SARIMA model load delivery prediction method and device considering time factor

Abstract

The embodiment of the invention provides an SARIMA model load delivery prediction method and device considering a time factor, and the method comprises the steps: constructing crawler data, and obtaining power market related data based on the crawler data; pre-processing the power market related data based on a pre-processing rule to obtain pre-processed data; determining abnormal data through a preset algorithm based on the preprocessed data; determining target data based on the data smooth replacement abnormal data, and determining an exogenous variable based on the time information of the abnormal data; performing model training based on the target data, the exogenous variables and the initial SARIMA model to determine a target SARIMA model; and acquiring real-time data, and determining prediction data based on the real-time data and the target SARIMA model. The prediction model is established by combining the time factor and the time sequence model, so that the model prediction precision can be improved, electricity market traders are helped to capture market opportunities, and optimal allocation of electricity resources is realized as far as possible.

Description

Technical Field

[0001] The embodiments of this specification relate to the field of data processing technology, and in particular to a SARIMA model load transmission prediction method considering time factors. Background Art

[0002] In recent decades, the power system has played an increasingly important role. Provincial power systems must not only meet growing domestic electricity demand but also potentially transmit electricity to other provinces to optimize regional power resource allocation. Therefore, accurate forecasting of power load and transmission is crucial for ensuring stable power system operation, rationally planning power production and transmission, improving energy efficiency, and reducing operating costs.

[0003] In the past, power load and transmission forecasts primarily relied on simple statistical methods and empirical evidence. For example, trend extrapolation, proportional coefficient methods, and empirical methods based on historical power load and transmission data were used. These simple statistical and empirical methods have played a role in previous power load and transmission forecasts. However, these traditional methods have numerous limitations. First, they often struggle to accurately capture the complex dynamics and uncertainties inherent in power load and transmission data. Power load is influenced by a variety of factors, including seasonal variations, weather conditions, economic activity, holidays, and policy changes, exhibiting significant nonlinear and random characteristics. Traditional methods are limited in their ability to handle these complex factors, resulting in low forecast accuracy and inaccurate results. Second, with the continuous development of modern power trading markets and the increasing complexity of power systems, traditional methods are struggling to adapt to new circumstances and requirements.

[0004] With the rapid development of information technology, data analysis, and artificial intelligence, a range of approaches, such as time series and machine learning, have emerged, providing new ideas and methods for forecasting power load and transmission data. Time series analysis methods, such as the autoregressive moving average (ARMA) model and its extended autoregressive integrated moving average (ARIMA) model, can predict external power load and transmission over time by modeling historical data, effectively capturing trends and regular patterns in time series data. Machine learning algorithms, such as support vector machines (SVMs), random forests, and artificial neural networks (ANNs), have also demonstrated their strengths in forecasting. They are capable of processing large amounts of data and exploiting complex nonlinear relationships within the data for learning and prediction.

[0005] Existing forecasting methods have great potential, but they still face several challenges in practical application. First, data quality is a key factor. Existing data is highly fragmented, residing in provincial power trading centers and public accounts. The multi-source and inconsistent data limits our understanding of the overall market situation. Second, different regions and power systems may have different characteristics, necessitating the selection of appropriate models. There is a lack of practical models capable of forecasting power market load and power transmission for a specific region. Furthermore, changes in power market policies and time factors can also introduce significant uncertainty into power load and power transmission forecasts.

[0006] Therefore, a better solution is urgently needed. Summary of the Invention

[0007] In light of this, embodiments of this specification provide a SARIMA model load delivery forecasting method that considers time factors. One or more embodiments of this specification also relate to a SARIMA model load delivery forecasting apparatus that considers time factors, a computing device, a computer-readable storage medium, and a computer program to address technical deficiencies in the prior art.

[0008] According to a first aspect of an embodiment of this specification, a SARIMA model load delivery forecasting method considering time factors is provided, comprising:

[0009] Build crawler data and obtain electricity market related data based on crawler data;

[0010] Preprocessing the electricity market related data based on preprocessing rules to obtain preprocessed data;

[0011] Determine abnormal data through a preset algorithm based on preprocessed data;

[0012] Determine target data based on data smoothing to replace abnormal data, and determine exogenous variables based on the time information of abnormal data;

[0013] Perform model training based on target data, exogenous variables and initial SARIMA model to determine the target SARIMA model;

[0014] Acquire real-time data and determine forecast data based on the real-time data and the target SARIMA model.

[0015] In one possible implementation, data related to the electricity market is obtained based on crawler data, including:

[0016] Based on crawler data, load data and outbound data are obtained from the power trading center and network interface;

[0017] Determine electricity market-related data based on load data and export data.

[0018] In one possible implementation, the power market related data is preprocessed based on preprocessing rules to obtain preprocessed data, including:

[0019] Determine standardization processing rules and data processing rules based on preprocessing rules;

[0020] Perform unified field naming and type conversion on power market-related data based on standardized processing rules to determine standardized data;

[0021] Based on the data processing rules, the standardized data is processed for missing values ​​and data duplication is removed to determine the preprocessed data.

[0022] In one possible implementation, determining abnormal data using a preset algorithm based on preprocessed data includes:

[0023] The model is trained using the isolation forest algorithm and the local anomaly factor algorithm to determine the abnormal data processing model;

[0024] Abnormal data is determined from the preprocessed data based on the abnormal data processing model.

[0025] In one possible implementation, determining target data based on data smoothing to replace abnormal data, and determining exogenous variables based on time information of the abnormal data, includes:

[0026] determining an averaging window, and determining an average value based on the averaging window;

[0027] Abnormal data is replaced based on the average value and time information is determined, and exogenous variables are determined based on the time information.

[0028] In one possible implementation, model training is performed based on target data, exogenous variables, and an initial SARIMA model to determine a target SARIMA model, including:

[0029] Determine training data based on target data, exogenous variables and preset proportions;

[0030] Model training is performed based on the training data and the initial SARIMA model to determine the target SARIMA model.

[0031] In a possible implementation, the method further includes:

[0032] Evaluate the forecast data and determine the evaluation data;

[0033] The target SARIMA model is tuned based on the evaluation data.

[0034] According to a second aspect of an embodiment of this specification, a SARIMA model load transmission forecasting device considering time factors is provided, comprising:

[0035] a data acquisition module configured to construct crawler data and acquire power market related data based on the crawler data;

[0036] a data preprocessing module configured to preprocess the power market related data based on preprocessing rules to obtain preprocessed data;

[0037] an abnormality determination module, configured to determine abnormal data using a preset algorithm based on the preprocessed data;

[0038] an exogenous variable module configured to determine target data based on data smoothing to replace abnormal data, and to determine exogenous variables based on time information of the abnormal data;

[0039] A model training module is configured to perform model training based on target data, exogenous variables and the initial SARIMA model to determine a target SARIMA model;

[0040] The data prediction module is configured to obtain real-time data and determine predicted data based on the real-time data and a target SARIMA model.

[0041] According to a third aspect of an embodiment of this specification, a computing device is provided, including:

[0042] memory and processor;

[0043] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the SARIMA model load delivery prediction method considering time factors are implemented.

[0044] According to a fourth aspect of the embodiments of this specification, a computer-readable storage medium is provided, which stores computer-executable instructions. When the instructions are executed by a processor, the steps of the SARIMA model load delivery prediction method considering time factors are implemented.

[0045] According to a fifth aspect of the embodiments of this specification, a computer program is provided, wherein when the computer program is executed in a computer, the computer is caused to execute the steps of the above-mentioned SARIMA model load transmission prediction method considering time factors.

[0046] The embodiments of this specification provide a SARIMA model load delivery forecasting method and device that considers time factors. The SARIMA model load delivery forecasting method that considers time factors includes: constructing crawler data and obtaining power market-related data based on the crawler data; preprocessing the power market-related data based on preprocessing rules to obtain preprocessed data; determining abnormal data using a preset algorithm based on the preprocessed data; determining target data based on data smoothing and replacing abnormal data, and determining exogenous variables based on the time information of the abnormal data; training a model based on the target data, exogenous variables, and an initial SARIMA model to determine a target SARIMA model; obtaining real-time data, and determining predicted data based on the real-time data and the target SARIMA model. By combining the time factor with a time series model to establish a forecasting model, the model's forecasting accuracy can be improved, helping power market traders seize market opportunities and achieve optimal allocation of power resources as much as possible. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a flow chart of a SARIMA model load delivery prediction method considering time factors provided by an embodiment of this specification;

[0048] Figure 2a This is a first prediction diagram of a SARIMA model load delivery prediction method considering time factors provided by an embodiment of this specification;

[0049] Figure 2b This is a second prediction diagram of a SARIMA model load delivery prediction method considering time factors provided by an embodiment of this specification;

[0050] Figure 2c This is a third prediction diagram of a SARIMA model load delivery prediction method considering time factors provided by an embodiment of this specification;

[0051] Figure 3 This is a schematic diagram of the structure of a SARIMA model load delivery prediction device considering time factors provided by an embodiment of this specification;

[0052] Figure 4 This is a structural block diagram of a computing device provided by one embodiment of this specification. DETAILED DESCRIPTION

[0053] The following description sets forth many specific details to facilitate a thorough understanding of this specification. However, this specification can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the scope of this specification. Therefore, this specification is not limited to the specific implementations disclosed below.

[0054] The terms used in one or more embodiments of this specification are for the purpose of describing specific embodiments only and are not intended to limit one or more embodiments of this specification. The singular forms "a," "an," and "the" used in one or more embodiments of this specification and the appended claims are also intended to include plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.

[0055] It should be understood that although the terms first, second, etc. may be used to describe various information in one or more embodiments of this specification, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of this specification, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0056] In this specification, a SARIMA model load delivery prediction method considering time factors is provided. This specification also relates to a SARIMA model load delivery prediction device considering time factors, a computing device, and a computer-readable storage medium, which are described in detail one by one in the following embodiments.

[0057] See also Figure 1 , Figure 1 A flow chart of a SARIMA model load delivery prediction method considering time factors provided according to an embodiment of this specification is shown, which specifically includes the following steps.

[0058] Step 101: construct crawler data, and obtain electricity market related data based on the crawler data.

[0059] In one possible implementation, obtaining electricity market related data based on crawler data includes: obtaining load data and outbound data from a power trading center and a network interface based on the crawler data; and determining electricity market related data based on the load data and outbound data.

[0060] In actual application, a specific crawler program interface is constructed to obtain data related to the present invention on the power trading center and official accounts. A total of 30 months of load and delivery data are obtained from 2021-04-01 to 2024-06-30.

[0061] The embodiments of this specification can obtain relevant data from multiple sources such as power trading centers and public accounts, and integrate and process power market-related data. Through data capture strategies and data timing collection mechanisms, the latest power market-related data can be effectively obtained.

[0062] Step 102: Preprocess the electricity market related data based on the preprocessing rules to obtain preprocessed data.

[0063] In one possible implementation, power market related data is preprocessed based on preprocessing rules to obtain preprocessed data, including: determining standardized processing rules and data processing rules based on the preprocessing rules; performing unified field naming and type conversion on power market related data based on the standardized processing rules to determine standardized data; performing missing value processing and data deduplication on standardized data based on the data processing rules to determine preprocessed data.

[0064] In practical applications, the acquired data is standardized, such as unified field naming and type conversion, and missing value processing and data deduplication are performed during the standardization process.

[0065] Step 103: Determine abnormal data using a preset algorithm based on the preprocessed data.

[0066] In one possible implementation, abnormal data is determined based on preprocessed data using a preset algorithm, including: performing model training using an isolation forest algorithm and a local anomaly factor algorithm to determine an abnormal data processing model; and determining abnormal data from the preprocessed data based on the abnormal data processing model.

[0067] In practical applications, a small portion of data is clearly distinct from the rest. This data often occurs in May and November. This is due to distinct seasonal characteristics and specific holidays. These data can be treated as exogenous variables. A column of exogenous variable data is added, all set to 0. The model is trained using the Isolation Forest algorithm and the Local Anomaly Factor algorithm to identify these outliers that are clearly distinct from the rest.

[0068] Specifically, Isolation Forest gradually isolates data points into independent subspaces by randomly selecting features and dividing thresholds. Outliers are quickly isolated to shallower tree nodes because their feature values ​​deviate from the normal range.

[0069] The specific steps are as follows:

[0070] (1) For a data point x, the path length h(x) in a single isolation tree is defined as the number of edges from the root node to the leaf node.

[0071] (2) Abnormal score determination: The abnormal score S(x) is calculated as:

[0072]

[0073] Where E(h(x)) is the average path length of data point x in all trees; c(n) is a normalization constant related to the number of samples n.

[0074] (3) If S(x)≈1, it is considered an outlier; otherwise, it is considered normal data.

[0075] Local outlier factor: By comparing the local density of each data point with the density of its neighbors, it identifies outliers whose density is significantly lower than that of the surrounding area.

[0076] The calculation formula of local reachability density is as follows:

[0077]

[0078] where N k (x): the set of k nearest neighbors of x; dist(x,o) is the reachable distance from point x to point o;

[0079] The local outlier factor LOF is calculated as follows:

[0080]

[0081] If the value is greater than the neighbor threshold, it is considered an outlier.

[0082] If a data point is detected as an outlier by both the isolation forest algorithm and the local outlier factor algorithm, then the data point is considered to be an outlier.

[0083] Furthermore, this solution also includes a method for dynamically adjusting the abnormal data ratio, optimizing the ratio based on actual conditions and historical data. Specifically, a clustering algorithm can be used to classify data into normal and abnormal categories, and then the abnormal data ratio can be dynamically adjusted based on the clustering results.

[0084] This example analyzes the temporal impact of abnormal market data using the results of anomaly detection using isolation forests and anomalous local factors. It also analyzes the trend and seasonality of data through parameter estimation and diagnostic analysis using the SARIMA model. This more comprehensive analysis can help traders and decision makers better understand market dynamics and develop more scientific and rational trading strategies and decisions.

[0085] Step 104: Determine target data based on the data smoothing replacement of abnormal data, and determine exogenous variables based on time information of the abnormal data.

[0086] In one possible implementation, target data is determined based on replacing abnormal data with data smoothing, and exogenous variables are determined based on time information of the abnormal data, including: determining an average window, determining an average value based on the average window; replacing abnormal data based on the average value and determining time information, and determining the exogenous variables based on the time information.

[0087] In practical applications, these abnormal data are moved averaged, the average window size is selected as 5, and the average value is taken as the value of the abnormal data. At the same time, the value of the time exogenous variable data is set to 1.

[0088] The embodiments of this specification take into account the influence of exogenous variables such as time factors and specific holidays. In summer, the nights are short and the days are long, while in winter, the nights are long and the days are short. This phenomenon determines the temporary occurrence of abnormal load and external transmission data in some months. Considering the influence of the exogenous variable time can more accurately grasp the dynamic changes in the electricity market. At the same time, due to special holidays or climatic conditions in some months, abnormal load and external transmission data may also occur. By deeply analyzing the influence of the exogenous variable time, a more accurate basis can be provided for electricity trading decisions, the allocation of electricity resources can be optimized, the operating efficiency of the electricity market can be improved, and traders can better understand market trends, reduce trading risks, and achieve stable and sustainable development of the electricity market.

[0089] Step 105: Perform model training based on the target data, exogenous variables, and the initial SARIMA model to determine the target SARIMA model.

[0090] In one possible implementation, model training is performed based on target data, exogenous variables, and an initial SARIMA model to determine a target SARIMA model, including: determining training data based on the target data, the exogenous variables, and a preset ratio; and determining the target SARIMA model by model training based on the training data and the initial SARIMA model.

[0091] In practical applications, the processed data from 2021-04-01 to 2024-06-01 are normalized, and the SARIMA model based on the time series containing exogenous variables is trained and tested in a ratio of 9:1 to obtain the final learning model.

[0092] The embodiments of this specification combine machine learning algorithms with time series models for analysis, improving forecast accuracy. Machine learning algorithms are used to identify anomalies. By comprehensively considering data anomalies and time series characteristics, forecast accuracy and reliability can be improved, providing more precise decision support for power market traders and decision makers.

[0093] Step 106: Acquire real-time data, and determine predicted data based on the real-time data and the target SARIMA model.

[0094] In practical applications, the final learning model is used to predict market trends based on current market conditions. Figure 2a and Figure 2b , Figure 2a This is the data predicted for the next day, where the left side is for load and the right side is for delivery. Figure 2b To predict the data for the next 7 days, the left side is the surrounding load and the right side is the delivery. Figure 2c To predict the data for the next 15 days, the left side is the surrounding load and the right side is the delivery.

[0095] In a possible implementation, the method further includes: evaluating the predicted data to determine evaluation data; and adjusting the target SARIMA model based on the evaluation data.

[0096] In practical applications, the performance of the prediction model is continuously updated, and the latest data is fed back and continuously optimized to improve the long-term performance of the model.

[0097] The embodiments of this specification provide a SARIMA model load delivery forecasting method and device that considers time factors. The SARIMA model load delivery forecasting method that considers time factors includes: constructing crawler data and obtaining power market-related data based on the crawler data; preprocessing the power market-related data based on preprocessing rules to obtain preprocessed data; determining abnormal data using a preset algorithm based on the preprocessed data; determining target data based on data smoothing and replacing abnormal data, and determining exogenous variables based on the time information of the abnormal data; training a model based on the target data, exogenous variables, and an initial SARIMA model to determine a target SARIMA model; obtaining real-time data, and determining predicted data based on the real-time data and the target SARIMA model. By combining the time factor with a time series model to establish a forecasting model, the model's forecasting accuracy can be improved, helping power market traders seize market opportunities and achieve optimal allocation of power resources as much as possible.

[0098] Corresponding to the above method embodiment, this specification also provides an embodiment of a SARIMA model load delivery prediction device considering time factors. Figure 3 FIG. 1 shows a schematic diagram of a SARIMA model load delivery prediction device considering time factors provided by an embodiment of this specification. Figure 3 As shown, the device includes:

[0099] The data acquisition module 301 is configured to construct crawler data and acquire power market related data based on the crawler data;

[0100] The data preprocessing module 302 is configured to preprocess the power market related data based on the preprocessing rules to obtain preprocessed data;

[0101] An abnormality determination module 303 is configured to determine abnormal data based on the preprocessed data using a preset algorithm;

[0102] The exogenous variable module 304 is configured to determine target data based on data smoothing to replace abnormal data, and to determine exogenous variables based on time information of the abnormal data;

[0103] The model training module 305 is configured to perform model training based on the target data, the exogenous variables and the initial SARIMA model to determine the target SARIMA model;

[0104] The data prediction module 306 is configured to obtain real-time data and determine predicted data based on the real-time data and the target SARIMA model.

[0105] In one possible implementation, data related to the electricity market is obtained based on crawler data, including:

[0106] Based on crawler data, load data and outbound data are obtained from the power trading center and network interface;

[0107] Determine electricity market-related data based on load data and export data.

[0108] In one possible implementation, the power market related data is preprocessed based on preprocessing rules to obtain preprocessed data, including:

[0109] Determine standardization processing rules and data processing rules based on preprocessing rules;

[0110] Perform unified field naming and type conversion on power market-related data based on standardized processing rules to determine standardized data;

[0111] Based on the data processing rules, the standardized data is processed for missing values ​​and data duplication is removed to determine the preprocessed data.

[0112] In one possible implementation, determining abnormal data using a preset algorithm based on preprocessed data includes:

[0113] The model is trained using the isolation forest algorithm and the local anomaly factor algorithm to determine the abnormal data processing model;

[0114] Abnormal data is determined from the preprocessed data based on the abnormal data processing model.

[0115] In one possible implementation, determining target data based on data smoothing to replace abnormal data, and determining exogenous variables based on time information of the abnormal data, includes:

[0116] determining an averaging window, and determining an average value based on the averaging window;

[0117] Abnormal data is replaced based on the average value and time information is determined, and exogenous variables are determined based on the time information.

[0118] In one possible implementation, model training is performed based on target data, exogenous variables, and an initial SARIMA model to determine a target SARIMA model, including:

[0119] Determine training data based on target data, exogenous variables and preset proportions;

[0120] Model training is performed based on the training data and the initial SARIMA model to determine the target SARIMA model.

[0121] In a possible implementation, the method further includes:

[0122] Evaluate the forecast data and determine the evaluation data;

[0123] The target SARIMA model is tuned based on the evaluation data.

[0124] The embodiments of this specification provide a SARIMA model load delivery prediction method and device that considers time factors, wherein the SARIMA model load delivery prediction device that considers time factors includes: constructing crawler data, and obtaining power market-related data based on the crawler data; preprocessing the power market-related data based on preprocessing rules to obtain preprocessed data; determining abnormal data using a preset algorithm based on the preprocessed data; determining target data based on data smoothing and replacing abnormal data, and determining exogenous variables based on the time information of the abnormal data; training a model based on the target data, exogenous variables, and an initial SARIMA model to determine a target SARIMA model; obtaining real-time data, and determining predicted data based on the real-time data and the target SARIMA model. By combining the time factor with the time series model to establish a prediction model, the model's prediction accuracy can be improved, helping power market traders seize market opportunities and achieve optimal allocation of power resources as much as possible.

[0125] The above is a schematic diagram of a SARIMA model load delivery forecasting device that considers time factors in this embodiment. It should be noted that the technical solution of this SARIMA model load delivery forecasting device that considers time factors and the technical solution of the SARIMA model load delivery forecasting method that considers time factors are based on the same concept. For details not described in detail in the technical solution of the SARIMA model load delivery forecasting device that considers time factors, please refer to the description of the technical solution of the SARIMA model load delivery forecasting method that considers time factors.

[0126] Figure 4The block diagram of a computing device 400 according to one embodiment of the present disclosure is shown. Components of the computing device 400 include, but are not limited to, a memory 410 and a processor 420. The processor 420 is connected to the memory 410 via a bus 430, and a database 450 is used to store data.

[0127] The computing device 400 also includes an access device 440 that enables the computing device 400 to communicate via one or more networks 460. Examples of such networks include a public switched telephone network (PSTN), a local area network (LAN), a wide area network (WAN), a personal area network (PAN), or a combination of communication networks such as the Internet. The access device 440 may include one or more of any type of network interface (e.g., a network interface card (NIC)) whether wired or wireless, such as an IEEE 802.11 wireless local area network (WLAN) wireless interface, a Worldwide Interoperability for Microwave Access (Wi-MAX) interface, an Ethernet interface, a universal serial bus (USB) interface, a cellular network interface, a Bluetooth interface, or a near field communication (NFC) interface.

[0128] In one embodiment of the present specification, the above components of the computing device 400 and Figure 4 Other components not shown in the figure may also be connected to each other, for example, via a bus. Figure 4 The computing device structure block diagram shown is for illustrative purposes only and is not intended to limit the scope of this specification. Those skilled in the art may add or replace other components as needed.

[0129] Computing device 400 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., a tablet computer, a personal digital assistant, a laptop computer, a notebook computer, a netbook computer, etc.), a mobile phone (e.g., a smartphone), a wearable computing device (e.g., a smartwatch, smart glasses, etc.), or other types of mobile devices, or a stationary computing device such as a desktop computer or a personal computer (PC). Computing device 400 may also be a mobile or stationary server.

[0130] Processor 420 is configured to execute the following computer-executable instructions, which, when executed by the processor, implement the steps of the aforementioned SARIMA model load delivery forecasting method that considers time factors. The above is a schematic diagram of a computing device according to this embodiment. It should be noted that the technical solution of this computing device and the technical solution of the aforementioned SARIMA model load delivery forecasting method that considers time factors are based on the same concept. For details not described in detail in the technical solution of the computing device, please refer to the description of the technical solution of the aforementioned SARIMA model load delivery forecasting method that considers time factors.

[0131] An embodiment of the present specification further provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the SARIMA model load delivery prediction method considering time factors.

[0132] The above is a schematic diagram of a computer-readable storage medium according to this embodiment. It should be noted that the technical solution of this storage medium is based on the same concept as the technical solution of the SARIMA model load outbound forecasting method considering time factors. For details not described in detail in the technical solution of the storage medium, please refer to the description of the technical solution of the SARIMA model load outbound forecasting method considering time factors.

[0133] An embodiment of the present specification further provides a computer program, wherein when the computer program is executed in a computer, the computer is caused to execute the steps of the above-mentioned SARIMA model load transmission forecasting method considering time factors.

[0134] The above is an illustrative embodiment of a computer program. It should be noted that the technical solution of this computer program is based on the same concept as the technical solution of the SARIMA model load delivery forecasting method that considers time factors. For details not described in detail in the technical solution of the computer program, please refer to the description of the technical solution of the SARIMA model load delivery forecasting method that considers time factors.

[0135] The foregoing description of this specification describes specific embodiments. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different from that described in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order shown or the sequential order to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0136] The computer instructions include computer program code, which may be in source code form, object code form, executable file, or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0137] It should be noted that for the aforementioned method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the embodiments of this specification are not limited by the order of the actions described, because according to the embodiments of this specification, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the embodiments of this specification.

[0138] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0139] The preferred embodiments disclosed above are intended only to help illustrate this specification. The optional embodiments do not exhaustively describe all details, nor do they limit the invention to the specific embodiments described. Obviously, many modifications and variations can be made based on the content of the embodiments of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the embodiments of this specification, so that those skilled in the art can better understand and utilize this specification. This specification is limited only by the claims and their full scope and equivalents.

Claims

1. A SARIMA model load delivery forecasting method considering time factors, characterized in that: include: Building crawler data, and acquiring power market related data based on the crawler data; preprocessing the electricity market related data based on preprocessing rules to obtain preprocessed data; Determining abnormal data using a preset algorithm based on the preprocessed data; Determining target data based on data smoothing replacement of the abnormal data, and determining exogenous variables based on time information of the abnormal data; Performing model training based on the target data, the exogenous variables and the initial SARIMA model to determine a target SARIMA model; Real-time data is acquired, and forecast data is determined based on the real-time data and the target SARIMA model.

2. The method according to claim 1, characterized in that The acquiring of power market related data based on the crawler data includes: Based on the crawler data, load data and outbound data are obtained from the power trading center and the network interface; Power market related data is determined based on the load data and the transmission data.

3. The method according to claim 1, characterized in that The preprocessing of the electricity market related data based on the preprocessing rules to obtain preprocessed data includes: Determine standardization processing rules and data processing rules based on preprocessing rules; Performing unified field naming and type conversion on the power market related data based on the standardized processing rules to determine standardized data; Perform missing value processing and data deduplication on the standardized data based on the data processing rules to determine preprocessed data.

4. The method according to claim 1, wherein The determining abnormal data based on the pre-processed data by using a preset algorithm includes: The model is trained using the isolation forest algorithm and the local anomaly factor algorithm to determine the abnormal data processing model; Abnormal data is determined from the pre-processed data based on the abnormal data processing model.

5. The method according to claim 1, characterized in that The step of replacing the abnormal data based on data smoothing to determine target data, and determining an exogenous variable based on time information of the abnormal data, includes: determining an averaging window, and determining an average value based on the averaging window; The abnormal data is replaced based on the average value and time information is determined, and an exogenous variable is determined based on the time information.

6. The method according to claim 1, characterized in that The performing model training based on the target data, the exogenous variables and the initial SARIMA model to determine the target SARIMA model includes: determining training data based on the target data, the exogenous variables, and a preset ratio; Model training is performed based on the training data and the initial SARIMA model to determine a target SARIMA model.

7. The method according to claim 1, characterized in that Also includes: Evaluating the predicted data to determine evaluation data; The target SARIMA model is adjusted based on the evaluation data.

8. A SARIMA model load delivery forecasting device considering time factors, characterized in that: include: a data acquisition module configured to construct crawler data and acquire power market related data based on the crawler data; a data preprocessing module, configured to preprocess the power market related data based on preprocessing rules to obtain preprocessed data; an abnormality determination module, configured to determine abnormal data using a preset algorithm based on the preprocessed data; an exogenous variable module configured to determine target data based on data smoothing replacement of the abnormal data and to determine exogenous variables based on time information of the abnormal data; A model training module is configured to perform model training based on the target data, the exogenous variables and the initial SARIMA model to determine a target SARIMA model; The data prediction module is configured to acquire real-time data and determine predicted data based on the real-time data and the target SARIMA model.

9. A computing device, characterized in that include: memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the SARIMA model load delivery prediction method considering time factors as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions, when executed by a processor, implement the steps of the SARIMA model load delivery forecasting method considering time factors as described in any one of claims 1 to 7.

Images