Calculation method and system for step tariff residual electric quantity of low-voltage user

Abstract

The invention discloses a method and a system for calculating the residual electric quantity of step tariff for low-voltage users, and the method comprises the steps: introducing a machine learning algorithm, dynamically adjusting an early warning proportion threshold value according to the historical power consumption behaviors, seasonal power consumption rules and holiday and festival factors of the users, and guaranteeing that early warning is timely but not too frequent; setting a threshold adjustment rule, and when it is predicted that the user will enter the next step electricity price interval within the first preset time, gradually reducing an early warning threshold in advance so as to early warn the user in advance; calculating the step electricity price residual electricity quantity in real time, namely subtracting the current electricity consumption from the target step electricity consumption; according to the current electricity consumption time period and the step electricity price, the current electricity price is accurately calculated, and segmented charging details are included; when the power utilization proportion reaches the early warning threshold value after dynamic adjustment, early warning information is automatically generated; and sending the early warning information to the user through a short message, an e-mail, an APP push or an intelligent sound box voice communication channel.

Description

Technical Field

[0001] The present invention belongs to the technical field of electricity calculation, and in particular, relates to a method and system for calculating the remaining electricity at tiered electricity prices for low-voltage users. Background Art

[0002] With the increasing refinement of energy management and the improvement of environmental awareness, tiered electricity prices have been widely used around the world as a pricing mechanism to encourage users to save energy and reduce emissions. Especially among low-voltage users, tiered electricity prices set different electricity prices according to different ranges of user electricity consumption. The more electricity you use, the higher the price. The purpose is to guide users to use electricity rationally and reduce energy waste through economic levers. However, in order to fully utilize this mechanism, accurately calculating the remaining electricity of users' tiered electricity prices and giving users early warnings at the right time have become key technical challenges.

[0003] Traditional static early warning systems often face the problem of improper selection of early warning timing: if the early warning threshold is set too high, users may not receive timely reminders when the tiered electricity price is close to adjustment, missing the best time to adjust their electricity strategy; if the threshold is too low, frequent early warnings may cause unnecessary anxiety to users, reducing user experience. Therefore, there is an urgent need for a more intelligent and personalized early warning system that can accurately predict user electricity consumption trends and dynamically adjust early warning strategies based on user characteristics and external environment to ensure that early warnings are both timely and appropriate. Summary of the invention

[0004] In order to solve the problems in the background technology, the basic concept of the technical solution adopted by the present invention is:

[0005] A method for calculating the remaining power of a tiered electricity price for a low-voltage user comprises the following steps:

[0006] Obtain user information of target low-voltage users through automatic collection by smart meters or manual reporting by users; determine the current electricity consumption of users this month;

[0007] According to the user's electricity consumption type, query and apply the corresponding tiered electricity price strategy, which includes different tiered electricity prices and corresponding electricity consumption thresholds;

[0008] Calculate the target step electricity consumption, that is, the next step electricity threshold corresponding to the current electricity consumption, and the corresponding step electricity price;

[0009] Introducing machine learning algorithms to dynamically adjust the warning ratio threshold based on users' historical electricity consumption behavior, seasonal electricity consumption patterns, and holiday factors to ensure that warnings are both timely and not too frequent;

[0010] Set threshold adjustment rules. When it is predicted that a user will enter the next tiered electricity price range within the first preset time, the warning threshold is gradually lowered in advance to warn the user in advance.

[0011] Real-time calculation of the remaining electricity at the tiered electricity price, i.e. the target tiered electricity consumption minus the current electricity consumption;

[0012] According to the current electricity consumption period and tiered electricity prices, the current electricity charges are accurately calculated, including the details of segmented billing;

[0013] When the electricity consumption ratio reaches the dynamically adjusted warning threshold, warning information is automatically generated;

[0014] Warning information will be delivered to users via SMS, email, APP push or smart speaker voice communication channels.

[0015] Furthermore, a machine learning algorithm is introduced to dynamically adjust the warning ratio threshold according to the user's historical electricity consumption behavior, seasonal electricity consumption patterns, and holiday factors. Specifically, it includes: collecting users' electricity consumption data, cleaning abnormal data, filling missing values ​​in the data preprocessing stage, and standardizing time series data to facilitate machine learning model training;

[0016] Based on the collected data, a feature set for machine learning is constructed. The key features include:

[0017] Historical electricity consumption trends, seasonal patterns, holiday effects, user category impacts, and time series characteristics.

[0018] Furthermore, a machine learning algorithm is introduced to dynamically adjust the warning ratio threshold according to the user's historical electricity consumption behavior, seasonal electricity consumption patterns, and holiday factors. The specific details include: Model selection and training: Select the XGBoost model and use historical data to train the model. The goal is to predict the user's electricity consumption in a certain period of time in the future and determine when it is close to the next step electricity threshold. During the training process, cross-validation and grid search techniques are used to optimize the model parameters to ensure the accuracy of the prediction;

[0019] After the model predicts the user's future electricity consumption trend, the warning threshold is dynamically adjusted according to the prediction results. The threshold is set as follows: when it is predicted that there is a 90% probability that the user will reach the threshold within the next week, a warning will be issued.

[0020] Furthermore, the user information of the target low-voltage user includes user ID, electricity usage type, and electricity usage time period.

[0021] Furthermore, the warning information includes remaining power, expected upgrade time, and recommended power saving measures.

[0022] Furthermore, the "first preset time" is set as a configurable parameter, and the time is set to 1 week, 2 weeks or one month before the user expects to reach the next step power threshold.

[0023] The system for calculating the remaining power of the tiered electricity prices for low-voltage users is used to execute the above-mentioned method for calculating the remaining power of the tiered electricity prices for low-voltage users.

[0024] Compared with the prior art, the present invention has the following beneficial effects:

[0025] This technical solution is aimed at the calculation method and system of the remaining power of the tiered electricity price for low-voltage users. The solution comprehensively uses modern information technology, big data analysis and machine learning algorithms to achieve in-depth insight and accurate prediction of users' electricity consumption behavior, and then dynamically adjusts the warning threshold, and efficiently conveys warning information through multiple channels, so that users can enjoy the tiered electricity price while achieving efficient management of electricity consumption, providing strong technical support for promoting social energy conservation and emission reduction and optimizing the efficiency of power grid operation.

[0026] Intelligent fusion data collection: A dual-track data acquisition mode that combines automatic data collection by smart meters with manual reporting by users ensures the comprehensiveness and accuracy of data, adapts to different user scenarios, and reduces blind spots in data collection.

[0027] Fine-grained tiered electricity price management: Build a dynamically updated tiered electricity price strategy database to achieve accurate matching of user electricity usage types and electricity price strategies, ensuring fair and transparent electricity fee calculations.

[0028] Machine learning-driven early warning optimization: Advanced machine learning algorithms such as XGBoost are introduced to dynamically adjust the early warning threshold according to the user's historical electricity consumption behavior, seasonal patterns, and holiday effects, so as to achieve personalized and intelligent early warning.

[0029] Multi-dimensional warning strategy: Through the gradient warning mechanism and personalized threshold adjustment algorithm, timely and appropriate warning information is provided to avoid excessive or insufficient warnings and improve user satisfaction.

[0030] Interactive user interface and multi-channel push: Create a user-friendly interface to intuitively display electricity consumption information and power saving suggestions, and ensure the efficient communication of early warning information through multiple channels such as SMS, email, APP push, etc., to enhance user interaction and response.

[0031] System security and data protection: Implement strict data encryption and privacy protection measures to ensure the security of user data in all aspects of collection, processing, storage and transmission.

[0032] The implementation of the technical solution not only solves the information asymmetry problem faced by low-voltage users under the tiered electricity price system, but also promotes the transformation of users' energy-saving behavior through technological empowerment, providing strong technical support for building a green, intelligent and efficient modern energy management system. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 This is a flow chart of the calculation method of the tiered electricity price remaining power for low-voltage users in this application. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments will be clearly and completely described below in conjunction with the drawings and embodiments of the present invention. The following embodiments are used to illustrate the present invention.

[0035] This application discloses a method for calculating the remaining power of the tiered electricity price for low-voltage users, such as Figure 1 , the method comprises the steps of:

[0036] Step 1: Obtain user information and electricity data, obtain user information of target low-voltage users through automatic collection by smart meters or manual reporting by users, including but not limited to user ID, electricity usage type, and electricity usage period; (Automatic collection by smart meters: Using Internet of Things technology, smart meters automatically record users' real-time electricity usage data at certain time intervals (such as 15 minutes, 30 minutes or 1 hour), and upload the data to the cloud server through secure wireless communication protocols (such as LoRa, NB-IoT or Wi-Fi). This process does not require human intervention, reduces errors and delays in data collection, and improves the real-time and accuracy of data. Smart meters must also have a timestamp function to ensure that they can distinguish between working days and non-working days and electricity consumption in various time periods (peak, flat, and valley).

[0037] Supplementary mechanism for manual reporting by users: For users who do not have smart meters installed or in special circumstances (such as smart meter failure), a channel for manual reporting is provided. Users can enter or confirm their electricity consumption information through mobile applications, web portals or telephone customer service, including total electricity consumption on a specific day and estimated consumption by time period (peak, flat and valley). The system is designed with a user-friendly interface and guidance process to ensure that users can easily complete the reporting, and at the same time, the rationality of the reported data is verified to ensure data quality.

[0038] Electricity usage time segmentation: When obtaining electricity usage time data, ensure that the data entered by the smart meter or the user manually can accurately reflect the electricity consumption in each time period: peak (e.g. 8am to 10pm), flat (e.g. 10pm to 8am the next day, excluding valley time), and valley (e.g. specific time period late at night). Smart meters have advanced metering functions, which can automatically classify and record data according to pre-set time periods, or provide users with the function of selecting the corresponding time period when manually reporting.

[0039] In the process of obtaining user information, ensure that all data transmission is encrypted and uses the HTTPS protocol or other industry standard security protocols to protect user privacy and data security. The server storing user information must meet industry security standards and conduct regular security audits and maintenance to prevent data leakage.

[0040] In addition to basic user ID and electricity usage period information, the user's electricity usage type should also be recorded in detail, including but not limited to residential users, commercial users, and industrial users, in order to facilitate the application of different tiered electricity price strategies.

[0041] Determine the user's current electricity consumption this month, taking into account the difference between working days and non-working days, and accurately calculate the electricity consumption in each period (peak, flat, and valley);

[0042] Step 2: Calculate the tiered electricity price and target electricity consumption. According to the user's electricity consumption type, query and apply the corresponding tiered electricity price strategy, including different tiered electricity prices and corresponding electricity consumption thresholds;

[0043] Construction of a tiered electricity price strategy database: First, establish or access a detailed tiered electricity price strategy database that covers the latest tiered electricity price policies for various types of electricity consumption. Including but not limited to residential electricity consumption, commercial electricity consumption, industrial electricity consumption, and special electricity consumption types (such as public service agencies and electric vehicle charging stations). Under each type of electricity consumption, clearly record the starting electricity threshold of different tiers, the corresponding electricity price, and any applicable preferential policies or rate adjustment rules.

[0044] Policy query interface design: Set up a query interface that can automatically retrieve and apply the corresponding tiered electricity price strategy based on user ID or electricity usage type. This interface has efficient data matching capabilities to ensure that the user's tiered electricity price details can be quickly and accurately located, and can seamlessly adapt even in the face of complex policy adjustments or new electricity usage types.

[0045] Dynamic policy update mechanism: Given that the tiered electricity price policy may change with the market, season or government decision, the system has a built-in policy update mechanism to synchronize the latest electricity price information from designated channels at regular intervals to ensure that the policies in the database are always consistent with the latest policies. When updating, the system should have version control and rollback functions to deal with any potential data inconsistency issues.

[0046] Personalized electricity price display: The user-side application or service intuitively displays the currently applicable tiered electricity price structure, including the starting and ending electricity consumption of each tier and the corresponding electricity price, so that users can clearly understand how their electricity charges are calculated. In addition, a simulation calculator is provided so that users can estimate the changes in electricity charges based on the estimated future electricity consumption, enhancing transparency and user participation.

[0047] Calculate the target step electricity consumption, that is, the next step electricity threshold corresponding to the current electricity consumption, and the corresponding step electricity price;

[0048] First, the system needs to extract all the tiered electricity thresholds and their corresponding electricity prices for the user's electricity consumption type from the tiered electricity price strategy database. For example, for residential users, the database records that the first tier of electricity consumption is 0-300 kWh, and the electricity price is X yuan / kWh; the second tier of electricity consumption is 301-500 kWh, and the electricity price is Y yuan / kWh, and so on.

[0049] Real-time power matching: The system compares the user's current total power consumption with the step power threshold level by level to determine which step the user is currently in. If the user's power consumption just reaches or exceeds the upper limit of a step, then that step becomes the current step, and the next higher power threshold becomes the starting point of the "target step power consumption".

[0050] Accurately calculate the target power threshold: Once it is determined that the user is about to enter the next level, the system needs to accurately calculate the starting power value of the level as the "target level power consumption". For example, if the user's current power consumption is 450 kWh, which is at the end of the second level, then the target level power consumption is the starting point of the next level, which is 501 kWh.

[0051] Corresponding tiered electricity price determination: The electricity price corresponding to the target tiered electricity consumption is the electricity price of the next tier. Continuing with the above example, if the target tiered electricity consumption belongs to the third tier, the corresponding electricity price is the electricity price of the third tier.

[0052] Cross-tier electricity consumption calculation logic: When a user's electricity consumption spans two tier intervals, the system has refined calculation logic to ensure accurate division of electricity consumption within the two tier intervals, and calculate the electricity bill using the electricity prices of their respective intervals to avoid simple averaging or erroneous accumulation.

[0053] Real-time feedback and display: The user interface displays in real time the current level, the remaining power to the next level, and the estimated increase in electricity bills that users will need to pay if they reach the next level, helping users to intuitively understand their electricity usage and potential costs.

[0054] Dynamic adjustment and optimization: As the user's electricity consumption changes, the system should automatically re-evaluate the user's step position and target step electricity consumption, especially when the user approaches the step boundary, and update the calculation results in real time to ensure the timeliness and accuracy of the information.

[0055] Through these detailed operations, users can clearly understand their specific gap to the next level of electricity consumption threshold and the electricity price changes they will face after reaching the threshold, thereby effectively guiding users' energy-saving behavior and electricity fee management;

[0056] Step 3: Dynamic warning threshold setting: introduce machine learning algorithms to dynamically adjust the warning ratio threshold based on the user's historical electricity consumption behavior, seasonal electricity consumption patterns, and holiday factors to ensure that the warning is both timely and not too frequent;

[0057] Specifically: Data collection and preprocessing: First, the system continuously collects users' electricity consumption data, including daily, weekly, monthly and even annual electricity consumption, and marks holidays, special events (such as extreme weather), and weekdays and weekends. The data preprocessing stage involves cleaning abnormal data, filling missing values, and standardizing time series data to facilitate machine learning model training.

[0058] Feature engineering: Based on the collected data, construct a feature set for machine learning. Key features include but are not limited to:

[0059] Historical electricity usage trends: Analyze users' electricity usage patterns in the same month or similar date ranges over the past year.

[0060] Seasonal patterns: Identify peaks and troughs in electricity consumption during summer and winter.

[0061] Holiday effect: considers changes in users’ electricity usage behavior before and after holidays, such as increased household electricity consumption during holidays.

[0062] Impact of user category: Consider the impact of user electricity usage type (such as residential, commercial) on electricity usage habits.

[0063] Time series characteristics: Use sliding window statistics (such as the average electricity consumption in the past week or month) as a predictor of real-time electricity consumption behavior.

[0064] Model selection and training: Select the XGBoost model. Use historical data to train the model, with the goal of predicting the user's electricity consumption in a certain period of time in the future and determining when the next step electricity consumption threshold is approaching. During the training process, use cross-validation and grid search techniques to optimize model parameters to ensure the accuracy of the prediction.

[0065] Dynamic warning threshold calculation: After the model predicts the user's future electricity consumption trend, the system dynamically adjusts the warning threshold based on the prediction results. The threshold setting takes into account the user's actual electricity consumption and the predicted fluctuation range, ensuring that the warning can be issued before the user is about to reach the next step of electricity threshold, while avoiding false warnings due to daily fluctuations. For example, it can be set to start issuing warnings when it is predicted that the user has a 90% probability of reaching the threshold in the next week.

[0066] Adaptive learning of threshold adjustment rules: By continuously receiving feedback on when users actually reach a new level, the machine learning model can self-optimize and adjust the warning rules. This means that the model will learn more precise warning trigger conditions, such as which special events will cause significant changes in users' electricity usage patterns, and how to more accurately predict these changes.

[0067] The XGBoost model is an advanced machine learning algorithm, the full name of which is "Extreme Gradient Boosting". It is optimized and extended based on the Gradient Boosting Machine (GBM). It is one of the models that currently performs well in many machine learning competitions and practical applications. In this application, the XGBoost model is used to predict the future power consumption of low-voltage users, and dynamically adjust the warning threshold accordingly to ensure the timeliness and effectiveness of the warning. The following are several key features and advantages of the XGBoost model:

[0068] Efficiency and high performance: XGBoost optimizes computing efficiency and greatly improves training speed and memory usage efficiency through technologies such as parallel computing, column-based data handling, and sparse-aware algorithms, making it perform well when processing large-scale data sets.

[0069] Regularization: XGBoost introduces L1 and L2 regularization during model training, which helps prevent overfitting, where the model performs well on the training data but has poor generalization ability on unknown data. Regularization terms can help simplify the model and improve the generalizability of the model.

[0070] Flexible loss function: XGBoost supports custom loss functions, which means it can adapt to various regression and classification problems. In this application scenario, the loss function design that may be involved can specifically measure the deviation between the predicted power consumption and the actual power consumption to ensure the accuracy of the prediction.

[0071] Learning rate adjustment: The model gradually iterates to build a tree model with a smaller learning rate. Each iteration only learns a small part of the residual of the previous step. This can better control the complexity of the model and further reduce the risk of overfitting.

[0072] Feature importance assessment: XGBoost can provide feature importance scores, which is very helpful for understanding which factors (such as historical electricity consumption trends, seasonality, holidays, etc.) are most critical to user electricity consumption forecasts, facilitating further optimization of models and strategy formulation.

[0073] Cross-validation and grid search: When training the XGBoost model, cross-validation is used to evaluate the robustness of the model performance to ensure that the model is not overfitting to the training data. Grid search technology is used to automatically find the best hyperparameter combination to further improve the prediction accuracy of the model.

[0074] By applying the XGBoost model, this technical solution can more accurately predict the electricity consumption trends of low-voltage users, and dynamically adjust the warning thresholds based on these predictions, ensuring that the warning system can both promptly remind users that the next threshold of the tiered electricity price is about to be reached and avoid user distress caused by improper threshold setting, thereby achieving intelligent management and guidance of user electricity consumption behavior.

[0075] Through the implementation of the above specific details, this technical solution not only uses machine learning algorithms to improve the intelligence level of the early warning mechanism, but also ensures that the early warning is both timely and highly personalized and accurate through a continuous learning and optimization process, effectively assisting users in rationally planning the use of electricity, and promoting energy conservation, emission reduction and optimization of power grid management.

[0076] The method for calculating the remaining power of the tiered electricity price for low-voltage users also includes the following steps:

[0077] Set threshold adjustment rules. For example, when it is predicted that a user will enter the next tiered electricity price range within the first preset time, gradually lower the warning threshold in advance to warn the user in advance.

[0078] Threshold adjustment strategy:

[0079] Time window definition: First, clarify the concept of "first preset time" and set it as a configurable parameter. For example, the system administrator or algorithm can set this time to 1 week, 2 weeks or one month before the user is expected to reach the next step power threshold based on the characteristics of the user group and past warning effect feedback. The selection of this time window needs to balance the advance amount and accuracy of the warning, so as to avoid unnecessary tension caused by early warning or untimely response of users due to late warning.

[0080] Gradient warning mechanism: Within the first preset time range, the system should gradually lower the warning threshold to form a gradient warning based on the urgency of the user's expected reaching the next level of power threshold. For example, if it is predicted that the user will reach the threshold within 2 weeks, the initial warning threshold can be set to 80% of the current power to the target power. As time approaches, it will gradually drop to 60%, 40%, and finally 20%, and so on. This gradient adjustment not only ensures the timeliness of the warning, but also avoids the abruptness of the warning, giving users room to gradually adjust their electricity usage behavior.

[0081] For example, this process can be implemented through Python function code, which receives the user's current power level, target power threshold, number of days before the warning starts (for example, the warning starts 2 weeks before the threshold is reached) and the number of steps by which the warning threshold decreases as input parameters, and then gradually reduces the warning threshold according to the described gradient.

[0082] Python def adjust_alert_threshold(current_usage,target_threshold,days_to_start_alert=14, steps=4):

[0083] """

[0084] Function to adjust the warning threshold.

[0085] parameter:

[0086] -current_usage: current power usage (unit: kWh)

[0087] -target_threshold: target power threshold (unit: kWh)

[0088] -days_to_start_alert: The number of days to start the alert, the default is 14 days

[0089] -steps: The number of decreasing steps of the step warning, the default is 4 steps

[0090] return:

[0091] - A list showing the warning thresholds for each day

[0092] """

[0093] #Calculate the daily power increment to ensure that the target power is reached before the warning starts

[0094] daily_increment=(target_threshold-current_usage) / days_to_start_alert

[0095] # Initialize the warning threshold list

[0096] alert_thresholds = []

[0097]

[0098] This code first calculates the amount of power that needs to be added each day before the warning starts, and then gradually reduces the percentage of the warning threshold according to the set number of warning steps until the warning threshold is close to 20% of the target power threshold. This example simplifies the calculation of power growth within the time window.

[0099] Personalized threshold adjustment algorithm: Combine the user's historical electricity consumption behavior pattern and machine learning prediction results to customize the warning threshold adjustment curve for each user. The algorithm needs to take into account the stability of the user's electricity consumption habits, seasonal changes, and personal response speed factors, and dynamically adjust the speed and amplitude of the threshold drop to ensure that the warning strategy meets the actual situation of each user.

[0100] Special handling of emergencies: When encountering special circumstances such as extreme weather or large-scale events that cause abnormal electricity consumption, the system has the ability to temporarily adjust the warning rules, such as starting the warning in advance or increasing the warning intensity, to cope with the sudden surge in electricity consumption and ensure that users have enough time to take countermeasures.

[0101] User feedback loop: Establish a user feedback mechanism to collect users' evaluation of the effectiveness of warning time, frequency and suggestions, and use these feedbacks to continuously optimize the warning threshold adjustment algorithm. For example, if a large number of users report that the warning is too frequent, the warning time window can be appropriately extended or the threshold drop rate can be adjusted.

[0102] Combination of automation and manual review: Although the adjustment of warning thresholds is mainly performed automatically by algorithms, the system should provide an interface for manual review and adjustment, allowing power company staff to intervene when necessary and manually fine-tune the warning thresholds based on actual power supply conditions or special policy requirements to ensure the flexibility and adaptability of the warning system.

[0103] Through these detailed threshold adjustment rules, the system can provide more humane services to low-voltage users while maintaining early warning efficiency, helping them to effectively manage their energy consumption while enjoying the preferential tiered electricity prices, thus promoting the sustainability of energy utilization and the stability of grid operation.

[0104] The method for calculating the remaining power of the tiered electricity price for low-voltage users also includes the following steps: Step 4: Real-time remaining power and electricity fee calculation, real-time calculation of the remaining power of the tiered electricity price, that is, the target tiered power consumption minus the current power consumption;

[0105] According to the current electricity consumption period and tiered electricity prices, the current electricity charges are accurately calculated, including the details of segmented billing;

[0106] Step 5: Warning information generation and multi-channel transmission. When the power consumption ratio reaches the warning threshold after dynamic adjustment, warning information is automatically generated, including remaining power, expected upgrade time, and recommended power saving measures. Warning information is quickly and accurately delivered to users through SMS, email, APP push, or smart speaker voice communication channels;

[0107] The details are as follows:

[0108] Design of early warning information content:

[0109] Remaining power reminder: Clearly inform users of the specific remaining power before the next tiered power threshold, such as: "You have 300 kWh remaining before the next tiered electricity price."

[0110] Estimated time to upgrade: Based on the user's current electricity consumption trend and forecast model, the system estimates and informs the user of the specific time when the next tier of electricity price is expected to be reached. For example, "Based on the current electricity consumption rate, it is expected that you will enter a higher tier of electricity price in 10 days."

[0111] Personalized power saving suggestions:

[0112] Time-based recommendations: Based on the analysis of the user's historical electricity usage time periods, recommendations are made to reduce unnecessary electricity usage during high electricity price periods, such as "It is recommended to reduce the use of air conditioners during the peak hours from 18:00 to 22:00 every day."

[0113] Alternative energy recommendations: For qualified users (such as those with solar power generation facilities), it is recommended to give priority to using solar energy when there is sufficient sunshine to save electricity bills.

[0114] Energy-saving equipment recommendations: Based on the user's electricity usage type, high-efficiency electrical appliances are recommended, such as LED lighting to replace traditional light bulbs, or smart thermostats to adjust air conditioning temperature to save energy.

[0115] Lifestyle adjustment: Provide life tips to encourage users to change certain power-consuming habits, such as using energy-saving mode to wash clothes, or using natural light instead of indoor lighting during the day.

[0116] Practical case sharing: Provide other users’ successful experience sharing or cases of energy saving to increase users’ practical confidence and motivation.

[0117] Interactive feedback portal: The warning information contains a simple and convenient feedback portal, allowing users to directly provide feedback on power saving effects or raise questions, thereby strengthening the interaction between users and the system and further optimizing personalized suggestions.

[0118] Optimization of early warning information format:

[0119] Multimedia presentation: Warning information is not limited to text, but can also include charts (such as power usage trend charts) and short video tutorials (demonstrations of power-saving tips) to improve user understanding and acceptance.

[0120] Multi-language support: In order to cover a wider user group, early warning information should support multiple languages ​​to ensure barrier-free transmission of information.

[0121] Multi-channel sending strategy:

[0122] Personalized channel selection: The system should automatically select the most effective information sending channel based on the user's preference settings or historical response behavior. For example, users who prefer SMS will receive SMS alerts first, while users who frequently use APPs will receive information through APP push.

[0123] Urgency grading: Dynamically adjust the priority of the sending channel based on the urgency of the warning and the user's response history. For example, for an emergency warning that is about to exceed the threshold, SMS and phone reminders can be used as a double insurance to ensure that notifications are in place.

[0124] Through the implementation of the above specific details, the early warning information can not only be conveyed to users accurately and promptly, but also provide electricity-saving guidelines with practical value, encouraging users to take effective measures to control electricity expenses, while improving user satisfaction and participation, and promoting a virtuous cycle of the entire system.

[0125] The present application discloses a system for calculating the remaining power of a low-voltage user at a tiered electricity price. The system for calculating the remaining power of a low-voltage user at a tiered electricity price is used to execute the above-mentioned method for calculating the remaining power of a low-voltage user at a tiered electricity price. The system for calculating the remaining power of a low-voltage user at a tiered electricity price includes:

[0126] Data acquisition and processing module:

[0127] Smart meter interface subsystem: responsible for automatically collecting smart meter data through IoT technologies (such as LoRa, NB-IoT, Wi-Fi), including user ID, power consumption period (peak, flat, valley) and real-time power consumption, to ensure the secure transmission of data to the cloud server.

[0128] User manual reporting subsystem: provides multiple channels such as web pages, mobile applications and customer service hotlines, allowing users to manually input electricity usage information, and integrates data cleaning and verification mechanisms to ensure data quality.

[0129] Data preprocessing subsystem: timestamps, processes outliers, fills in missing values, and standardizes the collected data to prepare for subsequent calculations and analysis.

[0130] Ladder electricity price strategy management module:

[0131] Tiered electricity price database: stores tiered electricity price strategies for various types of electricity consumption, synchronizes the latest policies in real time, and supports strategy query and dynamic update.

[0132] Electricity price calculation engine: Based on user type and real-time electricity data, query the applicable tiered electricity price strategy, calculate the current electricity fee and the threshold and expected electricity price for the remaining electricity to the next tier.

[0133] Machine learning early warning module:

[0134] Feature engineering and model training subsystem: Based on historical electricity consumption data, a model input containing features such as seasonality and holiday effects is constructed, and the XGBoost algorithm is used for training to optimize the warning threshold.

[0135] Early warning threshold dynamic adjustment subsystem: According to the prediction of user electricity consumption behavior, the early warning threshold is adjusted dynamically, and a gradient early warning mechanism is adopted to continuously optimize the early warning strategy in combination with user feedback loops.

[0136] Real-time calculation and display module:

[0137] Real-time electricity calculation engine: Calculates the remaining electricity of the user under the current tiered electricity price in real time and predicts the time to reach the next tier.

[0138] User interface: displays the user's current electricity usage level, remaining power, expected upgrade time and electricity fee changes, and provides power saving suggestions and simulation calculator.

[0139] Warning information generation and multi-channel push module:

[0140] Warning generation engine: When the user's power consumption approaches the warning threshold, warning information containing personalized power saving suggestions is automatically generated.

[0141] Multi-channel push service: Automatically select channels such as SMS, email, APP push or smart speakers according to user preferences to ensure accurate delivery of warning information, support emergency level classification and multiple languages.

[0142] Data security and privacy protection: Use HTTPS protocol and encryption algorithm to ensure data transmission security, strictly abide by data protection regulations, and implement access control and log auditing.

[0143] Algorithm optimization and model training: Continuously optimize XGBoost model parameters, use cross-validation and grid search technology to improve prediction accuracy, and ensure the timeliness and personalization of early warnings.

[0144] User interaction experience: Design a friendly and easy-to-use user interface to ensure intuitive information display, smooth feedback mechanism, and enhance user participation.

[0145] System compatibility and scalability: The microservice architecture is adopted to ensure that the system is modular, easy to expand and upgrade to adapt to future electricity consumption policies or technological changes.

[0146] This technical solution is aimed at the calculation method and system of the remaining power of the tiered electricity price for low-voltage users. The solution comprehensively uses modern information technology, big data analysis and machine learning algorithms to achieve in-depth insight and accurate prediction of users' electricity consumption behavior, and then dynamically adjusts the warning threshold, and efficiently conveys warning information through multiple channels, so that users can enjoy the tiered electricity price while achieving efficient management of electricity consumption, providing strong technical support for promoting social energy conservation and emission reduction and optimizing the efficiency of power grid operation.

[0147] Intelligent fusion data collection: A dual-track data acquisition mode that combines automatic data collection by smart meters with manual reporting by users ensures the comprehensiveness and accuracy of data, adapts to different user scenarios, and reduces blind spots in data collection.

[0148] Fine-grained tiered electricity price management: Build a dynamically updated tiered electricity price strategy database to achieve accurate matching of user electricity usage types and electricity price strategies, ensuring fair and transparent electricity fee calculations.

[0149] Machine learning-driven early warning optimization: Advanced machine learning algorithms such as XGBoost are introduced to dynamically adjust the early warning threshold according to the user's historical electricity consumption behavior, seasonal patterns, and holiday effects, so as to achieve personalized and intelligent early warning.

[0150] Multi-dimensional warning strategy: Through the gradient warning mechanism and personalized threshold adjustment algorithm, timely and appropriate warning information is provided to avoid excessive or insufficient warnings and improve user satisfaction.

[0151] Interactive user interface and multi-channel push: Create a user-friendly interface to intuitively display electricity consumption information and power saving suggestions, and ensure the efficient communication of early warning information through multiple channels such as SMS, email, APP push, etc., to enhance user interaction and response.

[0152] System security and data protection: Implement strict data encryption and privacy protection measures to ensure the security of user data in all aspects of collection, processing, storage and transmission.

[0153] The implementation of the technical solution not only solved the information asymmetry problem faced by low-voltage users under the tiered electricity price system, but also promoted the transformation of users' energy-saving behavior through technological empowerment, providing strong technical support for building a green, intelligent and efficient modern energy management system.

Claims

1. The method for calculating the remaining power of the tiered electricity price for low-voltage users is characterized by: Includes steps: Obtain user information of target low-voltage users through automatic collection by smart meters or manual reporting by users; determine the current electricity consumption of users this month; According to the user's electricity consumption type, query and apply the corresponding tiered electricity price strategy, which includes different tiered electricity prices and corresponding electricity consumption thresholds; Calculate the target step electricity consumption, that is, the next step electricity threshold corresponding to the current electricity consumption, and the corresponding step electricity price; Introducing machine learning algorithms to dynamically adjust the warning ratio threshold based on users' historical electricity consumption behavior, seasonal electricity consumption patterns, and holiday factors to ensure that warnings are both timely and not too frequent; Set threshold adjustment rules. When it is predicted that a user will enter the next tiered electricity price range within the first preset time, the warning threshold is gradually lowered in advance to warn the user in advance. Real-time calculation of the remaining electricity at the tiered electricity price, i.e. the target tiered electricity consumption minus the current electricity consumption; According to the current electricity consumption period and tiered electricity prices, the current electricity charges are accurately calculated, including the details of segmented billing; When the electricity consumption ratio reaches the dynamically adjusted warning threshold, warning information is automatically generated; Warning information will be delivered to users via SMS, email, APP push or smart speaker voice communication channels.

2. The method for calculating the remaining power of the tiered electricity price for low-voltage users according to claim 1 is characterized in that: Introducing machine learning algorithms to dynamically adjust the warning ratio threshold based on the user's historical electricity consumption behavior, seasonal electricity consumption patterns, and holiday factors. Specifically, it includes: collecting users' electricity consumption data, cleaning abnormal data, filling missing values ​​in the data preprocessing stage, and standardizing time series data to facilitate machine learning model training; Based on the collected data, a feature set for machine learning is constructed. The key features include: Historical electricity consumption trends, seasonal patterns, holiday effects, user category impacts, and time series characteristics.

3. The method for calculating the remaining power of the tiered electricity price for low-voltage users according to claim 1 is characterized in that: Introduce machine learning algorithms to dynamically adjust the warning ratio threshold based on the user's historical electricity consumption behavior, seasonal electricity consumption patterns, and holiday factors. The specifics include: Model selection and training: Select the XGBoost model and use historical data to train the model. The goal is to predict the user's electricity consumption in a certain period of time in the future and determine when it is close to the next step electricity threshold. During the training process, use cross-validation and grid search technology to optimize model parameters to ensure the accuracy of the prediction; After the model predicts the user's future electricity consumption trend, the warning threshold is dynamically adjusted according to the prediction results. The threshold is set as follows: when it is predicted that there is a 90% probability that the user will reach the threshold within the next week, a warning will be issued.

4. The method for calculating the remaining power of the tiered electricity price for low-voltage users according to claim 1 is characterized in that: The user information of the target low-voltage user includes user ID, electricity usage type, and electricity usage time period.

5. The method for calculating the remaining power of the tiered electricity price for low-voltage users according to claim 1 is characterized in that: The warning information includes remaining power, estimated upgrade time, and recommended power-saving measures.

6. The method for calculating the remaining power of the tiered electricity price for low-voltage users according to claim 1 is characterized in that: The "first preset time" is set as a configurable parameter, which is set to 1 week, 2 weeks or one month before the user expects to reach the next step power threshold.

7. A system for calculating the remaining power of a low-voltage user's tiered electricity price, characterized in that: Used to execute the calculation method of the tiered electricity price remaining power for low-voltage users as described in claim 1.

Images