in the deep learning network are configured based on the number of time periods of the extracted first sample data;

[0071] The sample data is extracted according to a second proportion, and the model loss value of the deep learning network after model training is determined using the extracted second sample data;

[0072] If the model loss value is less than a preset loss threshold, the model training of the deep learning network is completed;

[0073] If the model loss value is greater than or equal to the preset loss threshold, the step of extracting the sample data according to the first proportion and training the deep learning network using the extracted first sample data is re-executed.

[0074] In order to achieve accurate training of the deep learning network, thereby adapting to different load power demand prediction scenarios, the current execution end extracts the sample data according to a first proportion when performing model training, and uses the extracted first sample data as training samples to train the deep learning network. At the same time, the sample data is extracted according to a second proportion, and the extracted second sample data is used as a validation sample to determine the model loss value of the deep learning network after model training. At this time, the loss value can be calculated based on a loss function, which includes but is not limited to L1 loss, mean square error, cross-entropy, etc., and the present embodiment is not limited to specific. Further, if the calculated model loss value is less than a preset loss threshold, the model training of the deep learning network is completed, and if the model loss value is greater than or equal to the preset loss threshold, the step of extracting the sample data according to the first proportion and training the deep learning network using the extracted first sample data is re-executed for the next iteration training. In addition, the weight values of the plurality of neural network layers in the deep learning network in the present embodiment are configured based on the number of time periods of the extracted first sample data. Specifically, first, the target time period of the extracted sample data is determined as the number of time periods, for example, there are 5 target time periods that can be used as the time period of the extracted sample data, so the number of time periods is 5. When configuring the weights of the deep learning network constructed by three layers of neural networks, they are configured as 1 / 5, 3 / 5, and 1 / 5, respectively. In the present embodiment, the weight values of the first layer and the last layer are set to 1 / a, where a is the number of time periods.

[0075] In another embodiment of the present application, in order to further define and illustrate, the step of determining a reference period of the target period and adjusting the power demand reference data based on the power demand extreme value and the power demand average value of the reference period to obtain the power demand prediction result comprises:

[0076] In response to the period adjustment request, determining a reference period of the target period according to the period adjustment demand information carried in the period adjustment request;

[0077] If the reference extreme value in the power demand reference data is greater than the power demand extreme value, and the reference average value of the reference period is greater than the power demand average value, then the power demand reference data is adjusted downward;

[0078] If the reference extreme value in the power demand reference data is less than or equal to the power demand extreme value, and the reference average value of the reference period is less than or equal to the power demand average value, then the power demand reference data is adjusted upward.

[0079] To meet the flexible requirements for the effectiveness of electricity demand forecasting, the current execution terminal, when determining the reference period and adjusting the reference data for electricity demand, first triggers a period adjustment request through the power system. Upon receiving the request, the execution terminal uses the period adjustment requirement information carried in the request to determine the reference period for the target period. This period adjustment requirement information characterizes the specific period during which the target period needs adjustment. For example, if the target period is 2:00-3:00, the required adjustment information could be 1:00-2:00 daily. Therefore, 1:00-2:00 daily is used as the target period, and 2:00-3:00 is used as the reference period. The reference average value for this period is then obtained. Furthermore, the execution terminal uses the reference extreme values ​​and the extreme values ​​of electricity demand in the reference data to make judgments. Simultaneously, it compares the reference average value of the reference period with the average value of electricity demand to ensure that the electricity demand value is not too high or too low during the forecasting process, thus achieving the goal of accurate forecasting. Correspondingly, if the reference extreme value in the electricity demand reference data is greater than the actual electricity demand extreme value, and the reference average value for the reference period is greater than the actual electricity demand average value, it indicates that the electricity demand reference data is too large. Therefore, the electricity demand reference data will be adjusted downwards. In this case, the downward adjustment can be made by percentage, i.e., a preset threshold of 10% is used as the downward adjustment limit. After a 10% reduction, the reference extreme value in the electricity demand reference data will be compared with the actual electricity demand extreme value again. If the reference extreme value in the electricity demand reference data is less than or equal to the actual electricity demand extreme value, and the reference average value for the reference period is less than or equal to the actual electricity demand average value, then the electricity demand reference data will be adjusted upwards. In this case, the upward adjustment can also be made by percentage, i.e., a preset threshold of 10% is used as the upward adjustment limit. After a 10% increase, the reference extreme value in the electricity demand reference data will be compared with the actual electricity demand extreme value again.

[0080] In another embodiment of the invention, for further definition and explanation, the steps further include:

[0081] Output the electricity demand forecast results;

[0082] In response to the power demand forecast confirmation command, the power storage device or power generation device is driven to transmit power resources according to the power demand forecast result.

[0083] To meet the security and accuracy requirements of different power transmission methods, after receiving the power demand forecast results, the current execution terminal outputs these results to the front-end interface of the power system for operators to view. Since the power demand forecast results are for a target time period, when outputting the power demand forecast results, if... Figure 3As shown, the corresponding power demand prediction results are output according to different time periods and displayed in the form of a rectangular diagram, which greatly improves the visualization effect of the power demand prediction results. Correspondingly, the current execution end configures a confirmation button based on the power system front-end interface, which is used as a confirmation trigger after the operation personnel checks the power demand prediction. Therefore, after the operation personnel checks, the current execution end can receive the triggered power demand prediction confirmation instruction, so that the current execution end drives the storage device or the power generation device to transmit the power resource according to the power demand prediction result.

[0084] In some embodiments, the current execution end as a server of the power system can transmit the electric quantity to the load terminal through the connected power storage device such as an industrial battery or an accumulator, and can also transmit the electric quantity to the load terminal through the connected power generation device, which is not limited in the embodiments of the present application.

[0085] In another embodiment of the present application, in order to further limit and illustrate, after the step of adjusting the power demand reference data based on the power demand extreme value and the power demand average value of the reference period to obtain the power demand prediction result, the method further comprises:

[0086] Obtain the power transmission loss information of the power system, and calculate the transmission electric quantity data of different power transmission paths in the power system based on the power demand prediction result and the power transmission loss information, to generate a power transmission instruction request transmission electric quantity based on the transmission electric quantity data.

[0087] In order to avoid the situation that the electric quantity loss generated in the power transmission process reduces the electric quantity finally received by the load terminal, and improve the prediction accuracy of the power demand, the current execution end first obtains the power transmission loss information of the power system after obtaining the power demand prediction result. The power transmission loss information is the line loss electric quantity of different transmission routes in the power system, that is, the material and transmission distance of different lines will cause a certain electric quantity to disappear, so the specific electric quantity lost between the transmission electric quantity end and the load end can be calculated, and then the transmission electric quantity data of different power transmission paths in the power system is calculated based on the power demand prediction result and the power transmission loss information. In an embodiment scenario, the calculation of the transmission electric quantity data can be based on the transmission electric quantity formula, which is represented as: Wherein, δ is the power transmission loss value of the i th transmission path, q i is the power demand prediction result of the i th transmission path, and h is all the transmission paths.

[0088] The embodiment of the present application provides a power demand prediction method based on big data analysis, and the embodiment of the present application determines load information to be demanded and time period average use data corresponding to the load information, the time period average use data comprising average values of power scheduling use in multiple time periods; a corresponding demand prediction model is called based on the load information, and the time period average use data is processed based on the demand prediction model to determine power demand reference data of a target time period, the demand prediction model being obtained by model training of a deep learning network based on multiple time period use sample data; a reference time period of the target time period is determined, and the power demand reference data is adjusted based on a power demand extreme value and a power demand average value of the reference time period to obtain a power demand prediction result, so that prediction analysis is realized based on power of different time periods, and the power of multiple time periods is used as a prediction basis, the power demand prediction accuracy of different loads is greatly improved, and therefore the accuracy and effectiveness of power transmission are ensured.

[0089] Further, as an implementation of the above-mentioned Figure 1 method, the embodiment of the present application provides a power demand prediction system based on big data analysis, as shown in Figure 4 , the system comprises:

[0090] A determination module 21 is configured to determine load information to be demanded and time period average use data corresponding to the load information, the time period average use data comprising average values of power scheduling use in multiple time periods.

[0091] A processing module 22 is configured to call a corresponding demand prediction model based on the load information, and process the time period average use data based on the demand prediction model to determine power demand reference data of a target time period, the demand prediction model being obtained by model training of a deep learning network based on multiple time period use sample data.

[0092] An adjustment module 23 is configured to determine a reference time period of the target time period, and adjust the power demand reference data based on a power demand extreme value and a power demand average value of the reference time period to obtain a power demand prediction result.

[0093] This invention provides a power demand forecasting system based on big data analysis. The system determines the load information to be demanded and the corresponding average usage data for different time periods. The average usage data includes the average power dispatch usage over multiple time periods. Based on the load information, a corresponding demand forecasting model is retrieved, and the average usage data for different time periods is processed using this model to determine reference power demand data for a target time period. This demand forecasting model is obtained by training a deep learning network using usage sample data from multiple time periods. A reference time period is determined for the target time period, and the reference power demand data is adjusted based on the extreme and average power demand values ​​of that reference time period to obtain the power demand forecasting result. This system enables predictive analysis based on power consumption in different time periods and utilizes power consumption across multiple time periods as a basis for prediction, significantly improving the accuracy of power demand forecasting for different loads, thereby ensuring the precision and effectiveness of power transmission.

[0094] According to one embodiment of the present invention, a storage medium is provided, the storage medium storing at least one executable instruction, the computer-executable instruction being able to execute the power demand forecasting method based on big data analysis in any of the above method embodiments.

[0095] Figure 5 The diagram shows a structural schematic of a terminal according to an embodiment of the present invention. The specific implementation of the terminal is not limited by the specific embodiments of the present invention.

[0096] like Figure 5 As shown, the terminal may include: a processor 302, a communications interface 304, a memory 306, and a communications bus 308.

[0097] The processor 302, communication interface 304, and memory 306 communicate with each other via communication bus 308.

[0098] Communication interface 304 is used to communicate with other network elements such as clients or other servers.

[0099] The processor 302 is used to execute program 310, specifically to execute the relevant steps in the above-described embodiment of the power demand forecasting method based on big data analysis.

[0100] Specifically, program 310 may include program code that includes computer operation instructions.

[0101] The processor 302 can be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to perform the embodiments of the present application. The terminal can include one or more processors of the same type, such as one or more CPUs, or processors of different types, such as one or more CPUs and one or more ASICs.

[0102] The memory 306 is configured to store a program 310. The memory 306 can include a high-speed RAM memory, and can further include a non-volatile memory, such as at least one disk memory.

[0103] The program 310 can be specifically configured to enable the processor 302 to perform the following operations:

[0104] determining load information to be required, and time period average usage data corresponding to the load information, the time period average usage data including average values of power scheduling usage in multiple time periods;

[0105] based on the load information, calling a corresponding demand prediction model, and based on the demand prediction model, processing the time period average usage data to determine power demand reference data of a target time period, the demand prediction model being obtained by model training of a deep learning network based on multi-time period usage sample data;

[0106] determining a reference time period in the target time period, and based on a power demand extreme value and a power demand average value of the reference time period, adjusting the power demand reference data to obtain a power demand prediction result.

[0107] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present application can be realized by a general computing system, which can be concentrated on a single computing system, or distributed on a network composed of multiple computing systems, and optionally, they can be realized by program codes executable by a computing system, so that they can be stored in a storage system and executed by a computing system, and in some cases, the steps shown or described can be executed in different order, or they can be manufactured into individual integrated circuit modules, or multiple modules or steps can be manufactured into a single integrated circuit module. Thus, the present application is not limited to any specific combination of hardware and software.

[0108] The above merely provides the preferred embodiments of the present application, and is not used to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the principles and technical scope of the present application shall fall into the scope of the present application.

Claims

1. A method for power demand forecasting based on big data analysis, characterized in that, The method comprises the following steps: determining load information to be required, and time period average use data corresponding to the load information, the time period average use data comprising average values of power scheduling use in multiple time periods; creating a deep learning network comprising multiple neural network layers, and obtaining multiple time period use sample data; extracting matched sample data from the multiple time period use sample data according to different load information, and performing model training on the deep learning network based on the sample data to obtain a demand prediction model matched to different load information and completing model training, the load information comprising load voltage, load current, load power and load use duration; based on the load information, calling a corresponding demand prediction model, and processing the time period average use data based on the demand prediction model to determine power demand reference data of a target time period, the demand prediction model being obtained by performing model training on the deep learning network based on the multiple time period use sample data; determining a reference time period in the target time period, and adjusting the power demand reference data based on a power demand extreme value and a power demand average value of the reference time period to obtain a power demand prediction result; wherein the determination of the reference time period in the target time period and the adjustment of the power demand reference data based on the power demand extreme value and the power demand average value of the reference time period to obtain the power demand prediction result comprises: in response to a time period adjustment request, determining the reference time period in the target time period according to time period adjustment requirement information carried in the time period adjustment request; if a reference extreme value in the power demand reference data is greater than the power demand extreme value, and a reference average value of the reference time period is greater than the power demand average value, then the power demand reference data is adjusted downward; if the reference extreme value in the power demand reference data is less than or equal to the power demand extreme value, and the reference average value of the reference time period is less than or equal to the power demand average value, then the power demand reference data is adjusted upward.

2. The method of claim 1, wherein, the extraction of matched sample data from the multiple time period use sample data according to different load information comprises: querying corresponding target time periods from the multiple time period use sample data according to multiple extreme values of the load voltage, the load current, the load power and the load use duration; if the target time periods overlap, then sample data is extracted according to adjacent time periods of the target time periods; if the target time periods do not overlap, and the duration between multiple target time periods does not exceed a preset time length, then sample data is extracted according to the target time periods; if the target time periods do not overlap, and the duration between multiple target time periods exceeds the preset time length, then the target time periods are adjusted according to the preset time length, and sample data is extracted according to the adjusted target time periods.

3. The method of claim 1, wherein, the model training of the deep learning network based on the sample data to obtain the demand prediction model matched to different load information and completing model training comprises: The sample data is extracted according to a first proportion, and the deep learning network is trained by using the extracted first sample data, and weight values of a plurality of neural network layers in the deep learning network are configured based on a number of time periods of the extracted first sample data; The sample data is extracted according to a second proportion, and the model loss value of the deep learning network after model training is determined by using the extracted second sample data; If the model loss value is less than a preset loss threshold, the model training of the deep learning network is completed; If the model loss value is greater than or equal to the preset loss threshold, the step of extracting the sample data according to the first proportion and training the deep learning network by using the extracted first sample data is re-executed.

4. The method of claim 1, wherein, The method further comprises: outputting the power demand prediction result; in response to the power demand prediction confirmation instruction, driving the power storage device or the power generation device to transmit power resources according to the power demand prediction result.

5. The method according to any one of claims 1 to 4, characterized in that, After the power demand reference data is adjusted based on the power demand extreme value and the power demand average value of the reference period to obtain the power demand prediction result, the method further comprises: obtaining power transmission loss information of a power system, and calculating transmission power data of different power transmission paths in the power system based on the power demand prediction result and the power transmission loss information, to generate a power transmission instruction request transmission power based on the transmission power data.

6. A power demand prediction system based on big data analytics characterized in that, comprises: a determination module configured to determine load information to be demanded and time period average use data corresponding to the load information, the time period average use data comprising average values of power scheduling use in a plurality of time periods; a processing module configured to call a demand prediction model corresponding to the load information based on the load information, and process the time period average use data based on the demand prediction model to determine power demand reference data of a target time period, the demand prediction model being obtained by model training of a deep learning network based on multi-time period use sample data; an adjustment module configured to determine a reference period of the target time period, and adjust the power demand reference data based on a power demand extreme value and a power demand average value of the reference period to obtain a power demand prediction result; the processing module is further configured to create a deep learning network comprising a plurality of neural network layers, and obtain multi-time period use sample data; extract matched sample data from the multi-time period use sample data according to different load information, and train the deep learning network based on the sample data to obtain a demand prediction model matched to different load information and completed model training, the load information comprising load voltage, load current, load power, and load use time length; The adjustment module is further configured to, in response to a time period adjustment request, determine a reference time period of the target time period according to time period adjustment requirement information carried in the time period adjustment request; if a reference extreme value in the power demand reference data is greater than the power demand extreme value and a reference average value of the reference time period is greater than the power demand average value, the power demand reference data is adjusted downward; and if the reference extreme value in the power demand reference data is less than or equal to the power demand extreme value and the reference average value of the reference time period is less than or equal to the power demand average value, the power demand reference data is adjusted upward.

7. A storage medium, the storage medium storing at least one executable instruction, the executable instruction causing a processor to perform operations corresponding to the power demand prediction method based on big data analysis according to any one of claims 1-5. A processor, a memory, a communication interface, and a communication bus, the processor, the memory, and the communication interface performing communication with each other through the communication bus; 8. A terminal comprising: The memory is configured to store at least one executable instruction, the executable instruction causing the processor to perform operations corresponding to the power demand prediction method based on big data analysis according to any one of claims 1-5. ​

Images