Battery swapping recommendation method and system based on user behavior and load, and storage medium

Abstract

The application discloses a battery replacement recommendation method and system based on user behavior and load, and a storage medium, which comprises the following steps: obtaining electricity data of user battery replacement, user location data, electricity data of battery replacement of a battery replacement site, battery capacity data of the battery replacement site, location data of the battery replacement site and time data required by the user for battery replacement; processing and analyzing the obtained data to obtain input values of features of an online learning recommendation model based on a UCB algorithm; inputting the features into the online learning recommendation model based on the UCB algorithm which has been trained to recommend the location of an optimal battery replacement site for the user at this time; feeding back information according to the recommended battery replacement site of the user, correcting parameters of the online learning recommendation model based on the UCB algorithm, and recommending a more optimal battery replacement site for the user in the next battery replacement; and providing the user with a battery replacement site selection that can save electricity charges most when the user replaces the battery, thereby reducing the battery replacement cost of the user and maintaining the economic benefit requirement of the battery replacement site.

Description

Technical Field

[0001] This invention relates to the field of battery swapping technology, and in particular to a battery swapping recommendation method based on user behavior and load. Background Technology

[0002] With the increasing popularity of two-wheeled battery swapping, users are paying more and more attention to the selection of battery swapping stations. Existing battery swapping services typically lack an intelligent recommendation solution that comprehensively considers user behavior patterns, station load conditions, and electricity price fluctuations. Furthermore, traditional static recommendation methods cannot adapt to rapidly changing market conditions and user needs. At the same time, the distance between different battery swapping stations also affects the user's total cost. Therefore, an intelligent recommendation system that can comprehensively consider these factors is needed, while also taking into account the electricity cost and utilization rate of the battery swapping stations, satisfying both user needs and the economic requirements of the battery swapping stations. Summary of the Invention

[0003] Purpose of the invention: In order to overcome the shortcomings of the existing technology, the present invention provides a battery swapping recommendation method, system and storage medium based on user behavior and load. By collecting data and using advanced online learning algorithms, the recommendation strategy is dynamically adjusted to recommend better battery swapping sites to users.

[0004] Technical solution: To achieve the above objectives, the present invention provides a battery swapping recommendation method based on user behavior and load, comprising the following steps:

[0005] Step 1: Obtain user's battery swapping electricity cost data, user location data, battery swapping station's battery swapping electricity cost data, battery swapping station's power consumption data, battery swapping station's location data, number of battery swaps at the battery swapping station, and user's required battery swapping time data;

[0006] Step 2: Calculate the average electricity cost data of users' battery swapping over a period of time, the average electricity cost data of battery swapping at each battery swapping station, and the average battery swapping frequency at each battery swapping station. Calculate the utilization rate of the battery swapping station based on the battery swapping station's electricity consumption data and swapping frequency.

[0007] Step 3: Calculate the distance between the user's current location and the locations of each battery swapping station, and correlate the distance between the user's current location and the locations of each battery swapping station, the power data in each battery swapping station, and the battery swapping frequency to obtain several battery swapping behavior correlation data.

[0008] Step 4: Initialize the online learning recommendation model based on the UCB algorithm. Input the average electricity cost of user battery swapping, the average electricity cost of battery swapping at each battery swapping station, the battery swapping frequency of each battery swapping station, the utilization rate of the battery swapping station, the battery swapping behavior association data, and the user's battery swapping time data as features into the trained online learning recommendation model based on the UCB algorithm, and recommend the optimal location of the battery swapping station for the user at this time.

[0009] Step 5: Based on the feedback information from the user's recommended battery swapping sites, the parameters of the online learning recommendation model based on the UCB algorithm are corrected, and a better battery swapping site is recommended for the user's next battery swap.

[0010] Furthermore, in step two, the electricity cost variation pattern is obtained by analyzing the electricity expenditures of users at several battery swapping stations and at different time periods. The average value of user electricity cost data within each time period T and the average value of electricity cost data at each battery swapping station are calculated as follows:

[0011]

[0012] In the formula, E s，t E represents the electricity cost of a battery swapping station s at time t. u，t This represents the electricity cost for user u at time t.

[0013] The battery swapping frequency F of the battery swapping station s within each time period T s，T The calculation process is as follows:

[0014]

[0015] In the formula, F s，T R represents the battery swapping frequency of battery swapping station s within time period T. s，t This represents the number of battery swaps performed at battery swapping station s at time t.

[0016] Furthermore, in step two, a power supply and demand balance index (BAL) is defined based on the number of battery swaps and the power consumption data of each battery swapping station. s，t This refers to the utilization rate of the battery swapping station, calculated using the following formula:

[0017] BAL s，t =B s，t -R s，t

[0018] In the formula, B s，t This represents the number of batteries at battery swapping station s at time t.

[0019] Furthermore, in step three, the distance D from the current location of user u to each battery swapping station s is calculated based on the user location data and the battery swapping station location data. u，s , distance D u，s Power consumption data and battery swapping frequency F at battery swapping stations s，t The associated data of battery swapping behavior is obtained by performing related operations. The calculation process is as follows:

[0020]

[0021] In the formula, D u，s Let α represent the distance from user u to battery swapping station s at time t; α is the weighting parameter; B s，t This represents the number of batteries at battery swapping station s at time t, which is equivalent to the amount of electricity at battery swapping station s at time t.

[0022] Furthermore, correlation analysis and feature importance assessment are performed on all features input to the UCB-based online learning recommendation model to screen out features that have an impact on the recommendation results of the UCB-based online learning recommendation model; the correlation between all features is calculated, and the correlation between two features is measured by the Pearson correlation coefficient. When the absolute value of the calculated correlation value is greater than 0.8, it is determined that the correlation between the two features is too high, and one of the two features is removed.

[0023] Sort the features based on the standardized coefficients obtained from fitting all features, and obtain the final feature ranking y; based on feature x... n coefficient β n The magnitude of the absolute value after standardization reflects the characteristic x n The importance of the coefficient β n If the value is close to 0, the feature importance is too low, and this feature should be removed.

[0024] y = β0 + β1x1 + β2x2 + ... + β n x n +∈

[0025] In the formula, xn is the nth feature, βn is the coefficient of the nth feature, and ∈ is the error term.

[0026] Furthermore, in step four, the online learning recommendation model based on the UCB algorithm is initialized to initialize the number of attempts Ns(0) = 0 and the cumulative reward Hs(0) = 0 for each battery swapping station s;

[0027] At each time step t, for each battery swapping station s, calculate its average reward.

[0028]

[0029] Calculate the upper bound of the confidence interval UCBs(t):

[0030]

[0031] In the formula, c is a hyperparameter that controls the degree of exploration, and c>0; For the standard UCB exploration reward item, Ds is the distance from the user's current location to the battery swapping station s; λ is the distance penalty coefficient, and λ>0.

[0032] Furthermore, step four, which involves recommending battery swapping sites based on input features using an online learning recommendation model based on the UCB algorithm, includes the following steps:

[0033] S1. Input the acquired features into the online learning recommendation model based on the UCB algorithm for analysis, and select the battery swapping station according to the upper bound of the set confidence interval UCBs(t);

[0034] S2. Obtain the selected battery swapping sites s* = argmaxUCBs(t) and recommend them to the user;

[0035] S3. Users go to the battery swapping station s* to swap batteries and provide feedback on the actual cost after the swap.

[0036] S4. Generate expense data based on actual cost feedback. And update the number of attempts at the battery swapping station. and cumulative rewards

[0037] S5. Increment the duration step by t = t + 1, and repeat the above steps until the stopping condition is met;

[0038]

[0039] In the formula, Ethreshold is the electricity cost threshold, and Es(t) is the electricity cost of the battery swapping station.

[0040] Furthermore, the user feedback information includes accepting recommended battery swapping sites and rejecting recommended battery swapping sites; when a user accepts a recommended battery swapping site, it is recorded and stored based on the location of the recommended battery swapping site, the average electricity cost of the battery swapping site, the number of battery swaps at the battery swapping site, and the user's ID information; and based on the information of this recommendation, when the user swaps batteries again, it recommends the same or similar battery swapping sites to the user.

[0041] When a user refuses to accept a recommended battery swapping site, the system provides the reason for the refusal and sends feedback; it stores the user's feedback, the recommended information, and the user's ID; it analyzes the user's feedback and generates relevant feedback data; it adjusts the relevant parameters in the recommendation model based on the feedback data, and recommends battery swapping sites to the user based on the adjusted recommendation model.

[0042] Furthermore, a battery swapping recommendation system based on user behavior and load is provided. This system is used to implement the aforementioned electricity cost optimization battery swapping recommendation method based on the UCB algorithm, including:

[0043] The data collection module is used to acquire user electricity cost data for battery swapping, user location data, electricity cost data for battery swapping stations, power consumption data for battery swapping stations, location data for battery swapping stations, number of battery swaps at battery swapping stations, and time data when users need to swap batteries. The data processing and analysis module is used to analyze and process the acquired data. The online learning recommendation algorithm calculation module based on multi-armed slot machine (MAS) is used to recommend the optimal battery swapping station based on the processed data. The user feedback module is used to feed user feedback information back to the online learning recommendation algorithm calculation module based on multi-armed slot machine (MAS).

[0044] Furthermore, a storage medium storing an executable program, which, when executed by a processor, can implement the aforementioned battery swapping recommendation method based on user behavior and load.

[0045] Beneficial effects: The battery swapping recommendation method, system, and storage medium of this invention, based on user behavior and load, collect various battery swapping electricity cost data and battery swapping frequency data, and use advanced online learning algorithms to dynamically adjust the recommendation strategy to optimize the user's battery swapping cost; it can continuously learn and improve based on real-time user feedback, providing users with the most cost-effective battery swapping station selection; it introduces a distance penalty term to comprehensively consider the distance factor of the battery swapping station to achieve globally optimal recommendation; while reducing the user's battery swapping cost, it maintains the economic benefit requirements of the battery swapping station. Attached Figure Description

[0046] Figure 1 A flowchart illustrating a battery swapping recommendation method based on user behavior and load. Detailed Implementation

[0047] The invention will now be further described with reference to the accompanying drawings.

[0048] like Figure 1 As shown, the battery swapping recommendation method based on user behavior and load includes the following steps:

[0049] Step 1: Obtain data including user battery swapping electricity costs, user location data, battery swapping station electricity costs, battery swapping station power consumption data, battery swapping station location data, number of battery swaps at each station, and the time when the user needs to swap batteries. User battery swapping electricity costs record the specific electricity cost for each swap, calculated as electricity consumption multiplied by the unit electricity price of the battery swapping station. User location data shows the user's real-time location when they need to swap batteries. Battery swapping station electricity costs are recorded for each swapping station. Battery swapping station power consumption data shows the available power at each swapping station, calculated using the number of batteries available for swapping at the station, which represents the load at the swapping station. Battery swapping station location data shows the location of each swapping station.

[0050] Step 2: Calculate the average electricity cost data for user battery swapping over a period of time, the average electricity cost data for battery swapping at each battery swapping station, and the average battery swapping frequency at each battery swapping station. Calculate the battery swapping station utilization rate based on the station's electricity consumption data and swapping frequency. Use the data processing and analysis module to analyze electricity costs, number of battery swaps, and station load.

[0051] Step 3: Calculate the distance between the user's current location and the locations of each battery swapping station, and correlate the distance between the user's current location and the locations of each battery swapping station, the power data in each battery swapping station, and the battery swapping frequency to obtain several battery swapping behavior correlation data.

[0052] Step 4: Initialize the online learning recommendation model based on the UCB algorithm. Train the model on the historical data training set and evaluate its generalization ability through cross-validation to avoid overfitting. Evaluate the model's performance through A / B testing, using key metrics such as recommendation accuracy, user click-through rate, and electricity cost savings to compare the effects of different models and select the best model. Input the average electricity cost for user battery swapping, the average electricity cost for battery swapping at various battery swapping stations, the battery swapping frequency at each station, the utilization rate of battery swapping stations, battery swapping behavior correlation data, and user battery swapping time data as features into the trained UCB-based recommendation model. The algorithm's online learning recommendation model recommends the optimal battery swapping station location for the user at any given time; it also recommends the optimal battery swapping station when the user needs to swap batteries, and can recommend the optimal station based on the user's chosen swapping time; the online learning recommendation model based on the UCB algorithm, where the UCB algorithm is an algorithm that selects the best arm by calculating the upper confidence bound of each arm, is a confidence upper bound algorithm, and the upper confidence bound is calculated based on the arm's historical returns and the number of times it has been selected; it is a heuristic method for solving the multi-armed slot machine problem, namely, the calculation module of the online learning recommendation algorithm based on multi-armed slot machines.

[0053] Step 5: Feedback information on the user-recommended battery swapping sites is sent to the user feedback module to correct the parameters of the online learning recommendation model based on the UCB algorithm, and to recommend a better battery swapping site for the user's next battery swap.

[0054] Real-time data is acquired, including user battery swapping electricity costs, user location data, battery swapping station electricity costs, battery swapping station power consumption, battery swapping station swapping frequency, and battery swapping station location data. This data is then cleaned and normalized. The collected raw data is cleaned to remove duplicate records, fill in missing values, and handle outliers. For example, for missing electricity cost data, mean-based filling or interpolation using data from adjacent time points can be used. Simultaneously, statistical methods are used to identify and remove obviously abnormal electricity costs or battery swapping frequency to ensure data accuracy and reliability. Data from different sources is standardized to allow comparisons under the same dimensions. For numerical features such as electricity costs, normalization or standardization methods are used to convert the data into values ​​between 0 and 1, or a distribution with a mean of 0 and a variance of 1, to improve the model's convergence speed and accuracy.

[0055] The system comprises a data collection module, a data processing and analysis module, a multi-armed slot machine-based online learning recommendation algorithm calculation module, and a user feedback module, forming a battery swapping recommendation system based on user behavior and load. The data collection module gathers data such as user battery swapping electricity costs, user location data, battery swapping station electricity costs, battery swapping station power consumption data, battery swapping station location data, and user battery swapping time. This data is then transmitted to the data processing and analysis module for analysis and processing. The results from this module are used as features and transmitted to the multi-armed slot machine-based online learning recommendation algorithm calculation module and the user feedback module for further calculation and processing. The system recommends optimal battery swapping stations to users. Users provide feedback on the recommended stations through the user feedback module, which in turn optimizes and updates the model. Finally, the multi-armed slot machine-based online learning recommendation algorithm calculation module further recommends battery swapping stations.

[0056] Each time a user swaps batteries, the system updates its data, continuously updating relevant information about the user and the swapping station, including real-time electricity costs and user behavior records. This ensures the recommendation model makes decisions based on the latest data. The system continuously updates user and station data in real time, covering several key aspects. Real-time electricity costs change constantly with time and market supply and demand, accurately recording electricity prices at each station at different times. User behavior records include the user's charging history, the time and frequency of station visits, providing a basis for a comprehensive understanding of user habits. By continuously updating this data, the system ensures the recommendation model always makes decisions based on the latest information, thereby improving the accuracy and usability of recommendations.

[0057] The system also features system performance monitoring: regularly monitoring the performance metrics of each module, including data collection latency and algorithm efficiency. Anomaly detection: an anomaly detection mechanism is implemented to identify abnormal situations in system operation in real time and automatically issue alerts. Feedback mechanism: combining user feedback with system performance data, the recommendation strategy is adjusted regularly to improve user experience. Regular evaluation and optimization: regular system evaluation reports are generated to analyze the accuracy of recommendations, user feedback, and electricity cost savings. Based on the evaluation results, algorithm parameters are dynamically adjusted to adapt to market changes and user needs.

[0058] In step two, the electricity price change pattern is obtained by analyzing the electricity expenditures of users at several battery swapping stations and at different time periods. The average value of user electricity costs within each time period T and the average value of electricity costs at each battery swapping station are calculated as follows:

[0059]

[0060] In the formula, E s，t E represents the electricity cost of a battery swapping station s at time t. u，t This represents the electricity cost for user u at time t.

[0061] The battery swapping frequency F of the battery swapping station s within each time period T s，T The calculation process is as follows:

[0062]

[0063] In the formula, F s，T R represents the battery swapping frequency of battery swapping station s within time period T. s，t This represents the number of battery swaps performed at battery swapping station s at time t.

[0064] In step two, a power supply and demand balance index (BAL) is defined based on the number of battery swaps and the power consumption data of each battery swapping station. s，t This refers to the utilization rate of the battery swapping station. The smaller the value of the power supply and demand balance index, the higher the utilization rate of the battery swapping station. The calculation formula is as follows:

[0065] BAL s，t =B s，t -R s，t

[0066] In the formula, B s，t This represents the number of batteries at battery swapping station s at time t.

[0067] In step three, the distance D from the current location of user u to each battery swapping station s is calculated based on the user location data and the battery swapping station location data. u，s , distance D u，sPower consumption data and battery swapping frequency F at battery swapping stations s，t The associated data of battery swapping behavior is obtained by performing related operations. The calculation process is as follows:

[0068]

[0069] In the formula, D u，s This represents the distance from user u to battery swapping station s at time t; 'a' is a weighting parameter used to adjust the impact of swapping frequency on the overall index; B s，t This represents the number of batteries at battery swapping station s at time t, which is equivalent to the amount of electricity at battery swapping station s at time t.

[0070] Correlation analysis and feature importance assessment are performed on all features input to the UCB-based online learning recommendation model. Features that influence the recommendation results of the UCB-based online learning recommendation model are selected to reduce redundant features and improve the model's performance and interpretability. The correlation between all features is calculated, and the correlation between two features is measured by the Pearson correlation coefficient. If the absolute value of the calculated correlation is greater than 0.8, the two features are considered to be too highly correlated, and one of the two features is removed.

[0071]

[0072] In the formula, x i For the i-th feature x, y i The i-th y-feature and These are the average values ​​of the x-feature and the y-feature, respectively.

[0073] Sort the features based on the standardized coefficients obtained from fitting all features, and obtain the final feature ranking y; based on feature x... n coefficient β n The magnitude of the absolute value after standardization reflects the characteristic x n The importance of the coefficient β n If the value is close to 0, the feature importance is too low, and this feature should be removed.

[0074] y = β0 + β1x1 + β2x2 + ... + β nn x n +∈

[0075] In the formula, xn is the nth feature, βn is the coefficient of the nth feature, and ∈ is the error term, representing model error and other random factors and unexplainable parts.

[0076] In step four, the online learning recommendation model based on the UCB algorithm is initialized. For each battery swapping station s, the number of attempts Ns(0) = 0 and the cumulative reward Hs(0) = 0 are initialized, and the initial parameter of the model is alpha = 0.1.

[0077] At each time step t, for each battery swapping station s, calculate its average reward.

[0078]

[0079] Calculate the upper bound of the confidence interval UCBs(t):

[0080]

[0081] In the formula, c is a hyperparameter that controls the degree of exploration, and c>0; For the standard UCB exploration reward item, Ds is the distance from the user's current location to the battery swapping station s; λ is the distance penalty coefficient, and λ>0, used to balance the relationship between electricity savings and driving costs.

[0082] Step four, which involves recommending battery swapping sites based on input features using an online learning recommendation model based on the UCB algorithm, includes the following steps:

[0083] S1. Input the acquired features into the online learning recommendation model based on the UCB algorithm for analysis, and select the battery swapping station according to the upper bound of the set confidence interval UCBs(t);

[0084] S2. Obtain the selected battery swapping sites s* = argmaxUCBs(t) and recommend them to the user;

[0085] S3. Users go to the battery swapping station s* to swap batteries and provide feedback on the actual cost after the swap.

[0086] S4. Generate expense data based on actual cost feedback. And update the number of attempts at the battery swapping station. and cumulative rewards

[0087] S5. Increment the duration step by t = t + 1, and repeat the above steps until the stopping condition is met;

[0088]

[0089] In the formula, Ethreshold is the electricity cost threshold, which is a predetermined reference value used to measure whether the electricity cost of the battery swapping station is within an acceptable range; Es(t) is the electricity expenditure of the battery swapping station; when the stopping condition is met, the output location of the battery swapping station is based on the reduction of the user's battery swapping cost and ensuring that the battery swapping cost of the station will not cause the station to operate at a loss.

[0090] The user feedback information includes accepting recommended battery swapping sites and rejecting recommended battery swapping sites. When a user accepts a recommended battery swapping site, the information is recorded and stored based on the site's location, average electricity cost, real-time electricity cost, number of swaps, and the user's ID. Furthermore, based on this recommendation, the system recommends similar or identical battery swapping sites to the user the next time they need to swap batteries. The recommended site location, average electricity cost, real-time electricity cost, number of swaps, and user ID are used as features input to an online learning recommendation model based on the UCB algorithm. Based on this data, the model parameters are updated by minimizing the loss function, thus re-recommending a new optimal battery swapping site for the user's next battery swap.

[0091] When a user refuses a recommended battery swapping station, the system provides the reason for the refusal and feedback. The reason might be that the station is too far away, the cost of swapping is too high, or the station is too crowded. The system stores the user's feedback, the recommended information, and the user's ID. It analyzes the user's feedback and generates relevant feedback data. Based on this feedback data, it adjusts the relevant parameters in the recommendation model and recommends new battery swapping stations to the user. The system not only records the reason for refusal but also uses all data collected during this period as features input into an online learning recommendation model based on the UCB algorithm. During model updates, the reason for refusal is incorporated as a constraint into the optimization algorithm. For example, if the reason for dissatisfaction is that a station is too far away, a distance-related penalty term can be added to the loss function; or the model parameters can be optimized by adjusting the coefficients of the penalty and reward terms. The model adjusts its parameters by minimizing the new loss function to avoid recurring reasons that lead to user dissatisfaction, and then recommends new, optimal battery swapping station options.

[0092] A battery swapping recommendation system based on user behavior and load, the system being used to implement the aforementioned battery swapping recommendation method based on user behavior and load, is characterized by comprising:

[0093] The data collection module is used to acquire user electricity cost data for battery swapping, user location data, electricity cost data for battery swapping stations, power consumption data for battery swapping stations, location data for battery swapping stations, number of battery swaps at battery swapping stations, and user battery swapping time data. The data processing and analysis module is used to analyze and process the acquired data. The multi-armed slot machine-based online learning recommendation algorithm calculation module is used to recommend the optimal battery swapping station based on the processed data. The user feedback module is used to feed user feedback information back to the multi-armed slot machine-based online learning recommendation algorithm calculation module and optimize the model in the multi-armed slot machine-based online learning recommendation algorithm calculation module.

[0094] A storage medium storing an executable program, which, when executed by a processor, enables the aforementioned battery swapping recommendation method based on user behavior and load.

[0095] User feedback is used as new data points to input into the online learning algorithm, updating model parameters to reflect the latest user preferences and market dynamics. This feedback loop ensures continuous optimization of the model over long-term operation. Simultaneously, user profiles are built based on user feedback to achieve personalized recommendations; for example, users who frequently swap batteries during off-peak hours can be prioritized for recommendations of sites with lower electricity costs. Based on recorded user ID information, the parameters in the recommendation model are adjusted differently for different users, resulting in customized battery swapping site recommendations. This provides dynamic and personalized battery swapping recommendations, optimizing the user's battery swapping experience.

[0096] When recommending the optimal battery swapping site to users, we consider the discounts on users' electricity costs and the reduction in distance to the site. We also need to consider the site's utilization rate and electricity costs. Different sites have different electricity costs due to their location, and varying total capacity and number of swaps will affect their utilization rates. Therefore, while recommending sites with the lowest possible swapping cost, we must also consider site utilization and electricity costs. Using an online learning recommendation model based on the UCB algorithm, we recommend sites that are close to users, have low swapping costs, and low utilization rates. This approach improves site utilization, reduces user swapping costs, and avoids congestion during swapping sessions – this is the optimal recommendation solution. Of course, we also need to avoid recommending sites based on users' chosen swapping times to prevent congestion during peak swapping periods.

[0097] The collected data, after analysis and processing, is used as features input into an online learning recommendation model based on the UCB algorithm, significantly reducing users' battery swapping costs. Through intelligent recommendation algorithms, users can find nearby battery swapping stations during periods of lower electricity prices, effectively reducing overall battery swapping costs; the average battery swapping cost for users during the pilot period is predicted to be reduced by approximately 15% compared to before. Simultaneously, battery swapping waiting times are shortened. Because the system can monitor station load in real time and recommend stations with lower loads accordingly, user waiting times at battery swapping stations are significantly reduced, improving battery swapping efficiency. User satisfaction is improved. According to user feedback collected by the system, over 80% of users are satisfied with the new system's recommendations, especially those who frequently swap batteries at night, who generally believe the recommended stations better suit their travel habits. Operator revenue increases: Due to optimized station utilization, operators can increase revenue through more rational resource allocation; at the same time, by reducing electricity expenses, operators' overall operating costs also decrease. Contribution to energy conservation and emission reduction: By guiding users to swap batteries during off-peak hours, this system indirectly promotes the balanced use of electricity resources, helping to reduce energy waste during peak grid load periods and playing a positive role in environmental protection. Resource utilization is improved. Through dynamic analysis of the load status of battery swapping stations, the system helps operators to better allocate resources and avoid situations where some stations are overcrowded while other stations have low utilization rates.

[0098] The above description is merely a preferred embodiment of the present invention. Those skilled in the art can make several modifications and optimizations based on the above disclosure without departing from the basic principles described above. These modifications and optimizations should be considered within the scope of protection as understood by the present invention.

Claims

1. A battery swapping recommendation method based on user behavior and load, characterized in that: Includes the following steps: Step 1: Obtain user's battery swapping electricity cost data, user location data, battery swapping station's battery swapping electricity cost data, battery swapping station's power consumption data, battery swapping station's location data, number of battery swaps at the battery swapping station, and user's required battery swapping time data; Step 2: Calculate the average electricity cost data of users' battery swapping over a period of time, the average electricity cost data of battery swapping at each battery swapping station, and the average battery swapping frequency at each battery swapping station. Calculate the utilization rate of the battery swapping station based on the battery swapping station's electricity consumption data and swapping frequency. Step 3: Calculate the distance between the user's current location and the locations of each battery swapping station, and correlate the distance between the user's current location and the locations of each battery swapping station, the power data in each battery swapping station, and the battery swapping frequency to obtain several battery swapping behavior correlation data. Step 4: Initialize the online learning recommendation model based on the UCB algorithm. Input the average electricity cost of user battery swapping, the average electricity cost of battery swapping at each battery swapping station, the battery swapping frequency of each battery swapping station, the utilization rate of the battery swapping station, the battery swapping behavior association data, and the user's battery swapping time data as features into the trained online learning recommendation model based on the UCB algorithm, and recommend the optimal location of the battery swapping station for the user at this time. Step 5: Based on the feedback information from the user's recommended battery swapping sites, the parameters of the online learning recommendation model based on the UCB algorithm are corrected, and a better battery swapping site is recommended for the user's next battery swap.

2. The battery swapping recommendation method based on user behavior and load according to claim 1, characterized in that: In step two, the electricity price change pattern is obtained by analyzing the electricity expenditures of users at several battery swapping stations and at different time periods. The average value of user electricity costs within each time period T and the average value of electricity costs at each battery swapping station are calculated as follows: In the formula, E s，t E represents the electricity cost of a battery swapping station s at time t. u，t This represents the electricity cost for user u at time t. The battery swapping frequency F of the battery swapping station s within each time period T s，T The calculation process is as follows: In the formula, F s，T R represents the battery swapping frequency of battery swapping station s within time period T. s，t This represents the number of battery swaps performed at battery swapping station s at time t.

3. The battery swapping recommendation method based on user behavior and load according to claim 1, characterized in that: In step two, a power supply and demand balance index (BAL) is defined based on the number of battery swaps and the power consumption data of each battery swapping station. s，t This refers to the utilization rate of the battery swapping station, calculated using the following formula: BAL s，t =B s，t -R s，t In the formula, B s，t This represents the number of batteries at battery swapping station s at time t.

4. The battery swapping recommendation method based on user behavior and load according to claim 1, characterized in that: In step three, the distance D from the current location of user u to each battery swapping station s is calculated based on the user location data and the battery swapping station location data. u，s , distance D u，s Power consumption data and battery swapping frequency F at battery swapping stations s，t The associated data of battery swapping behavior is obtained by performing related operations. The calculation process is as follows: In the formula, D u，s Let α represent the distance from user u to battery swapping station s at time t; α is the weighting parameter; B s，t This represents the number of batteries at battery swapping station s at time t, which is equivalent to the amount of electricity at battery swapping station s at time t.

5. The battery swapping recommendation method based on user behavior and load according to claim 4, characterized in that: Relevance analysis and feature importance assessment are performed on all features input to the UCB-based online learning recommendation model to screen out features that have an impact on the recommendation results of the UCB-based online learning recommendation model; the correlation between all features is calculated, and the correlation between two features is measured by the Pearson correlation coefficient. When the absolute value of the calculated correlation value is greater than 0.8, it is determined that the correlation between the two features is too high, and one of the two features is removed. Sort the features based on the standardized coefficients obtained from fitting all features, and obtain the final feature ranking y; based on feature x... n coefficient β n The magnitude of the absolute value after standardization reflects the characteristic x n The importance of the coefficient β n If the value is close to 0, the feature importance is too low, and this feature should be removed. y=β0+β1x1+β2x2+…+β n x n +∈ In the formula, xn is the nth feature, βn is the coefficient of the nth feature, and ∈ is the error term.

6. The battery swapping recommendation method based on user behavior and load according to claim 1, characterized in that: In step four, the online learning recommendation model based on the UCB algorithm is initialized to initialize the number of attempts Ns(0) = 0 and the cumulative reward Hs(0) = 0 for each battery swapping station s; At each time step t, for each battery swapping station s, calculate its average reward. Calculate the upper bound of the confidence interval UCBs(t): In the formula, c is a hyperparameter that controls the degree of exploration, and c>0; For the standard UCB exploration reward item, Ds is the distance from the user's current location to the battery swapping station s; λ is the distance penalty coefficient, and λ>

0.

7. The battery swapping recommendation method based on user behavior and load according to claim 6, characterized in that: Step four, which involves recommending battery swapping sites based on input features using an online learning recommendation model based on the UCB algorithm, includes the following steps: S1. Input the acquired features into the online learning recommendation model based on the UCB algorithm for analysis, and select the battery swapping station according to the upper bound of the set confidence interval UCBs(t); S2. Obtain the selected battery swapping sites s* = argmaxUCBs(t) and recommend them to the user; S3. Users go to the battery swapping station s* to swap batteries and provide feedback on the actual cost after the swap. S4. Generate expense data based on actual cost feedback. And update the number of attempts at the battery swapping station. and cumulative rewards S5. Increment the duration step by t = t + 1, and repeat the above steps until the stopping condition is met; In the formula, Ethreshold is the electricity cost threshold, and Es(t) is the electricity cost of the battery swapping station.

8. The battery swapping recommendation method based on user behavior and load according to claim 1, characterized in that: The user feedback information includes accepting recommended battery swapping sites and rejecting recommended battery swapping sites; when a user accepts a recommended battery swapping site, the information is recorded and stored based on the location of the recommended battery swapping site, the average electricity cost of the battery swapping site, the number of battery swaps at the site, and the user's ID information. Based on the information provided in this recommendation, the system will recommend the same or similar battery swapping sites to the user the next time they need to swap batteries. When a user refuses to accept the recommended battery swapping site, provide the reason for the refusal and feedback. Store user feedback, recommendations, and user IDs; analyze user feedback and generate relevant feedback data. Based on the feedback data, the relevant parameters in the recommendation model are adjusted, and the adjusted recommendation model is used to recommend battery swapping sites to the user.

9. A battery swapping recommendation system based on user behavior and load, the system being used to implement the battery swapping recommendation method based on user behavior and load as described in any one of claims 1 to 8, characterized in that: include: Data collection module: used to acquire user battery swapping electricity cost data, user location data, battery swapping station electricity cost data, battery swapping station power consumption data, battery swapping station location data, number of battery swaps at the battery swapping station, and user battery swapping time data; Data processing and analysis module: used to analyze and process the acquired data; The online learning recommendation algorithm calculation module based on multi-armed slot machines is used to recommend the optimal battery swapping site based on the processed data. User feedback module: Used to feed user feedback information back to the online learning recommendation algorithm calculation module based on multi-armed slot machines.

10. A storage medium, characterized in that: It stores an executable program, which, when executed by a processor, can implement the battery swapping recommendation method based on user behavior and load as described in any one of claims 1 to 8.

Images