A power load forecasting system and method based on dual-mode communication

Abstract

The present invention relates to the technical field of power prediction, and discloses a power load prediction system and method based on dual-mode communication, including: obtaining the power consumption data of a user within a certain period, judging the usage attributes of each device of the user, obtaining the data of other users in the same area as a data reference, and integrating to form the power prediction data of the user. This power load prediction system and method based on dual-mode communication judges the power consumption habits of the user and the usage attributes of each device of the user. When a user temporarily does not use a certain device in the short term, resulting in the device having no usage habit of this user, the usage habit of this user for this device is predicted according to the usage habits of other users in the same area who use the same device, making the power load prediction data of the user more accurate, so as to minimize the deviation in power distribution as much as possible, avoid causing a certain shortage or redundancy in the distribution and use of power, and reduce the waste of resources.

Description

Technical Field

[0001] The present invention relates to the technical field of power prediction, and specifically to a power load prediction system and method based on dual-mode communication. Background Art

[0002] Power load prediction is a crucial part in the management of modern power systems, involving all aspects of power production, dispatching, distribution, and consumption. In order to effectively cope with the increasing power demand and improve the energy utilization efficiency, a power load prediction system has emerged. Accurate load prediction helps power companies reasonably dispatch power generation resources, avoid power waste or shortage, reduce unnecessary energy costs by reducing reserve capacity, and improve economic benefits. With the development of renewable energy (such as wind energy and solar energy), load prediction is crucial for optimizing its application. The power load prediction system based on dual-mode communication can effectively improve the flexibility and reliability of data transmission by combining two different communication methods, providing strong support for the real-time monitoring and prediction of power loads. The application of this technology can enhance the intelligent level of power management and provide an important guarantee for the sustainable development of future power systems;

[0003] The existing power load prediction system and method based on dual-mode communication cannot judge the user's electricity consumption habits according to the user's electricity consumption information within a certain period, nor can it judge the usage attributes of devices according to the user's usage habits of each device. When a user temporarily does not use a certain device in the short term, resulting in the device having no usage habits of this user, it is unable to estimate the usage habits of this user for this device based on the usage habits of other users in the same area who use the same device, resulting in certain deviations in the user's power load prediction data, thus causing deviations in power distribution, and further leading to possible shortages or redundancies in the distribution and use of power, causing certain waste of resources, and having certain limitations in its practicality. Summary of the Invention

[0004] The present invention provides a power load prediction system and method based on dual-mode communication to help solve the problems mentioned in the background art.

[0005] The present invention provides the following technical solutions: A power load prediction method based on dual-mode communication, including:

[0006] Obtain a target user;

[0007] Obtain a collection period;

[0008] Through a data collection strategy, obtain the electricity consumption data of the target user during the collection period to form historical electricity consumption data;

[0009] According to the historical electricity consumption data, through a preliminary analysis strategy, form preliminary estimation data;

[0010] Obtain all the electrical devices of the target user to form a device set;

[0011] Successively identify each element in the device set as an analysis device;

[0012] Execute a device analysis strategy on the analysis device to form device analysis data;

[0013] The device analysis data includes seasonal devices and non-seasonal devices;

[0014] If the device analysis data is a seasonal device, execute a first prediction strategy to form first prediction data;

[0015] If the device analysis data is a non-seasonal device, execute a second prediction strategy to form second prediction data;

[0016] If the preliminary prediction data = the first prediction data or the second prediction data, calculate the integrated prediction data, and the integrated prediction data = (preliminary prediction data × 5) × (1 + 30%);

[0017] If the preliminary prediction data ≠ the first prediction data or the second prediction data, calculate the integrated prediction data, and the integrated prediction data = (the first prediction data or the second prediction data × 5) × (1 + 30%)

[0018] Obtain the electricity consumption corresponding to the integrated prediction data, and generate a power prediction result for the target user.

[0019] As a power load prediction method based on dual-mode communication according to the present invention, wherein: the obtaining of the acquisition period is specifically as follows:

[0020] Obtain the current date;

[0021] Obtain the month in which the current date is located and define it as the current month;

[0022] Obtain all the dates of the current month to form a date set;

[0023] Extract the first element in the date set and define it as the sub-sampling date;

[0024] Extract all the interval elements between the current date and the sub-sampling date in the date set to form an interval set;

[0025] Obtain the number of elements in the interval set and define it as the interval number;

[0026] Obtain the number of elements in the date set and define it as the date number;

[0027] If the interval number + 2 ≥ 50% of the date number, then extract the median date of the current month and identify it as the judgment date;

[0028] If the number of intervals + 2 < 50% of the number of dates, then calculate the determination date, and the determination date = the current date - (the number of intervals + 2).

[0029] Set the acquisition duration.

[0030] Calculate the acquisition date, and the acquisition date = the determination date - the acquisition duration.

[0031] Taking the acquisition date as the start date and the determination date as the end date, form an acquisition period.

[0032] As a power load forecasting method based on dual-mode communication according to the present invention, wherein: the data acquisition strategy is specifically as follows:

[0033] Obtain the acquisition period.

[0034] Set the analysis duration.

[0035] Extract the acquisition date.

[0036] Taking the acquisition date as the start date, and at intervals of the analysis duration, form a data acquisition point within the acquisition period.

[0037] Obtain the acquisition date, the determination date, and all data acquisition points to form a time set.

[0038] Extract two adjacent elements in the time set in sequence, and respectively define them as the previous time element and the subsequent time element.

[0039] Define the time period formed between the previous time element and the subsequent time element as the acquisition time period.

[0040] Obtain all acquisition time periods within the acquisition period to form a time period set.

[0041] Obtain the electricity consumption data corresponding to each element in the time period set, and define it as historical data.

[0042] Make each historical data correspond to each element in the time period set one by one to form a historical data set.

[0043] As a power load forecasting method based on dual-mode communication according to the present invention, wherein: the preliminary analysis strategy is specifically as follows:

[0044] Obtain the device set.

[0045] Obtain the historical data set.

[0046] Obtain the acquisition period and the time period set.

[0047] Obtain the number of cycles of each element within the acquisition period, and define it as the element cycle.

[0048] Correspond each element period to each element within the acquisition period one by one to form a period set;

[0049] Set an analysis period set;

[0050] Extract each element in the analysis period set in sequence and define it as the analysis period number;

[0051] Extract all elements corresponding to the analysis period number within the period set to form a period element set;

[0052] Identify each element in the period element set as a determination period element in sequence;

[0053] Extract all time periods within the time period set that contain the determination period element to form a determination time period set;

[0054] Extract the element in the determination time period set where the determination period element is the end time of this time period and define it as the preliminary selection time period;

[0055] Extract all preliminary selection time periods within the acquisition period to form a preliminary selection set;

[0056] Extract the historical data corresponding to each element in the preliminary selection set within the historical data set and define it as the preliminary selection historical data;

[0057] Calculate the preliminary estimated value, Preliminary estimated value = sum of the preliminary selection historical data of all elements in the preliminary selection set ÷ number of elements in the preliminary selection set;

[0058] Extract the preliminary estimated value of each element in the analysis period set;

[0059] Correspond each preliminary estimated value to each element in the analysis period set one by one to form a preliminary estimation set;

[0060] Calculate the preliminary estimated data, Preliminary estimated data = sum of the preliminary estimated values of all elements in the preliminary estimation set.

[0061] As a power load forecasting method based on dual-mode communication according to the present invention, wherein: the device analysis strategy is specifically as follows:

[0062] Obtain the usage data of the target user for the analysis device and define it as the target usage data;

[0063] If there is no usage data for the analysis device, execute the reference analysis strategy;

[0064] If there is usage data for the analysis device, obtain all historical usage data of the analysis device to form a historical usage set;

[0065] Obtain the data acquisition time of each element in the historical usage set and define it as the historical time point;

[0066] Correspond each historical time point with each element in the historical usage set one by one to form a historical collection set;

[0067] Obtain the data collection date of each element in the historical collection set and define it as the historical date;

[0068] Correspond each historical date with each element in the historical collection set one by one to form a historical date set;

[0069] Obtain the data collection month of each element in the historical date set and define it as the historical month;

[0070] Correspond each historical month with each element in the historical date set one by one to form a historical month set;

[0071] Set a month set;

[0072] Successively identify each element in the historical month set as a month element;

[0073] If the month set contains the month element, determine that the month element is the first month element;

[0074] If the month set does not contain the month element, determine that the month element is the second month element;

[0075] Obtain the number of elements in the historical month set and define it as the month number;

[0076] Obtain the number of elements in the historical month set that determine the month element as the first month element and define it as the determined month number;

[0077] If the determined month number = the month number, determine that the device analysis data is seasonal equipment;

[0078] If the determined month number < the month number, determine that the device analysis data is non-seasonal equipment.

[0079] As a power load forecasting method based on dual-mode communication according to the present invention, wherein: the reference analysis strategy is specifically:

[0080] Obtain the area where the target user is located and define it as the target area;

[0081] Obtain all users using the analysis device in the target area to form a user set;

[0082] Successively identify each element in the user set as an analysis user;

[0083] Obtain all usage data of the analysis user for the analysis device, define it as the analysis usage data, and form an analysis usage set;

[0084] Obtain the data collection time of each element in the analysis usage set and define it as the analysis time point;

[0085] Correspond each analysis time point with each element in the analysis usage set one by one to form an analysis collection set;

[0086] Obtain the data collection date of each element in the analysis collection set and define it as the analysis date;

[0087] Correspond each analysis date with each element in the analysis collection set one by one to form an analysis date set;

[0088] Obtain the data collection month of each element in the analysis date set and define it as the analysis month;

[0089] Correspond each analysis month with each element in the analysis date set one by one to form an analysis month set;

[0090] Obtain the month set;

[0091] Successively identify each element in the analysis month set as a judgment month element;

[0092] If the month set contains the judgment month element, then determine that this month element is the first reference month;

[0093] If the month set does not contain the judgment month element, then determine that this month element is the second reference month;

[0094] Obtain the number of elements in the analysis month set and define it as the analysis month quantity;

[0095] Obtain the number of elements in the analysis month set that determine this month element as the first reference month and define it as the reference month quantity;

[0096] If the reference month quantity = the analysis month quantity, then determine that this user is a seasonal user;

[0097] If the reference month quantity < the analysis month quantity, then determine that this user is a non-seasonal user;

[0098] Obtain the number of elements in the user set and define it as the user quantity;

[0099] Obtain the number of elements in the user set that determine this user as a seasonal user and define it as the analysis user quantity;

[0100] If the user quantity = the analysis user quantity, then determine that the device analysis data is seasonal equipment;

[0101] If the user quantity > the analysis user quantity, then determine that the device analysis data is non-seasonal equipment.

[0102] As a power load forecasting method based on dual-mode communication according to the present invention, wherein: the first estimation strategy is specifically as follows:

[0103] Obtain a set of devices;

[0104] Obtain a set of periodic elements;

[0105] Successively identify each element in the set of periodic elements as an analysis periodic element;

[0106] Extract all time periods in the set of time periods that contain the analysis periodic element to form a set of periodic time periods;

[0107] Extract the element in the set of periodic time periods where the analysis periodic element is the end time of the time period as the selected time period;

[0108] Extract all the selected time periods within the acquisition period to form a selection set;

[0109] Successively identify each element in the set of devices as a selected device;

[0110] Obtain the usage data of the selected device corresponding to each element in the selection set as the selected usage data;

[0111] One-to-one correspond each selected usage data with each element in the selection set to form a device usage set;

[0112] Calculate the selected prediction value of the selected device in the number of analysis periods, where the selected prediction value = the sum of all the selected usage data in the device usage set ÷ the number of elements in the device usage set;

[0113] Calculate the analysis prediction value, where the analysis prediction value = the sum of the selected prediction values of all elements in the set of devices;

[0114] Extract the analysis prediction value of each element in the set of analysis periods;

[0115] One-to-one correspond each analysis prediction value with each element in the set of analysis periods to form an element prediction set;

[0116] Calculate the first prediction data, where the first prediction data = the sum of the analysis prediction values of all elements in the element prediction set.

[0117] As a power load forecasting method based on dual-mode communication according to the present invention, wherein: the second estimation strategy is specifically as follows:

[0118] Obtain a set of devices;

[0119] Obtain a set of periodic elements;

[0120] Successively identify each element in the set of periodic elements as an analysis periodic element;

[0121] Extract all time periods within the extraction time period set that contain the analysis period element to form a periodic time period set;

[0122] Extract the elements in the periodic time period set where the analysis period element is the termination time of the time period, and define them as the selected time periods;

[0123] Extract all the selected time periods within the acquisition period to form a selection set;

[0124] Successively identify each element in the device set as a selected device;

[0125] Obtain the user set corresponding to the selected device and define it as the selected user set;

[0126] Successively identify each element in the selected user set as a selected user;

[0127] Obtain the usage data of the selected users corresponding to the time periods of each element in the selection set for the selected device, and define it as the selected user data;

[0128] One-to-one correspond each selected user data with each element in the selection set to form a device-user set;

[0129] Calculate the determination pre-estimation value of the selected device in the number of analysis periods. The determination pre-estimation value = the sum of all the selected user data in the device-user set ÷ the number of elements in the device-user set;

[0130] Extract the determination pre-estimation value of the selected device for each element in the analysis period set;

[0131] One-to-one correspond each determination pre-estimation value with each element in the analysis period set to form a selection pre-estimation set;

[0132] Calculate the device pre-estimation value. The device pre-estimation value = the sum of the determination pre-estimation values of all elements in the selection pre-estimation set;

[0133] Extract the device pre-estimation value of each element in the selected user set;

[0134] One-to-one correspond each device pre-estimation value with each element in the selected user set to form a user pre-estimation set;

[0135] Calculate the reference pre-estimation value of the selected device. The reference pre-estimation value = the sum of the device pre-estimation values of all elements in the user pre-estimation set ÷ the number of elements in the user pre-estimation set;

[0136] Calculate the second pre-estimation data. The second pre-estimation data = the sum of the reference pre-estimation values of all elements in the device set.

[0137] As the present invention also discloses a system for implementing a power load forecasting method based on dual-mode communication, wherein: it includes:

[0138] Data acquisition module: used to obtain the power consumption data of the target user within the acquisition period;

[0139] Data analysis module: used to form preliminary estimated data according to historical power consumption data through preliminary analysis strategies;

[0140] Equipment analysis module: used to analyze each device of the target user to determine whether the device is a seasonal device;

[0141] Data integration module: used to calculate integrated estimated data according to the preliminary estimated data and the first estimated data or the second estimated data, and generate the power estimation result of the target user.

[0142] The present invention has the following beneficial effects:

[0143] 1. The power load forecasting system and method based on dual-mode communication obtain the power consumption information of users within a certain period, judge the power consumption habits of users, and preliminarily estimate the power load data of users for a week, so as to compare with the power load data estimated according to the equipment usage habits, and integrate to form more accurate power load data, making the power load forecasting data of users more accurate, thereby making the power distribution reduce deviations as much as possible, avoiding certain shortages or redundancies in the power distribution and use, and reducing the waste of resources.

[0144] 2. The power load forecasting system and method based on dual-mode communication obtain the usage conditions of each device of users, judge the usage habits of each device of users, and thus judge the usage attributes of the devices, that is, whether the devices are seasonal devices such as air conditioners, etc., and re-estimate the power load data of users for a week according to the usage conditions of the devices, and compare the power load data with the preliminarily estimated power load data, and integrate to form more accurate power load data, making the power load forecasting data of users more accurate, thereby making the power distribution reduce deviations as much as possible, avoiding certain shortages or redundancies in the power distribution and use, and reducing the waste of resources.

[0145] 3. For the power load prediction system and method based on dual-mode communication, when a user temporarily does not use a certain device in the short term, resulting in the device not having the user's usage habit, by obtaining the usage conditions of the same device by other users in the same area, comprehensively judging the usage attributes of the device, that is, whether the device is a seasonal device, and comprehensively estimating the usage situation of this user for the device according to the usage conditions of other users for the device, and then estimating the power load data of the user for a week again according to the usage situation of the device, and comparing the power load data with the preliminarily estimated power load data, integrating to form more accurate power load data, making the power load prediction data of the user more accurate, so as to make the power distribution reduce deviations as much as possible, avoid causing certain shortages or redundancies in the distribution and use of power, and reduce waste of resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0146] Figure 1 It is a flowchart of the power load prediction method based on dual-mode communication of the present invention;

[0147] Figure 2 It is a system block diagram for implementing the power load prediction method based on dual-mode communication of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0148] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0149] Embodiment 1, a power load prediction method based on dual-mode communication, refer to Figure 1 , including:

[0150] Obtain the target user;

[0151] Obtain the collection period;

[0152] Through the data collection strategy, obtain the power consumption data of the target user within the collection period to form historical power consumption data, and the power consumption data is the power consumption;

[0153] According to the historical power consumption data, form preliminary estimation data through the preliminary analysis strategy;

[0154] Obtain all the electrical devices of the target user to form a device set;

[0155] Successively identify each element in the device set as the analysis device;

[0156] Execute the device analysis strategy on the analysis device to form device analysis data;

[0157] The device analysis data includes seasonal devices and non-seasonal devices;

[0158] If the device analysis data is a seasonal device, execute the first prediction strategy to form the first prediction data;

[0159] If the device analysis data is a non-seasonal device, execute the second prediction strategy to form the second prediction data;

[0160] If the preliminary prediction data = the first prediction data or the second prediction data, calculate the integrated prediction data, and the integrated prediction data = (preliminary prediction data × 5) × (1 + 30%);

[0161] If the preliminary prediction data ≠ the first prediction data or the second prediction data, calculate the integrated prediction data, and the integrated prediction data = (the first prediction data or the second prediction data × 5) × (1 + 30%)

[0162] Obtain the power consumption corresponding to the integrated prediction data and generate the power prediction result of the target user.

[0163] Through the above method, judge the user's electricity consumption habits according to the user's electricity consumption information within a certain period, judge the usage attributes of the devices according to the user's usage habits of each device. When the user temporarily does not use a certain device in the short term, resulting in no usage habit of this device for this user, estimate the usage habit of this user for this device according to the usage habits of other users in the same region using the same device, making the user's power load prediction data more accurate, so as to make the power distribution reduce deviations as much as possible, avoid causing a certain shortage or redundancy in the power distribution and use, and reduce the waste of resources.

[0164] Embodiment 2. This embodiment is an improvement made on the basis of Embodiment 1. For the power load prediction method based on dual-mode communication, the obtaining of the acquisition period is specifically as follows:

[0165] Obtain the current date;

[0166] Obtain the month in which the current date is located and define it as the current month;

[0167] Obtain all the dates of the current month to form a date set;

[0168] Extract the first element in the date set and define it as the sub-sampling date;

[0169] Extract all the interval elements between the current date and the sub-sampling date in the date set to form an interval set, where the interval set does not include the sub-sampling date and the current date;

[0170] Obtain the number of elements in the interval set and define it as the interval number;

[0171] Obtain the number of elements in the date set, which is defined as the number of dates;

[0172] If the number of intervals + 2 ≥ 50% of the number of dates, then extract the median date of the current month and recognize it as the determination date, and the median date is the 15th;

[0173] If the number of intervals + 2 < 50% of the number of dates, then calculate the determination date, and the determination date = the current date - (the number of intervals + 2)

[0174] Set the acquisition duration, and the acquisition duration is 6 months;

[0175] Calculate the acquisition date, and the acquisition date = the determination date - the acquisition duration;

[0176] Form an acquisition period with the acquisition date as the start date and the determination date as the end date.

[0177] This embodiment also provides that the data acquisition strategy is specifically as follows:

[0178] Obtain the acquisition period;

[0179] Set the analysis duration, and the analysis duration is 24 hours;

[0180] Extract the acquisition date;

[0181] With the acquisition date as the start date, form a data acquisition point in the acquisition period every interval of the analysis duration;

[0182] Obtain the acquisition date, the determination date, and all data acquisition points to form a time set;

[0183] Extract two adjacent elements in the time set in sequence, and define them as the previous time element and the subsequent time element respectively;

[0184] Define the time period formed between the previous time element and the subsequent time element as the acquisition time period;

[0185] Obtain all acquisition time periods in the acquisition period to form a time period set;

[0186] Obtain the electricity consumption data corresponding to each element in the time period set, which is defined as historical data;

[0187] Correspond each historical data with each element in the time period set one by one to form a historical data set.

[0188] Embodiment 3, this embodiment is an improvement made on the basis of Embodiment 2. In this embodiment, the device analysis strategy is specifically as follows:

[0189] Obtain the usage data of the target user for the analysis device, and define it as the target usage data, where the usage data is the energy consumption data of the device;

[0190] If there is no usage data for the analysis device, execute the reference analysis strategy;

[0191] If there is usage data for the analysis device, obtain all the historical usage data of the analysis device to form a historical usage set;

[0192] Obtain the data collection time of each element in the historical usage set, and define it as the historical time point;

[0193] Correspond each historical time point with each element in the historical usage set one by one to form a historical collection set;

[0194] Obtain the data collection date of each element in the historical collection set, and define it as the historical date;

[0195] Correspond each historical date with each element in the historical collection set one by one to form a historical date set;

[0196] Obtain the data collection month of each element in the historical date set, and define it as the historical month;

[0197] Correspond each historical month with each element in the historical date set one by one to form a historical month set;

[0198] Set a month set, where the month set is from June to September and from December to March, with a total of 8 elements;

[0199] Successively identify each element in the historical month set as a month element;

[0200] If the month set contains the month element, determine that the month element is the first month element;

[0201] If the month set does not contain the month element, determine that the month element is the second month element;

[0202] Obtain the number of elements in the historical month set, and define it as the month number, where the number of elements is the number of months in the historical month set. For example, if the elements in the historical date set are January 3rd, January 4th, February 3rd, and June 3rd, and the elements in the historical month set are January, January, February, and June, then the number of elements in the historical month set is 3;

[0203] Obtain the number of elements in the historical month set that are determined to be the first month element, and define it as the determined month number;

[0204] If it is determined that the number of months = the number of months, it is determined that the device analysis data is for seasonal equipment, that is, the device is used only in specific months, indicating that the device may be seasonal equipment such as air conditioners;

[0205] If it is determined that the number of months < the number of months, it is determined that the device analysis data is for non-seasonal equipment, that is, the device is not used only in specific months, indicating that the device may be non-seasonal equipment such as televisions.

[0206] Among them, the reference analysis strategy is specifically as follows:

[0207] Obtain the area where the target user is located and define it as the target area;

[0208] Obtain all users who use the analysis device within the target area to form a user set;

[0209] Successively identify each element in the user set as an analysis user;

[0210] Obtain all usage data of the analysis user for the analysis device, define it as analysis usage data, and form an analysis usage set;

[0211] Obtain the data collection time of each element in the analysis usage set and define it as the analysis time point;

[0212] Correspond each analysis time point with each element in the analysis usage set one by one to form an analysis collection set;

[0213] Obtain the data collection date of each element in the analysis collection set and define it as the analysis date;

[0214] Correspond each analysis date with each element in the analysis collection set one by one to form an analysis date set;

[0215] Obtain the data collection month of each element in the analysis date set and define it as the analysis month;

[0216] Correspond each analysis month with each element in the analysis date set one by one to form an analysis month set;

[0217] Obtain a month set, and the month set is from June to September and from December to March, with a total of 8 elements;

[0218] Successively identify each element in the analysis month set as a determination month element;

[0219] If the month set contains the determination month element, it is determined that the month element is the first reference month;

[0220] If the month set does not contain the determination month element, it is determined that the month element is the second reference month;

[0221] Obtain the number of elements in the analysis month set, which is defined as the number of analysis months. Here, this number of elements is the number of months in the historical month set. For example, if the elements in the historical date set are January 3rd, January 4th, February 3rd, and June 3rd respectively, the elements in the historical month set are January, January, February, and June. Then the number of elements in the historical month set is 3;

[0222] Obtain the number of elements in the analysis month set that determine the month element as the first reference month, which is defined as the number of reference months;

[0223] If the number of reference months = the number of analysis months, then determine that the user is a seasonal user, that is, the user uses the device only in specific months;

[0224] If the number of reference months < the number of analysis months, then determine that the user is a non - seasonal user, that is, the user does not use the device only in specific months;

[0225] Obtain the number of elements in the user set, which is defined as the number of users;

[0226] Obtain the number of elements in the user set that determine the user as a seasonal user, which is defined as the number of analyzed users;

[0227] If the number of users = the number of analyzed users, then determine that the device analysis data is for a seasonal device, that is, the device is used only in specific months, indicating that the device may be a device with seasonality such as an air conditioner;

[0228] If the number of users > the number of analyzed users, then determine that the device analysis data is for a non - seasonal device, that is, the device is not used only in specific months, indicating that the device may be a device without seasonality such as a TV.

[0229] Embodiment 4. This embodiment is an improvement based on Embodiment 3. In this embodiment, the preliminary analysis strategy is specifically as follows:

[0230] Obtain the device set;

[0231] Obtain the historical data set;

[0232] Obtain the collection period and the time period set;

[0233] Obtain the number of cycles of each element within the collection period, which is defined as the element cycle. The number of cycles is the number corresponding to each element within a week. For example, if the element is Monday, then the number of cycles of this element is 1;

[0234] Correspond each element cycle with each element within the collection period to form a cycle set;

[0235] Set an analysis period set, where the analysis period set is a set formed by seven elements from Monday to Sunday;

[0236] Extract each element in the analysis period set in turn and define it as the analysis period number;

[0237] Extract all elements corresponding to the analysis period number in the period set to form a period element set;

[0238] Identify each element in the period element set as a determination period element in turn;

[0239] Extract all time periods in the time period set that contain the determination period element to form a determination time period set;

[0240] Extract the element in the determination time period set where the determination period element is the end time of this time period and define it as the preliminary selection time period;

[0241] Extract all preliminary selection time periods within the collection period to form a preliminary selection set;

[0242] Extract the historical data corresponding to each element in the preliminary selection set in the historical data set and define it as the preliminary selection historical data;

[0243] Calculate the preliminary estimated value. The preliminary estimated value = the sum of the preliminary selection historical data of all elements in the preliminary selection set ÷ the number of elements in the preliminary selection set;

[0244] Extract the preliminary estimated value of each element in the analysis period set;

[0245] Correspond each preliminary estimated value with each element in the analysis period set one by one to form a preliminary estimation set;

[0246] Calculate the preliminary estimated data. The preliminary estimated data = the sum of the preliminary estimated values of all elements in the preliminary estimation set.

[0247] This embodiment also provides that the first estimation strategy is specifically:

[0248] Obtain a device set;

[0249] Obtain a period element set;

[0250] Identify each element in the period element set as an analysis period element in turn;

[0251] Extract all time periods in the time period set that contain the analysis period element to form a period time period set;

[0252] Extract the element in the period time period set where the analysis period element is the end time of this time period and define it as the selected time period;

[0253] Extract all selected time periods within the collection period to form a selected set;

[0254] Identify each element in the device set as a selected device in sequence;

[0255] Obtain the usage data of the selected device corresponding to each element in the selection set for each period, and define it as the selected usage data;

[0256] Correspond each selected usage data to each element in the selection set one by one to form a device usage set;

[0257] Calculate the selected prediction value of the selected device in the analysis period number, and the selected prediction value = the sum of all selected usage data in the device usage set ÷ the number of elements in the device usage set;

[0258] If the current month is within the month set, calculate the analysis prediction value, and the analysis prediction value = the sum of the selected prediction values of all elements in the device set;

[0259] If the current month is not within the month set, calculate the analysis prediction value, and the analysis prediction value = the sum of the selected prediction values of the elements corresponding to all non-seasonal devices in the device set;

[0260] Extract the analysis prediction value of each element in the analysis period set;

[0261] Correspond each analysis prediction value to each element in the analysis period set one by one to form an element prediction set;

[0262] Calculate the first prediction data, and the first prediction data = the sum of the analysis prediction values of all elements in the element prediction set.

[0263] This embodiment also provides that the second prediction strategy is specifically:

[0264] Obtain the device set;

[0265] Obtain the periodic element set;

[0266] Identify each element in the periodic element set as an analysis periodic element in sequence;

[0267] Extract all periods in the period set that contain the analysis periodic element to form a periodic period set;

[0268] Extract the element in the periodic period set where the analysis periodic element is the end time of this period, and define it as the selected period;

[0269] Extract all selected periods within the acquisition period to form a selection set;

[0270] Identify each element in the device set as a selected device in sequence;

[0271] Obtain the user set corresponding to the selected device, and define it as the selected user set;

[0272] Each element in the selected user set is sequentially identified as a selected user;

[0273] Obtain the usage data of the selected users corresponding to each element in the selection set for the selected device during each period, and define it as the selected user data;

[0274] One-to-one correspondence is established between each selected user data and each element in the selection set to form a device-user set;

[0275] Calculate the determination prediction value of the selected device for the number of analysis periods. The determination prediction value = the sum of all selected user data in the device-user set ÷ the number of elements in the device-user set;

[0276] Extract the determination prediction value of the selected device for each element in the analysis period set;

[0277] One-to-one correspondence is established between each determination prediction value and each element in the analysis period set to form a selection prediction set;

[0278] Calculate the device prediction value. The device prediction value = the sum of the determination prediction values of all elements in the selection prediction set;

[0279] Extract the device prediction value of each element in the selected user set;

[0280] One-to-one correspondence is established between each device prediction value and each element in the selected user set to form a user prediction set;

[0281] Calculate the reference prediction value of the selected device. The reference prediction value = the sum of the device prediction values of all elements in the user prediction set ÷ the number of elements in the user prediction set;

[0282] If the current month is within the month set, then calculate the second prediction data. The second prediction data = the sum of the reference prediction values of all elements in the device set;

[0283] If the current month is not within the month set, then calculate the second prediction data. The second prediction data = the sum of the reference prediction values of the elements corresponding to all non-seasonal devices in the device set.

[0284] Example 5. This example also discloses a system for implementing a power load forecasting method based on dual-mode communication. Refer to Figure 2 , including:

[0285] Data acquisition module: used to obtain the electricity consumption data of the target users during the acquisition period;

[0286] Data analysis module: used to form preliminary prediction data according to the historical electricity consumption data through preliminary analysis strategies;

[0287] Device analysis module: used to analyze each device of the target user to determine whether the device is a seasonal device;

[0288] Data integration module: used to calculate the integrated estimated data based on the preliminary estimated data and the first estimated data or the second estimated data, and generate the electricity estimated result of the target user.

[0289] In this embodiment, the user's electricity consumption habits are judged according to the electricity consumption information of the user within a certain period, and the usage attributes of the device are judged according to the user's usage habits of each device. When the user temporarily does not use a certain device in the short term, resulting in the device having no usage habits of the user, the usage habits of this user for the device are estimated according to the usage habits of other users in the same region who use the same device, making the user's electricity load prediction data more accurate, so that the power distribution can reduce deviations as much as possible, avoid causing a certain shortage or redundancy in the distribution and use of electricity, and reduce the waste of resources.

Claims

1. A method for power load forecasting based on dual-mode communication, characterized in that: include: Get target users; Get the collection cycle; Through data collection strategies, the target user's electricity consumption data within the collection period is obtained to form historical electricity consumption data; Based on historical electricity consumption data, preliminary estimated data is formed through preliminary analysis strategies; Obtain all electrical devices of the target user to form a device set; Identify each element of the device set in turn as an analysis device; Execute device analysis strategy on the analysis device to form device analysis data; The equipment analysis data includes seasonal equipment and non-seasonal equipment; If the device analysis data is a seasonal device, the first estimation strategy is executed to form the first estimation data, and the power load data of the user for one week is estimated again according to the user's use of the device. When the user does not use a certain device in a short period of time and the device does not have the user's usage habits, the use of the device by the user is comprehensively estimated by obtaining the use of the device by other users in the same area who use the same device, and the power load data of the user for one week is estimated again according to the use of the device; If the device analysis data is a non-seasonal device, the second estimation strategy is executed to form the second estimation data, and the power load data of the user for one week is estimated again according to the user's use of the device. When the user does not use a certain device in a short period of time and the device does not have the user's usage habits, the user's use of the device is comprehensively estimated by obtaining the use of the device by other users in the same area who use the same device, and the power load data of the user for one week is estimated again according to the use of the device; If the preliminary estimated data = the first estimated data or the second estimated data, then calculate the integrated estimated data, the integrated estimated data = (preliminary estimated data × 5) × (1 + 30%); If the preliminary estimated data ≠ the first estimated data or the second estimated data, then calculate the integrated estimated data, integrated estimated data = (first estimated data or second estimated data × 5) × (1 + 30%) Obtain the power consumption corresponding to the integrated estimated data and generate the power estimation results for the target users.

2. The method for power load forecasting based on dual-mode communication according to claim 1, characterized in that: The acquisition collection cycle is specifically: Get the current date; Get the month of the current date and set it as the current month; Get all dates of the current month to form a date collection; Extract the first element in the date set and set it as the sampling date; Extract all interval elements between the current date and the sub-date in the date set to form an interval set; Get the number of elements in the interval set and set it as the number of intervals; Get the number of elements in the date collection and set it as the number of dates; If the number of intervals + 2 ≥ the number of dates × 50%, the median date of the current month is extracted and identified as the judgment date; If the number of intervals + 2 < the number of dates × 50%, then calculate the judgment date, judgment date = current date - (number of intervals + 2) Set the collection duration; Calculate the collection date, collection date = determination date - collection duration; The collection period is formed by taking the collection date as the starting date and the determination date as the ending date.

3. The method for power load forecasting based on dual-mode communication according to claim 1, characterized in that: The data collection strategy is specifically as follows: Get the collection cycle; Set the analysis duration; Extract collection date; Taking the collection date as the starting date, each interval analysis duration forms a data collection point within the collection period; Obtain the collection date, determination date and all data collection points to form a time set; Sequentially extract two adjacent elements in the time set and define them as the previous time element and the next time element respectively; The period formed between the previous time element and the subsequent time element is defined as the acquisition period; Obtain all collection time periods within the collection cycle to form a time period set; Obtain the electricity consumption data corresponding to each element in the time period set and define it as historical data; Each historical data is matched one by one with each element in the time period set to form a historical data set.

4. The method for power load forecasting based on dual-mode communication according to claim 1, characterized in that: The preliminary analysis strategy is specifically as follows: Get the device collection; Get historical data collection; Get the collection cycle and time period set; Get the period number of each element in the acquisition period, which is defined as the element period; Each element period is matched one by one with each element in the acquisition period to form a period set; Set the analysis period set; Extract each element in the analysis period set in turn and set it as the analysis period number; Extract and analyze all elements corresponding to the periodic number in the periodic set to form a periodic element set; Identify each element in the periodic element set as a determination periodic element in turn; Extract all time periods containing determination period elements in the time period set to form a determination time period set; Extract the element whose determination period element is the end time of the period from the determination period set, and define it as the primary selection period; Extract all the preliminary selection periods within the collection period to form a preliminary selection set; Extract the historical data corresponding to each element in the preliminary selection set in the historical data set, and define it as the preliminary selection historical data; Calculate the preliminary estimated value, which is the sum of the preliminary historical data of all elements in the preliminary set ÷ the number of elements in the preliminary set; Extracting a preliminary estimate of each element in the analysis period set; Each preliminary estimated value is matched one-to-one with each element in the analysis period set to form a preliminary estimated set; Calculate the preliminary estimated data, the preliminary estimated data = the sum of the preliminary estimated values ​​of all elements in the preliminary estimated set.

5. The method for power load forecasting based on dual-mode communication according to claim 1, characterized in that: The device analysis strategy is specifically: Obtain target user's usage data of the analysis device and set it as target usage data; If there is no usage data for the analyzed device, the reference analysis strategy is executed; If the analysis device has usage data, all historical usage data of the analysis device are obtained to form a historical usage set; Get the data collection time of each element in the historical usage set and set it as the historical time point; Each historical time point is matched one by one with each element in the historical usage set to form a historical collection set; Get the data collection date of each element in the historical collection set and set it as the historical date; Match each historical date with each element in the historical collection set one by one to form a historical date set; Get the data collection month of each element in the historical date set and set it as the historical month; Match each historical month with each element in the historical date set one by one to form a historical month set; Set month collection; Identify each element in the historical month set as a month element in turn; If the month set contains a month element, the month element is determined to be the first month element; If the month set does not contain a month element, the month element is determined to be the second month element; Get the number of elements in the historical month collection, set as the number of months; Get the number of elements in the historical month set that determine the month element as the first month element, and define it as the number of determined months; If the number of judged months = the number of months, the equipment analysis data is judged to be seasonal equipment; If the number of judged months is less than the number of months, the equipment analysis data is judged to be non-seasonal equipment.

6. The method for power load forecasting based on dual-mode communication according to claim 5, characterized in that: The reference analysis strategy is specifically: Get the area where the target user is located and set it as the target area; Obtain all users who use analysis devices in the target area to form a user set; Identify each element in the user set as an analysis user in turn; Obtain all usage data of the analysis user on the analysis device, define it as analysis usage data, and form an analysis usage collection; Get the data collection time of each element in the analysis collection and set it as the analysis time point; Each analysis time point is matched one by one with each element in the analysis use set to form an analysis collection set; Get the data collection date of each element in the analysis collection set and set it as the analysis date; Match each analysis date with each element in the analysis collection set one by one to form an analysis date set; Get the data collection month of each element in the analysis date set and set it as the analysis month; Match each analysis month with each element in the analysis date set one by one to form an analysis month set; Get the month collection; Identify each element in the analysis month set as a determination month element in turn; If the month set contains a determination month element, the month element is determined to be the first reference month; If the month set does not contain the determined month element, the month element is determined to be the second reference month; Get the number of elements in the analysis month collection and set it as the number of analysis months; Get the number of elements in the analyzed month set that determine the month element as the first reference month, and set it as the number of reference months; If the number of reference months = the number of analysis months, the user is considered a seasonal user; If the number of reference months is less than the number of analysis months, the user is considered a non-seasonal user; Get the number of elements in the user set and set it as the number of users; Obtain the number of elements in the user set that determine that the user is a seasonal user, and define it as the number of analyzed users; If the number of users = the number of analyzed users, the device analysis data is determined to be a seasonal device; If the number of users is greater than the number of analyzed users, the device analysis data is determined to be a non-seasonal device.

7. The method for power load forecasting based on dual-mode communication according to claim 1, characterized in that: The first estimation strategy is specifically: Get the device collection; Get a collection of periodic elements; Identify each element in the set of periodic elements as an analysis periodic element in turn; Extract all time periods that contain the analysis period elements in the time period set to form a period time period set; Extract the element whose analysis period element is the end time of the period from the period time set and define it as the selected period; Extract all selected time periods within the collection cycle to form a selection set; Identify each element in the device set as a selection device in turn; Obtaining usage data of the selected device in the time period corresponding to each element in the selected set, and defining it as selected usage data; Each selected usage data is matched one-to-one with each element in the selection set to form a device usage set; Calculate the selection estimate of the selected device in the number of analysis cycles, where the selection estimate = the sum of all the selection usage data in the device usage set ÷ the number of elements in the device usage set; Calculate the estimated analysis value, where the estimated analysis value = the sum of the selected estimated values ​​of all elements in the device set; Extract the analysis estimate for each element in the analysis period set; Match each analysis estimate value with each element in the analysis period set one by one to form an element estimate set; Calculate the first estimated data, where the first estimated data=the sum of the analyzed estimated values ​​of all elements in the element estimation set.

8. The method for power load forecasting based on dual-mode communication according to claim 1, characterized in that: The second estimation strategy is specifically: Get the device collection; Get a collection of periodic elements; Identify each element in the set of periodic elements as an analysis periodic element in turn; Extract all time periods that contain the analysis period elements in the time period set to form a period time period set; Extract the element whose analysis period element is the end time of the period from the period time set and define it as the selected period; Extract all selected time periods within the collection cycle to form a selection set; Identify each element in the device set as a selection device in turn; Obtain the user set corresponding to the selected device and define it as the selected user set; Identify each element in the selected user set as a selected user in turn; Obtaining usage data of the selected user of the selected device in the time period corresponding to each element in the selected set, and defining it as the selected user data; Each selected user data is matched one-to-one with each element in the selected set to form a device user set; Calculate the judgment estimate value of the selected device in the number of analysis cycles, the judgment estimate value = the sum of all selected user data in the device user set ÷ the number of elements in the device user set; Extracting the judgment estimate of each element in the analysis period set of the selected device; Each judgment estimate is matched one-to-one with each element in the analysis period set to form a selection estimate set; Calculate the device estimate value, where the device estimate value = the sum of the judgment estimate values ​​of all elements in the selected estimate set; Extract the device estimate for each element in the selected user set; Match each device estimate value with each element in the selected user set one by one to form a user estimate set; Calculate the reference estimated value of the selected device, where the reference estimated value = the sum of the device estimated values ​​of all elements in the user estimated set / the number of elements in the user estimated set; Calculate the second estimated data, where the second estimated data=the sum of the reference estimated values ​​of all elements in the device set.

9. A system for executing the power load forecasting method based on dual-mode communication according to claim 1, characterized in that: include: Data collection module: used to obtain the target user's electricity consumption data within the collection period; Data analysis module: used to form preliminary estimated data based on historical electricity consumption data through preliminary analysis strategies; Device analysis module: used to analyze each device of the target user and determine whether the device is a seasonal device; Data integration module: used to calculate and integrate the estimated data based on the preliminary estimated data and the first estimated data or the second estimated data, and generate the power estimation result of the target user.

Images