Peak-valley adaptive scheduling method and system for household electrical load

Abstract

The invention discloses a household electrical load peak-valley adaptive scheduling method and system, and relates to the technical field of load optimization. The method comprises the following steps: collecting historical power utilization data of each electric device in a family, analyzing power utilization characteristics, marking rigid loads, transferable loads and interruptible loads, and generating power utilization characteristic marks; establishing an electrical load prediction model of the family in a future preset time period; and on the basis of the output of the power utilization load prediction model and the peak-valley electricity price time period information, in combination with the power utilization characteristic mark, with the minimization of the power utilization cost, the minimization of the user comfort deviation and the minimization of the peak-valley difference of the load curve as targets, solving the optimal scheduling strategy of each power utilization device in the future preset time period. The technical problem that in the prior art, the household electricity consumption load scheduling is difficult to dynamically optimize and adjust in combination with the peak-valley electricity price and the user preference, and consequently the electricity consumption cost is high is solved, and the technical effect of reducing the household electricity consumption cost on the premise of meeting the electricity consumption requirement of the user is achieved.

Description

Technical Field

[0001] This invention relates to the field of load optimization technology, specifically to a peak-valley adaptive scheduling method and system for household electricity load. Background Technology

[0002] With the improvement of residents' living standards, the types and number of household electrical appliances are constantly increasing, and household electricity load is characterized by diversified electricity consumption behavior, concentrated time periods, and significant load fluctuations. Especially during peak hours when electricity prices are high, the simultaneous operation of multiple high-power electrical appliances can easily lead to increased household electricity costs and put significant pressure on the power distribution network. Existing household electricity load scheduling schemes are mostly based on fixed time period rules or simple electricity price thresholds for control. They typically only start, stop, or delay the operation of some electrical appliances according to preset peak and valley periods, lacking effective differentiation of the load characteristics of different electrical appliances within the household. At the same time, existing schemes often fail to fully incorporate users' personalized electricity consumption preferences and predicted household electricity load information during the scheduling process, resulting in insufficient adaptability of the scheduling strategy and difficulty in dynamically adjusting according to changes in electricity consumption behavior and electricity price fluctuations. In addition, in household electricity scenarios where multiple devices operate in parallel, existing technologies usually focus on a single optimization objective, emphasizing the reduction of electricity costs, while ignoring factors such as user comfort and load curve smoothing. This can easily lead to a decline in the electricity experience or excessive concentration of load during low-price periods, thereby weakening the regulatory effect of the peak-valley electricity pricing mechanism. Summary of the Invention

[0003] This application provides a peak-valley adaptive scheduling method and system for household electricity load, which solves the technical problem in the prior art that it is difficult to dynamically optimize and adjust household electricity load scheduling in combination with peak-valley electricity prices and user preferences, resulting in high electricity costs.

[0004] A first aspect of this application provides a peak-valley adaptive scheduling method for household electricity load, the method comprising:

[0005] Historical electricity consumption data of each household appliance is collected, and the electricity consumption characteristics are analyzed to label rigid loads, transferable loads, and interruptible loads, generating electricity consumption characteristic labels. Historical load data of each household appliance, user-defined preference data, and real-time peak-valley electricity price information released by the power grid are collected. Based on the historical load data and user-defined preference data, an electricity load prediction model for the household's future preset time period is established. Based on the output of the electricity load prediction model, the peak-valley electricity price information, and the electricity consumption characteristic labels, the optimal scheduling strategy for each appliance in the future preset time period is solved with the objectives of minimizing electricity costs, minimizing user comfort deviations, and minimizing the peak-valley difference of the load curve.

[0006] A second aspect of this application provides a peak-valley adaptive scheduling system for household electricity loads, the system comprising:

[0007] Data Analysis Module: Collects historical electricity consumption data of various electrical devices in the household, analyzes electricity consumption characteristics, and labels rigid loads, transferable loads, and interruptible loads to generate electricity consumption characteristic labels; Model Building Module: Collects historical load data of various electrical devices in the household, user-defined preference data, and real-time peak and valley electricity price information released by the power grid, and establishes a household electricity load prediction model for a future preset period based on historical load data and user-defined preference data; Solution Module: Based on the output of the electricity load prediction model, the peak and valley electricity price information, and the electricity consumption characteristic labels, and with the objectives of minimizing electricity costs, minimizing user comfort deviations, and minimizing the peak-valley difference of the load curve, solves for the optimal scheduling strategy for each electrical device in the future preset period.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] First, historical electricity consumption data for each household appliance is collected, and its characteristics are analyzed to label rigid loads, transferable loads, and interruptible loads, generating electricity characteristic tags. Then, historical load data for each household appliance, user-defined preference data, and real-time peak-valley electricity price information released by the power grid are collected. Based on the historical load data and user-defined preference data, a future preset electricity load prediction model for the household is established. Finally, based on the output of the electricity load prediction model, peak-valley electricity price information, and electricity characteristic tags, the optimal scheduling strategy for each appliance within the future preset period is solved, aiming to minimize electricity costs, user comfort deviations, and the peak-valley difference in the load curve. This solves the technical problem in existing technologies where household electricity load scheduling is difficult to dynamically optimize and adjust in conjunction with peak-valley electricity prices and user preferences, leading to high electricity costs. It achieves the technical effect of reducing household electricity costs while meeting user electricity needs. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 A schematic flowchart of a peak-valley adaptive scheduling method for household electricity load provided in an embodiment of this application;

[0012] Figure 2This is a schematic diagram of a peak-valley adaptive scheduling system for household electricity load provided in an embodiment of this application.

[0013] Figure labeling: Data analysis module 11, model building module 12, solution module 13. Detailed Implementation

[0014] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0015] Example 1, as Figure 1 As shown, this application provides a peak-valley adaptive scheduling method for household electricity load, wherein the method includes:

[0016] Historical electricity consumption data of various electrical devices in the household is collected, and the electricity consumption characteristics are analyzed to mark rigid loads, transferable loads, and interruptible loads, generating electricity consumption characteristic tags.

[0017] In this embodiment, historical electricity consumption data of various electrical devices in a household is collected. This historical electricity consumption data includes at least the operating power, start / stop time, continuous operating duration, and corresponding timestamp information of each device within a preset statistical period. The statistical period can be set to daily, weekly, or monthly. The collected historical electricity consumption data is preprocessed, including anomaly removal, missing data completion, and time alignment, to form a standardized device electricity consumption dataset. Based on this standardized dataset, the power variation characteristics, start / stop frequency characteristics, and user operation time distribution characteristics of each device are analyzed. The power variation characteristics characterize whether the device's power output is stable or adjustable during operation; the start / stop frequency characteristics characterize whether the device frequently starts and stops within the statistical period; and the user operation time distribution characteristics characterize whether the device's operation is concentrated in a fixed time period. The extracted characteristic parameters are matched with preset load type determination rules. When a device has a fixed operating time, non-adjustable power, and is not allowed to be interrupted during operation, it is marked as a rigid load. When the device's operating time can be adjusted within a preset time window without affecting its function, it is marked as a transferable load. When the device is allowed to adjust its power or perform phased start-stop operations while meeting preset comfort or performance constraints, it is marked as an interruptible load. Through the above analysis and matching process, corresponding electricity consumption characteristic tags are generated for each electrical device in the household, and these tags are stored for subsequent electricity load prediction and peak-valley adaptive scheduling decisions.

[0018] Furthermore, historical electricity consumption data of various electrical devices within the household is collected, and the electricity consumption characteristics are analyzed to label rigid loads, transferable loads, and interruptible loads, generating electricity consumption characteristic tags, including:

[0019] Based on the historical electricity consumption data, the power change pattern, start-stop pattern, and user operation time distribution of each electrical device are extracted as key characteristic parameters. The key characteristic parameters are matched with a preset classification rule library, and load type labels are assigned to each electrical device according to the matching results, thereby generating the electricity consumption characteristic labels.

[0020] Based on historical electricity consumption data, the operating power sequence and start-stop records of each electrical device within a preset statistical period are analyzed to extract key characteristic parameters characterizing the device's electricity consumption behavior. These key characteristic parameters include at least the power change pattern, start-stop pattern, and user operation time distribution. Specifically, the power change pattern reflects whether the power output during device operation is constant, continuous, or adjustable; the start-stop pattern reflects whether the device's start-stop is highly correlated with real-time user operations, or exhibits periodic or task-oriented characteristics; and the user operation time distribution reflects the concentration and dispersion of device operating time over a one-day or multi-day timeline.

[0021] The extracted key characteristic parameters are input into a preset classification rule base for matching and judgment. This rule base predefines characteristic combination rules corresponding to different load types: when a device's power is basically constant during operation, and its start-stop time strictly follows the user's immediate operation commands, without time shifting or operational interruption conditions, the device is judged and marked as a rigid load; when a device's power curve exhibits continuous and stable task-oriented operation characteristics, and its start-stop time has a discrete distribution within a preset time window, allowing for overall shifting without affecting task completion, the device is judged and marked as a transferable load; when a device's operating power can be adjusted within a preset range, or the device can periodically start and stop or operate in segments according to set values, the device is judged and marked as an interruptible load. Based on the above matching and judgment results, corresponding load type labels are assigned to each electrical device in the household, generating the aforementioned electricity characteristic labels, which are then used as the basic input parameters for subsequent electricity load prediction and peak-valley adaptive scheduling decisions.

[0022] Furthermore, when assigning load type labels to each electrical device, transferable loads and interruptible loads can be labeled simultaneously on the same electrical device.

[0023] Preferably, when assigning load type tags to each electrical device, it is allowed to assign both transferable load tags and interruptible load tags to the same electrical device simultaneously. This is to characterize that the electrical device, under the premise of meeting the operating constraints, has both the scheduling capability of overall shifting of electricity consumption time and the scheduling capability of power regulation or phased start-stop during operation. Among them, rigid load tags are mutually exclusive with transferable load tags and interruptible load tags, that is, electrical devices marked as rigid loads are not assigned other load type tags.

[0024] Through the aforementioned composite labeling method, in the subsequent scheduling process, the scheduling strategy can simultaneously execute power consumption time shifting scheduling and operating power adjustment scheduling for electrical equipment that simultaneously has transferable load labels and interruptible load labels, combined with peak and valley electricity price time information and load optimization objectives, thereby improving the flexibility and optimization space of power load scheduling.

[0025] Historical load data of various electrical devices in the household, user-defined preference data, and peak-valley electricity price information released by the power grid in real time are collected. Based on the historical load data and user-defined preference data, a prediction model for the household's electricity load during the future preset time period is established.

[0026] In this embodiment, historical load data of various electrical devices within a household, user-defined preference data, and real-time peak-valley electricity price information released by the power grid are collected. The historical load data includes at least the power consumption sequence, electricity consumption period distribution, and corresponding timestamp information of each electrical device within a historical period. The user-defined preference data includes the user's preferred operating periods, acceptable delay durations, comfort preference parameters, and allowed power adjustment ranges set for different electrical devices. The peak-valley electricity price information includes the electricity price level and time boundaries corresponding to each billing period. The collected data undergoes time alignment, normalization, and outlier correction to construct a unified household electricity consumption data sample set.

[0027] Based on the aforementioned household electricity consumption data sample set, a household electricity load prediction model for a preset future time period is established. Specifically, historical load data is used as the primary time-series input to characterize the load variation patterns of the entire household and individual electrical appliances; user-defined preference data is used as auxiliary feature input to reflect the impact of user electricity consumption behavior on load changes. During the model training phase, historical load data is divided into training samples using a sliding window according to a preset prediction period, and supervised learning is employed to train the prediction model's parameters, enabling the model to output prediction results for the overall household electricity load and the load of each electrical appliance within the preset future time period. Through the above methods, a household electricity load prediction model that comprehensively reflects historical electricity consumption behavior and user preference characteristics is constructed, providing a predictive basis for subsequent load scheduling optimization based on peak-valley electricity pricing.

[0028] Furthermore, based on historical load data and user-defined preference data, a model for predicting household electricity load for a preset period in the future is established, including:

[0029] A hybrid neural network prediction model is constructed, comprising a long short-term memory network layer for capturing the temporal variation pattern of load, and a fully connected layer for fusing user preferences and environmental features; historical load data is processed in a time series manner, and combined with corresponding user-defined preference data as training samples, the hybrid neural network prediction model is trained under supervision to establish the electricity load prediction model.

[0030] Preferably, a hybrid neural network prediction model is constructed, which includes at least a long short-term memory network layer for processing load time-series data and a fully connected layer connected to it. The long short-term memory network layer is used to perform time-series modeling of historical household load data to capture the long-term dependencies and short-term fluctuations in the load of the entire household and individual electrical appliances over time. The fully connected layer is used to fuse the time-series features from the long short-term memory network layer with user-defined preference data to generate a comprehensive load feature representation reflecting the user's electricity consumption behavior.

[0031] During model training, the historical load data is first segmented into time-series segments according to a preset time step to construct corresponding load sequence samples. Each load sequence sample is then associated with user-defined preference data within the same time window to form training sample pairs. These training sample pairs are input into the hybrid neural network prediction model, and supervised learning is used to iteratively train the model parameters, gradually converging the prediction error between the model output and the actual electricity load for the corresponding time period. Once the model training meets the preset convergence condition, an electricity load prediction model is obtained, which outputs the predicted electricity load for households in the future for a preset time period.

[0032] Based on the output of the electricity load prediction model and the peak-valley electricity price information, combined with the electricity consumption characteristic markers, the optimal scheduling strategy for each electrical device in the future preset time period is solved with the objectives of minimizing electricity costs, minimizing user comfort deviations, and minimizing the peak-valley difference of the load curve.

[0033] After obtaining the predicted output of the electricity load forecasting model, based on the predicted output and the peak-valley electricity price information released by the power grid in real time, and combined with the electricity consumption characteristic tags corresponding to each electrical device, the scheduling optimization of the operation of electrical devices in the household within a future preset time period is performed. Specifically, according to the output of the electricity load forecasting model, the predicted power consumption curve or predicted electricity consumption sequence of each electrical device within the future preset time period is obtained, and the electricity price level corresponding to each time period is determined according to the peak-valley electricity price information.

[0034] In the scheduling optimization process, a scheduling optimization objective system is constructed with minimizing the overall household electricity cost, minimizing user comfort deviation, and minimizing the peak-to-valley difference of the load curve as joint optimization objectives. Electricity cost measures the total electricity cost incurred by operating each electrical device at the corresponding electricity price in each time period; user comfort deviation measures the degree of deviation between the actual scheduling result and the user's set preferences; and the peak-to-valley difference of the load curve measures the difference between the peak and valley values ​​of the overall household load within a preset time period. Based on the electricity consumption characteristic markers, differentiated scheduling constraints are applied to different types of electrical devices. Rigid loads maintain their original operating sequence without adjustment; transferable loads are allowed to shift their operating time as a whole, provided that the time window constraint is met; and interruptible loads are allowed to adjust their operating power or perform phased start-stop operations, provided that the preset comfort range is met.

[0035] Based on the above optimization objectives and scheduling constraints, a preset scheduling solution mechanism is adopted to iteratively adjust the operating time and operating power of each power-consuming device, calculate the objective function evaluation results corresponding to different scheduling schemes, and select the scheduling result that satisfies all constraints and has the best joint objective evaluation as the output of the optimal scheduling strategy for each power-consuming device in the future preset time period.

[0036] Furthermore, based on the output of the electricity load prediction model and the peak-valley electricity price information, combined with the electricity consumption characteristic markers, and with the objectives of minimizing electricity costs, minimizing user comfort deviations, and minimizing the peak-valley difference in the load curve, the optimal scheduling strategy for each electrical device within a future preset time period is solved, including:

[0037] Based on the output of the electricity load prediction model, the predicted scheduling sequence of power-consuming equipment is obtained; based on the electricity consumption characteristic tags, candidate scheduling equipment with tags of transferable load and interruptible load are extracted, and the scheduling priority is determined and sorted in descending order to generate a scheduling equipment sequence; based on the peak and valley electricity price time period information, an objective function is established with minimizing electricity cost, minimizing user comfort deviation, and minimizing the peak-valley difference of the load curve; based on the preset scheduling mechanism, the power consumption time of the equipment is adjusted according to the scheduling equipment sequence to generate the optimal scheduling strategy.

[0038] Preferably, firstly, based on the output of the electricity load forecasting model, the predicted power consumption curves and corresponding predicted operating sequences of each household's electrical equipment within a preset future time period are obtained, serving as the initial electrical equipment scheduling sequence. Secondly, based on the electricity consumption characteristic markers, the schedulability of each household's electrical equipment is screened, extracting electrical equipment that simultaneously or separately possesses transferable load markers and interruptible load markers as candidate scheduling equipment. These candidate scheduling equipment are then sorted according to their corresponding schedulability priority, which can be comprehensively determined based on equipment power consumption, time flexibility range, and impact on user comfort. The candidate scheduling equipment is arranged in descending order to generate a scheduling equipment sequence for subsequent scheduling adjustments. Then, based on the peak-valley electricity price information released by the power grid in real time, a multi-objective optimization objective function is constructed, including an electricity cost term, a user comfort deviation term, and a load curve peak-valley difference term. The electricity cost term characterizes the total electricity cost under each scheduling scheme, the user comfort deviation term characterizes the degree of deviation of the scheduling adjustment from the user's set preferences, and the load curve peak-valley difference term characterizes the impact of the scheduling results on the overall load smoothness of the household. Based on this, according to the preset scheduling mechanism, the power consumption time and operation mode of the candidate scheduling equipment are adjusted sequentially according to the order of the scheduling equipment sequence. Under the premise of meeting the corresponding operation constraints and comfort constraints of each power consumption equipment, the transferable load is prioritized to be shifted to the low electricity price period, and the operating power is adjusted or the interruptible load is started and stopped in stages. After the adjustment of each candidate scheduling equipment is completed, the objective function evaluation result of the corresponding scheduling scheme is calculated, and the scheduling scheme with the optimal joint objective function value is selected to generate the optimal scheduling strategy for each power consumption equipment in the future preset period.

[0039] Furthermore, the preset scheduling mechanism includes:

[0040] Transferable loads can be shifted to lower-cost periods while meeting time window constraints, and interruptible loads can adjust their operating power or start / stop within a preset comfort range.

[0041] The preset scheduling mechanism includes differentiated scheduling rules set for different load types. For electrical equipment marked as transferable loads, without changing its total operating tasks, the overall operating period is adjusted according to the corresponding operating time window constraints. This ensures that, while meeting the earliest allowed start-up time and latest completion time constraints, the equipment is preferentially shifted to lower-cost periods within the peak-valley electricity price range. The time window constraints can be determined by the equipment's operating characteristics or user-defined preferences, and are used to limit the adjustable time range of transferable loads.

[0042] For electrical equipment marked as interruptible loads, its operating power or operating status is adjusted under the condition of meeting the preset comfort range constraints. Specifically, the preset comfort range is used to limit the allowable power change range, number of start-stop operations, or continuous operating duration of the interruptible load during the scheduling process. When the power change or start-stop operation caused by the scheduling adjustment does not exceed the preset comfort range, it is allowed to perform power reduction, power increase, or phased start-stop operations on the electrical equipment to reduce the electricity load during high electricity price periods and smooth the overall load curve.

[0043] By setting scheduling rules for transferable and interruptible loads respectively, while ensuring the normal operation of electrical equipment and the user's electricity experience, the household electricity load can be flexibly adjusted during peak and off-peak electricity price periods, providing an executable scheduling basis for subsequent multi-objective optimization scheduling.

[0044] Furthermore, before adjusting the equipment power consumption time through the preset scheduling mechanism based on the power consumption characteristic marker, the following steps are also included:

[0045] Based on the peak and off-peak electricity price information, the current peak and off-peak electricity periods are determined; the load balance index of the peak and off-peak electricity periods is analyzed according to the predicted power equipment scheduling sequence; if the load balance index meets the preset balance index, no adjustment of equipment power consumption time is made; if it does not meet the preset balance index, the equipment power consumption time is adjusted through the preset scheduling mechanism of the power consumption characteristic mark.

[0046] First, based on the real-time peak-valley electricity price information released by the power grid, the electricity price level corresponding to each billing period is analyzed. According to preset electricity price grading rules, the peak and off-peak electricity consumption periods within the current forecast cycle are determined. Peak consumption periods correspond to time intervals where the electricity price is higher than a preset high-price threshold, and off-peak electricity periods correspond to time intervals where the electricity price is lower than a preset low-price threshold. Second, based on the predicted power equipment scheduling sequence output by the electricity load forecasting model, the predicted load distribution of households during the peak and off-peak electricity consumption periods is calculated. Based on the predicted load distribution results, a load balance index is constructed to characterize the degree of load equilibrium. This load balance index can be used to quantify the degree of load difference between peak and off-peak periods. Then, the load balance index is compared with the preset balance index. When the load balance index meets the preset balance index, it is determined that the current predicted power load distribution is in an acceptable balanced state, and the equipment power consumption time adjustment is not triggered. When the load balance index does not meet the preset balance index, it is determined that the current predicted power load distribution has a significant peak-valley imbalance, and the preset scheduling mechanism based on the power consumption characteristic mark is triggered to perform power consumption time adjustment and operation mode adjustment on the power equipment corresponding to the transferable load and interruptible load, so as to improve the load distribution and generate an optimized scheduling result.

[0047] Furthermore, after generating the optimal scheduling strategy, the process also includes:

[0048] Receive feedback information from home users regarding the optimal scheduling strategy; compile records of manual changes made by users to the optimal scheduling strategy based on the feedback information; adjust the weights of each evaluation item in the objective function based on the manual change records, and simultaneously adjust the priority of schedulable devices.

[0049] After generating the optimal scheduling strategy, the process further includes adaptively correcting the scheduling model based on user feedback. Specifically, after the optimal scheduling strategy is distributed and applied to the operation of household appliances, user feedback is received. This feedback includes at least user confirmation and rejection of the scheduling results, as well as records of manual modifications to the operating time or mode of individual appliances. Based on the received feedback, user manual modification records are statistically analyzed to identify user acceptance of different scheduling adjustments. These manual modification records can characterize user preferences regarding electricity costs, electricity comfort, or load balancing effects during actual use. Based on these records, the weight parameters corresponding to each evaluation item in the objective function are adjusted. When frequent cancellation or rollback of an addition scheduling adjustment is detected, the weight of the corresponding evaluation item in the objective function is reduced; when a user consistently accepts a certain type of scheduling adjustment, the weight of the corresponding evaluation item in the objective function is increased. At the same time, based on the adjustment results of the weight parameters, the schedulable priority of the power equipment is updated synchronously, so that the subsequent scheduling process is more in line with the user's actual power consumption habits and preferences.

[0050] Through the aforementioned weight adjustment and schedulable priority adjustment mechanism based on user feedback, the household electricity load scheduling strategy can be continuously and adaptively optimized, so that the scheduling results can better meet the actual electricity needs and user experience of users while reducing electricity costs.

[0051] Furthermore, the schedulable priority of candidate scheduling devices is set based on the power consumption and time elasticity of the corresponding devices.

[0052] Specifically, based on the rated power or average operating power consumption of each candidate scheduling device obtained from historical operating data, the impact of the device on the overall household electricity cost and load curve when participating in scheduling adjustments is evaluated. At the same time, based on the adjustable time range allowed for each candidate scheduling device under user-defined preferences and device operating constraints, the time flexibility level of the device is evaluated. The time flexibility is used to characterize the flexibility of the device's operating period to be shifted or split.

[0053] In the calculation of schedulable priority, the power consumption level and time flexibility are comprehensively considered. Equipment with higher power consumption and greater time flexibility is assigned a higher schedulable priority, allowing it to participate in scheduling adjustments first in the scheduling equipment sequence. Equipment with lower power consumption or limited time flexibility is assigned a relatively lower schedulable priority. Through this priority rule based on a combination of power consumption and time flexibility, scheduling resources are prioritized for equipment that has a significant impact on overall electricity costs and load balance, thereby improving the efficiency and effectiveness of scheduling optimization.

[0054] In summary, the embodiments of this application have at least the following technical effects:

[0055] First, historical electricity consumption data for each household appliance is collected, and its characteristics are analyzed to label rigid loads, transferable loads, and interruptible loads, generating electricity characteristic tags. Then, historical load data for each household appliance, user-defined preference data, and real-time peak-valley electricity price information released by the power grid are collected. Based on the historical load data and user-defined preference data, a future preset electricity load prediction model for the household is established. Finally, based on the output of the electricity load prediction model, peak-valley electricity price information, and electricity characteristic tags, the optimal scheduling strategy for each appliance within the future preset period is solved, aiming to minimize electricity costs, user comfort deviations, and the peak-valley difference in the load curve. This solves the technical problem in existing technologies where household electricity load scheduling is difficult to dynamically optimize and adjust in conjunction with peak-valley electricity prices and user preferences, leading to high electricity costs. It achieves the technical effect of reducing household electricity costs while meeting user electricity needs.

[0056] Example 2, based on the same inventive concept as the peak-valley adaptive scheduling method for household electricity load in the foregoing examples, such as... Figure 2 As shown, this application provides a peak-valley adaptive scheduling system for household electricity load, wherein the system includes:

[0057] Data Analysis Module 11: Collects historical electricity consumption data of each electrical device in the household, analyzes electricity consumption characteristics, and marks rigid loads, transferable loads, and interruptible loads to generate electricity consumption characteristic tags; Model Building Module 12: Collects historical load data of each electrical device in the household, user-defined preference data, and peak-valley electricity price information released by the power grid in real time, and establishes an electricity load prediction model for the household in the future preset period based on historical load data and user-defined preference data; Solution Module 13: Based on the output of the electricity load prediction model, the peak-valley electricity price information, and the electricity consumption characteristic tags, and with the objectives of minimizing electricity costs, minimizing user comfort deviations, and minimizing the peak-valley difference of the load curve, solves the optimal scheduling strategy for each electrical device in the future preset period.

[0058] Furthermore, the solver module 13 is used to perform the following method:

[0059] Based on the output of the electricity load prediction model, the predicted scheduling sequence of power-consuming equipment is obtained; based on the electricity consumption characteristic tags, candidate scheduling equipment with tags of transferable load and interruptible load are extracted, and the scheduling priority is determined and sorted in descending order to generate a scheduling equipment sequence; based on the peak and valley electricity price time period information, an objective function is established with minimizing electricity cost, minimizing user comfort deviation, and minimizing the peak-valley difference of the load curve; based on the preset scheduling mechanism, the power consumption time of the equipment is adjusted according to the scheduling equipment sequence to generate the optimal scheduling strategy.

[0060] Furthermore, the solver module 13 is used to perform the following method:

[0061] Transferable loads can be shifted to lower-cost periods while meeting time window constraints, and interruptible loads can adjust their operating power or start / stop within a preset comfort range.

[0062] Furthermore, the solver module 13 is used to perform the following method:

[0063] Based on the peak and off-peak electricity price information, the current peak and off-peak electricity periods are determined; the load balance index of the peak and off-peak electricity periods is analyzed according to the predicted power equipment scheduling sequence; if the load balance index meets the preset balance index, no adjustment of equipment power consumption time is made; if it does not meet the preset balance index, the equipment power consumption time is adjusted through the preset scheduling mechanism of the power consumption characteristic mark.

[0064] Furthermore, the solver module 13 is used to perform the following method:

[0065] Receive feedback information from home users regarding the optimal scheduling strategy; compile records of manual changes made by users to the optimal scheduling strategy based on the feedback information; adjust the weights of each evaluation item in the objective function based on the manual change records, and simultaneously adjust the priority of schedulable devices.

[0066] Furthermore, the solver module 13 is used to perform the following method:

[0067] The schedulable priority of candidate scheduling devices is set based on the power consumption and time elasticity of the corresponding devices.

[0068] Furthermore, the data analysis module 11 is used to perform the following methods:

[0069] Based on the historical electricity consumption data, the power change pattern, start-stop pattern, and user operation time distribution of each electrical device are extracted as key characteristic parameters. The key characteristic parameters are matched with a preset classification rule library, and load type labels are assigned to each electrical device according to the matching results, thereby generating the electricity consumption characteristic labels.

[0070] Furthermore, the data analysis module 11 is used to perform the following methods:

[0071] When assigning load type labels to each electrical device, transferable loads and interruptible loads can be labeled simultaneously on the same electrical device.

[0072] Furthermore, the model building module 12 is used to perform the following methods:

[0073] A hybrid neural network prediction model is constructed, comprising a long short-term memory network layer for capturing the temporal variation pattern of load, and a fully connected layer for fusing user preferences and environmental features; historical load data is processed in a time series manner, and combined with corresponding user-defined preference data as training samples, the hybrid neural network prediction model is trained under supervision to establish the electricity load prediction model.

[0074] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A peak-valley adaptive scheduling method for household electricity load, characterized in that, The method includes: Collect historical electricity consumption data of various electrical devices in the home, analyze the electricity consumption characteristics to mark rigid loads, transferable loads and interruptible loads, and generate electricity consumption characteristic tags; Collect historical load data of various electrical devices in the household, user-defined preference data, and peak-valley electricity price information released by the power grid in real time. Based on the historical load data and user-defined preference data, establish a predictive model for the household's electricity load during the future preset time period. Based on the output of the electricity load prediction model and the peak-valley electricity price information, combined with the electricity consumption characteristic markers, the optimal scheduling strategy for each electrical device in the future preset time period is solved with the objectives of minimizing electricity costs, minimizing user comfort deviations, and minimizing the peak-valley difference of the load curve.

2. The peak-valley adaptive scheduling method for household electricity load as described in claim 1, characterized in that, Based on the output of the electricity load prediction model and the peak-valley electricity price information, combined with the electricity consumption characteristic markers, and with the objectives of minimizing electricity costs, minimizing user comfort deviations, and minimizing the peak-valley difference in the load curve, the optimal scheduling strategy for each electrical device within a future preset time period is solved, including: Based on the output of the power load prediction model, the predicted power equipment scheduling sequence is obtained; Based on the electricity consumption characteristic tags, candidate scheduling devices with tags of transferable load and interruptible load are extracted, and the scheduling priority is determined and sorted in descending order to generate a scheduling device sequence. Based on the peak and valley electricity price time information, an objective function is established to minimize electricity costs, user comfort deviation, and load curve peak-valley difference. Based on the preset scheduling mechanism, the power consumption time of the equipment is adjusted according to the scheduling equipment sequence to generate the optimal scheduling strategy.

3. The peak-valley adaptive scheduling method for household electricity load as described in claim 2, characterized in that, The preset scheduling mechanism includes: Transferable loads can be shifted to lower-cost periods while meeting time window constraints, and interruptible loads can adjust their operating power or start / stop within a preset comfort range.

4. The peak-valley adaptive scheduling method for household electricity load as described in claim 2, characterized in that, Before adjusting the equipment power consumption time through the preset scheduling mechanism based on the power consumption characteristic marker, the following steps are also included: Based on the peak-valley electricity price time information, determine the current peak electricity consumption period and the off-peak electricity consumption period; Based on the predicted power equipment scheduling time sequence, analyze the load balance index of the peak power consumption period and the off-peak power consumption period; If the load balance index meets the preset balance index, no adjustment is made to the equipment power consumption time; otherwise, the equipment power consumption time is adjusted through the preset scheduling mechanism of the power consumption characteristic mark.

5. The peak-valley adaptive scheduling method for household electricity load as described in claim 2, characterized in that, After generating the optimal scheduling strategy, the following is also included: Receive feedback information from home users regarding the optimal scheduling strategy; Based on the response feedback information, statistical records of users' manual changes to the optimal scheduling strategy are compiled; Based on the manually changed records, the weights of each evaluation item in the objective function are adjusted, and the priority of device schedulability is adjusted simultaneously.

6. The peak-valley adaptive scheduling method for household electricity load as described in claim 1, characterized in that, The schedulable priority of candidate scheduling devices is set based on the power consumption and time elasticity of the corresponding devices.

7. The peak-valley adaptive scheduling method for household electricity load as described in claim 1, characterized in that, Historical electricity consumption data of various electrical devices in the household is collected, and the electricity consumption characteristics are analyzed to label rigid loads, transferable loads, and interruptible loads, generating electricity consumption characteristic tags, including: Based on the historical electricity consumption data, the power change pattern, start-stop rule and user operation time distribution of each electrical device are extracted as key characteristic parameters. Based on the key characteristic parameters and the preset classification rule base, load type labels are assigned to each electrical device according to the matching results, and the electrical characteristic labels are generated.

8. The peak-valley adaptive scheduling method for household electricity load as described in claim 7, characterized in that, When assigning load type labels to each electrical device, transferable loads and interruptible loads can be labeled simultaneously on the same electrical device.

9. The peak-valley adaptive scheduling method for household electricity load as described in claim 1, characterized in that, A household electricity load forecasting model for a preset time period is established based on historical load data and user-defined preference data, including: A hybrid neural network prediction model is constructed, which includes a long short-term memory network layer for capturing the temporal variation pattern of load, and a fully connected layer for fusing user preferences and environmental features. Historical load data is processed in a time series manner, and combined with corresponding user-defined preference data as training samples to supervise the training of the hybrid neural network prediction model, thereby establishing the electricity load prediction model.

10. A peak-valley adaptive scheduling system for household electricity load, characterized in that, For implementing the peak-valley adaptive scheduling method for household electricity load according to any one of claims 1-9, the system comprises: Data Analysis Module: Collects historical electricity consumption data of various electrical devices in the home, analyzes electricity consumption characteristics to mark rigid loads, transferable loads and interruptible loads, and generates electricity consumption characteristic tags; Model building module: Collects historical load data of various electrical devices in the household, user-defined preference data, and peak and valley electricity price information released by the power grid in real time, and builds a prediction model of the household's electricity load for the future preset period based on historical load data and user-defined preference data; Solution module: Based on the output of the electricity load prediction model and the peak-valley electricity price time period information, combined with the electricity consumption characteristic markers, with the objectives of minimizing electricity costs, minimizing user comfort deviations, and minimizing the peak-valley difference of the load curve, the module solves for the optimal scheduling strategy of each electrical device in the future preset time period.

Images