Unattended hydrogen ion battery charging station

Abstract

The invention discloses an unattended hydrogen ion battery charging station, and relates to the technical field of energy management, and the unattended hydrogen ion battery charging station comprises the steps: monitoring battery state parameters and environmental conditions through a sensor, and collecting multi-dimensional data; transmitting the multi-dimensional data to an automatic operation platform for cleaning and feature extraction, and training an adaptive learning model; determining a charging strategy by applying an optimization algorithm according to the charging demand prediction result; the charging strategy is converted into a control instruction to be sent to a controller in the automatic operation platform, and the controller adjusts the working mode of the charging equipment after receiving the instruction; the running state of the battery is detected in real time, and when a potential fault is detected, an early warning mechanism is triggered; according to the method, the optimal charging strategy is determined through the differential evolution algorithm, multiple factors such as the charging demand intensity, the power level and the battery health are comprehensively considered, the global optimal solution is quickly found, the charging efficiency and the resource utilization rate are improved, and the operation cost is reduced.

Description

Technical Field

[0001] The invention relates to the technical field of energy management, in particular to an unattended hydrogen ion battery charging station. Background Art

[0002] With the global emphasis on renewable energy and low-carbon economy, the development of electric vehicles (EVs) and their supporting infrastructure has become an important part of the energy policies of various countries. As a new type of energy storage device, hydrogen-ion batteries have shown significant advantages in energy density, charging speed and environmental friendliness, and have therefore received widespread attention in the electric vehicle industry. At the same time, the advancement of smart grid technology has provided solid technical support for distributed energy management, making power distribution more efficient and flexible. However, most traditional charging stations rely on manual operation, which not only increases operating costs, but also makes it difficult to meet the growing demand for automation and service quality requirements. In recent years, unmanned charging stations have gradually become a research hotspot. They combine advanced technologies such as the Internet of Things (IoT), big data analysis, and machine learning to achieve full automation of the charging process, improve charging efficiency, and reduce maintenance costs.

[0003] In the existing technology, although there are some unattended charging solutions, they show obvious limitations when dealing with complex and changeable actual application scenarios. Traditional charging methods usually use fixed charging algorithms. This approach cannot dynamically adjust the charging strategy according to the battery status and external environment. The charging strategy has poor flexibility, which may lead to longer charging time or shorter battery life. Insufficient data quality and information security is another key issue that needs to be solved urgently. Most current charging methods fail to fully consider the importance of data encryption during communication. Once a network attack occurs, the user's sensitive information, such as charging records and personal identity information, may be leaked. When the equipment has an abnormal condition, it is often not discovered and processed in time, which in turn affects the stable operation of the entire system. Existing unattended charging stations rarely involve the application of deep learning models, and there is still insufficient research on how to use historical data for accurate prediction and optimization decision-making, which limits the further improvement of their intelligence level and service capabilities. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides an unattended hydrogen ion battery charging station to solve the problems of insufficient data quality and information security and poor charging strategy flexibility in the prior art.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides an unattended hydrogen ion battery charging station, which includes monitoring battery status parameters and environmental conditions through sensors and collecting multi-dimensional data; the specific steps of monitoring battery status parameters and environmental conditions through sensors and collecting multi-dimensional data are as follows:

[0008] Select and install current, voltage, temperature, humidity, light intensity, and grid load sensors;

[0009] Configure the sensor's operating mode, set the measurement range and accuracy;

[0010] Connect the sensor to the automated operation platform, configure the communication protocol and set the sampling frequency;

[0011] The automated operation platform triggers data collection at regular intervals to obtain multi-dimensional data on current, voltage, temperature, internal resistance, charging rate, discharge depth, and grid load conditions from sensors.

[0012] As a preferred solution of the unattended hydrogen ion battery charging station of the present invention, the multi-dimensional data is transmitted to the automated operation platform for cleaning and feature extraction, and the adaptive learning model is trained. The specific steps are:

[0013] The sensor data is transmitted to the automated operation platform in real time using an encrypted communication protocol. The automated operation platform receives and stores data packets through HTTP protocol POST requests.

[0014] Apply low-pass filtering and smoothing, and store the packets in a distributed database;

[0015] The automated operation platform identifies data packets in the distributed database and removes outliers and fills in missing values;

[0016] Extract features from processed data packets and select predictive features to optimize the input of the adaptive learning model;

[0017] Initialize the long short-term memory model network parameters, set the initial learning rate and batch size, use historical data to train the adaptive learning model, and dynamically adjust the hyperparameters.

[0018] As a preferred solution of the unattended hydrogen ion battery charging station of the present invention, the adaptive learning model based on training automatically adjusts the prediction parameters according to the real-time environmental changes, and the specific steps are as follows:

[0019] Extract historical data and real-time sensor data from distributed databases and perform normalization and standardization;

[0020] Combining historical feature changes and current environmental impacts, the i-th predicted feature value f of the historical data and real-time sensor data that have been normalized and standardized over time is i (t), the jth environmental factor value e that affects charging demand in the real-time environment j and the weight w corresponding to the i-th prediction feature i Introduce an expression to calculate the dynamically adjusted hyperparameter α′;

[0021] Using the initial hyperparameter α, the future charging demand is predicted through the long short-term memory model, and the dynamically adjusted hyperparameter α′ is optimized;

[0022] Apply the obtained dynamically adjusted hyperparameter α′ to the initial parameter setting of the optimization algorithm;

[0023] Update the population parameters in the optimization algorithm, and dynamically adjust the initial learning rate and batch size based on the dynamic adjustment hyperparameter α′.

[0024] As a preferred solution of the unattended hydrogen ion battery charging station of the present invention, wherein: the charging strategy is determined by applying an optimization algorithm according to the charging demand prediction result, and the specific steps are:

[0025] Using the dynamically adjusted hyperparameter α′, the charging demand forecast results are generated through the long short-term memory model to obtain the charging demand intensity Q and power level P at the time point;

[0026] Combined with extracting demand characteristic values ​​g that change over time from multi-dimensional data i (t), and assign weight d i Form the characteristic value of total demand;

[0027] Extract the health feature value k that affects charging demand from the real-time environment j (t), and assign weight b j Forming a total health characteristic value;

[0028] The charging demand intensity Q and power level P are introduced into the expression to calculate the optimal charging strategy S * ;

[0029] Differential evolution algorithm is used to solve the charging strategy S * , initialize the population parameters and dynamically adjust them, and iteratively update the population individuals until they converge to the optimal solution.

[0030] As a preferred solution of the unattended hydrogen ion battery charging station of the present invention, wherein: the differential evolution algorithm is used to solve the charging strategy S * , initialize the population parameters and dynamically adjust them, iteratively update the population individuals until they converge to the optimal solution. The specific steps are:

[0031] The automated operation platform generates multiple candidate charging strategies based on the charging demand forecast results;

[0032] The automated operation platform defines comprehensive evaluation criteria, analyzes each candidate charging strategy, evaluates whether it meets charging requirements, grid load balancing, and battery health impact, and draws evaluation results;

[0033] The automated operation platform uses the differential evolution algorithm to iteratively optimize the population, generates new individuals through selection, mutation, and crossover operations, and replaces the original individuals, gradually approaching the optimal solution;

[0034] Improve optimization efficiency and robustness according to the adaptive differential evolution algorithm parameters;

[0035] When the maximum number of iterations is reached, the adaptive differential evolution algorithm stops running and outputs the final optimal charging strategy.

[0036] As a preferred solution of the unattended hydrogen ion battery charging station of the present invention, the charging strategy is converted into a control instruction and sent to a controller in the automated operation platform. After receiving the instruction, the controller adjusts the working mode of the charging device. The specific steps are as follows:

[0037] According to the final optimal charging strategy, it is organized into control instructions in JSON format;

[0038] Use TLS / SSL encryption protocol to send control instructions from the automated operation platform to the controller through the POST request method of the HTTP protocol. The controller receives and parses the instructions and extracts parameters.

[0039] The controller adjusts the working mode of the charging device according to the extracted parameters;

[0040] The controller executes the adjusted working mode and monitors the charging effect through sensors;

[0041] The controller continuously monitors the charging process through sensors. When an abnormal situation is detected, the controller information is fed back to the automated operation platform, and the automated operation platform records the charging data.

[0042] As a preferred solution of the unattended hydrogen ion battery charging station of the present invention, the real-time detection of the operating status of the battery, when a potential fault is detected, triggers the early warning mechanism, the specific steps are:

[0043] The controller continuously monitors the operating parameters of the battery through sensors and transmits the data to the automated operation platform;

[0044] The automated operations platform conducts comparative analysis based on historical data and real-time data to identify abnormal patterns;

[0045] If a potential fault is displayed, the early warning mechanism is triggered;

[0046] If no potential fault is detected, the current charging state is maintained and the battery operating status continues to be monitored;

[0047] When the early warning mechanism is triggered, the automated operation platform pushes it to relevant personnel via email, text message, and application.

[0048] In a second aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the unattended hydrogen ion battery charging station as described in the first aspect of the present invention is implemented.

[0049] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the unattended hydrogen ion battery charging station as described in the first aspect of the present invention is implemented.

[0050] The beneficial effects of the present invention are as follows: by deploying high-precision sensors to monitor battery status and environmental conditions in real time, multi-dimensional data is collected to ensure the quality and reliability of input data, enhance the adaptability of the automated operation platform to complex environments, and achieve instant response to changes to optimize charging strategies. The data is securely transmitted to the automated operation platform using an encrypted communication protocol, cleaned and feature extracted on the platform, and the adaptive learning model is trained using a long short-term memory model, thereby improving the accuracy and security of data analysis. Based on the trained model, the automated operation platform automatically adjusts the prediction parameters according to real-time environmental changes to generate accurate charging demand prediction results, reduce prediction errors and enhance the flexibility and reliability of the automated operation platform. Finally, the optimal charging strategy is determined by a differential evolution algorithm, taking into account multiple factors such as charging demand intensity, power level and battery health, and quickly finding the global optimal solution, thereby improving charging efficiency and resource utilization and reducing operating costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0052] Figure 1 This is a flow chart of the unattended hydrogen ion battery charging station in Example 1.

[0053] Figure 2This is a flow chart for collecting multi-dimensional data in Example 1. DETAILED DESCRIPTION

[0054] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0055] Example 1, reference Figure 1 , which is the first embodiment of the present invention, provides an unattended hydrogen ion battery charging station, comprising the following steps:

[0056] S1. Monitor battery status parameters and environmental conditions through sensors and collect multi-dimensional data.

[0057] Further, current, voltage, temperature, humidity, light intensity and grid load sensors are selected and installed;

[0058] Specifically, in order to ensure that the unattended hydrogen-ion battery charging station can fully monitor the battery status and its operating environment, professional-grade sensors suitable for different application scenarios are selected and accurately installed and calibrated, including:

[0059] Current sensor: used to measure charging current;

[0060] Voltage sensor: used to detect battery terminal voltage;

[0061] Temperature sensor: used to monitor the internal and external temperature of the battery;

[0062] Humidity sensor: used to assess air humidity;

[0063] Light intensity sensor: used to record changes in natural light sources;

[0064] Grid load sensor: used to monitor power supply.

[0065] By selecting and installing current, voltage, temperature, humidity, light intensity and grid load sensors, the automated operation platform achieves accurate measurement of battery status and environmental conditions. High-precision sensors ensure the reliability and accuracy of data, providing a solid foundation for subsequent data analysis. Diverse sensors cover physical quantities and environmental variables, enhancing the environmental adaptability of the automated operation platform and enabling it to maintain stable performance under complex conditions.

[0066] Real-time monitoring and dynamic adjustment capabilities enable the automated operation platform to respond to changes immediately, optimize charging strategies, improve efficiency and reduce energy waste. Comprehensive monitoring can also identify potential problems in advance, trigger early warning mechanisms, and perform preventive maintenance, thereby improving the safety and reliability of the automated operation platform. Ultimately, this approach builds an efficient, intelligent and safe automated operation platform, significantly improving the performance of unmanned hydrogen-ion battery charging stations and reflecting the value of technological innovation.

[0067] Configure the sensor's operating mode, set the measurement range and accuracy;

[0068] It should be noted that by configuring the working mode of the sensor, setting the measurement range and accuracy, accurate monitoring of the battery status and environmental conditions can be achieved, including:

[0069] Current sensor: set the measurement range to 0-50A, accuracy ±0.5%;

[0070] Voltage sensor: set the measurement range to 0-400V, accuracy ±0.2%;

[0071] Temperature sensor: set the measurement range to -40℃ to +125℃, accuracy ±0.3℃;

[0072] Humidity sensor: set the measurement range to 0%-100%RH, accuracy ±2%;

[0073] Light intensity sensor: set the measurement range to 0-200,000 lux, accuracy ±5%;

[0074] Grid load sensor: The measurement range is set to 0-500kW, with an accuracy of ±1%.

[0075] A reasonable measurement range ensures that the sensor works stably under expected operating conditions, while high-precision standards ensure data accuracy, providing a reliable basis for subsequent data analysis. This helps to formulate a scientific and reasonable charging strategy, and setting an appropriate measurement range avoids resource waste, improves resource utilization, reduces unnecessary energy consumption and equipment redundancy, and reduces operating costs.

[0076] High-quality data input enables machine learning models to more accurately predict future demand and behavior patterns, guide more intelligent charging strategies, and improve the intelligence level of the automated operation platform. High-precision measurement enables the automated operation platform to maintain stable performance in complex environments, enhancing the environmental adaptability of the automated operation platform. Accurate data input helps to identify potential problems in advance and take timely measures, ensuring the stability and security of the automated operation platform. Reasonable design of the measurement range not only improves resource utilization, but also reduces hardware costs, reduces unnecessary energy consumption, and improves economic benefits.

[0077] Connect the sensor to the automated operation platform, configure the communication protocol and set the sampling frequency;

[0078] It should be noted that the current, voltage, temperature, humidity, light intensity and grid load sensors are connected to the central automation operation platform by wired or wireless means, and a secure encrypted communication protocol (such as TLS / SSL) is selected to ensure the security and integrity of data transmission. The sampling frequency is set as follows:

[0079] Current sensor: collects data once per second (1Hz) to ensure real-time monitoring of current changes during charging;

[0080] Voltage sensor: collects data once per second (1Hz) to capture voltage fluctuations and ensure charging stability;

[0081] Temperature sensor: collects data once a minute (1 / 60Hz) to monitor battery and ambient temperature to prevent overheating;

[0082] Humidity sensor: collects data every 10 minutes (1 / 600Hz) to evaluate the impact of ambient humidity on the battery;

[0083] Light intensity sensor: collects data every 30 minutes (1 / 1800Hz), records changes in natural light sources, and assists in predicting solar charging efficiency;

[0084] Grid load sensor: collects data every 5 minutes (1 / 300Hz) to monitor power supply and optimize resource allocation.

[0085] Through secure encrypted communication protocols, data transmission from sensors to the automated operation platform is ensured to be fast and secure, preventing data from being tampered with or leaked, and flexibly adapting to different needs: different sampling frequency settings are adjusted according to actual needs to ensure high-frequency monitoring of key parameters, while reducing unnecessary data volume, improving the operating efficiency of the automated operation platform, optimizing resource utilization and reducing latency: reasonable sampling frequency avoids data overload, reduces the processing burden of the automated operation platform, reduces the response time of the automated operation platform, and improves overall operating efficiency.

[0086] Real-time transmission and high-precision data collection provide a solid foundation for subsequent data analysis, enhance the stability and reliability of the automated operation platform, and strengthen decision-making support capabilities: high-quality data input enables machine learning models to more accurately predict future demand and behavior patterns, guide more intelligent charging strategies, and enhance the intelligence level of the automated operation platform, improve operational efficiency and reduce costs: by optimizing resource utilization, it reduces unnecessary energy consumption and equipment redundancy, reduces hardware costs and maintenance costs, and improves economic benefits.

[0087] The automated operation platform triggers data collection at regular intervals to obtain multi-dimensional data on current, voltage, temperature, internal resistance, charging rate, discharge depth, and grid load conditions from sensors.

[0088] It should be noted that the automated operation platform automatically triggers data collection tasks according to preset time intervals (such as every second, every minute or longer), and acquires multi-dimensional data: key parameters including current (1Hz), voltage (1Hz), temperature (1 / 60Hz), internal resistance (1 / 300Hz), charging rate (1 / 60Hz), discharge depth (1 / 3600Hz) and grid load (1 / 300Hz) are collected from different types of sensors.

[0089] Timed-triggered data collection ensures all-round monitoring of the battery and its environmental conditions, provides real-time and continuous data flow, and provides detailed basic information for subsequent analysis and decision-making. Through timed collection, data loss or omission is avoided, the integrity and consistency of the data are guaranteed, and the reliability and stability of the automated operation platform are enhanced. The acquisition of multi-dimensional data enables the automated operation platform to capture more influencing factors, providing rich input for machine learning models and optimization algorithms, and supporting more intelligent and accurate charging strategy formulation.

[0090] S2. Transmit multi-dimensional data to the automated operation platform for cleaning and feature extraction, and train the adaptive learning model.

[0091] Furthermore, the sensor data is transmitted to the automated operation platform in real time using an encrypted communication protocol, and the automated operation platform receives and stores the data packets via a POST request of the HTTP protocol;

[0092] Specifically, encrypted communication protocols such as TLS / SSL are selected to ensure that the security and confidentiality of data are not threatened during the data transmission from sensors to the automated operation platform. The data collected by the sensors are transmitted to the automated operation platform in real time through encrypted channels, ensuring the immediacy and continuity of the data. The automated operation platform uses the POST request of the HTTP protocol to receive data packets and stores them in a distributed database to ensure effective data management and subsequent analysis.

[0093] The encrypted communication protocol prevents data from being stolen or tampered with during transmission, enhancing the security of the automated operation platform. It is particularly suitable for unattended environments. Real-time transmission enables the automated operation platform to immediately obtain the latest battery status and environmental parameters, adjust the charging strategy in a timely manner, and improve the response speed of the automated operation platform. The HTTP POST request method simplifies the data transmission process, ensuring that data packets can reach the automated operation platform quickly and reliably, and are effectively stored for subsequent processing and analysis.

[0094] Encrypted communication ensures the security of data transmission, reduces problems caused by data loss or errors, and enhances the stability and reliability of the automated operation platform. Real-time data transmission provides the latest and most accurate input for machine learning models, supports more intelligent and precise charging strategy formulation, and improves the intelligence level of the automated operation platform. Efficient transmission and storage mechanisms reduce the burden on the automated operation platform, reduce unnecessary hardware costs and maintenance work, and improve operational efficiency. Real-time monitoring and immediate response mechanisms enable the automated operation platform to discover and solve problems in the first place, ensuring service continuity and the quality of user experience.

[0095] Apply low-pass filtering and smoothing, and store the packets in a distributed database;

[0096] It should be noted that a low-pass filter is applied to the data collected by the sensor to remove high-frequency noise and instantaneous fluctuations and retain useful signals. A smoothing algorithm (such as the moving average method) is further used to reduce random disturbances in the data and make the data more stable and reliable. The processed data packets are stored in a distributed database to ensure data security, scalability and efficient access.

[0097] Low-pass filters and smoothing processing effectively remove noise and outliers in the data, improve the quality and reliability of the data, and provide clean basic data for subsequent analysis. High-quality data input enables machine learning models and prediction algorithms to capture trends and patterns more accurately, supporting more scientific and reasonable charging strategy formulation. Distributed databases not only provide high availability and fault tolerance, but also support rapid access and processing of large-scale data, enhancing the flexibility and responsiveness of the automated operation platform.

[0098] Through purification processing, noise interference is reduced, making the data more accurate and stable, thereby improving the reliability of decisions made based on these data. After removing outliers and smoothing, the automated operation platform can better respond to environmental changes and emergencies, enhancing overall robustness and stability. The use of distributed databases not only improves the efficiency of data storage and retrieval, but also reduces the risk of single point failures, reduces maintenance costs, and improves operational efficiency. The processed data can more realistically reflect the battery status and environmental conditions, help identify potential problems in advance and take preventive measures, and ensure the long-term performance and reliability of the automated operation platform.

[0099] The automated operation platform identifies data packets in the distributed database and removes outliers and fills in missing values;

[0100] It should be noted that the automated operation platform extracts data packets transmitted by sensors from the distributed database to ensure that all data are correctly received and stored, applies statistical methods or machine learning algorithms to identify and eliminate abnormal data points that are beyond the normal range, prevents these outliers from affecting subsequent analysis, and uses interpolation, mean filling or prediction models to fill in missing data caused by collection failure or other reasons to ensure data continuity and integrity.

[0101] By eliminating outliers, the impact of noise and erroneous data on the automated operation platform is eliminated, the quality and reliability of the data are improved, and missing values ​​are filled to keep the data sequence complete, avoiding analysis bias caused by data discontinuity, providing a reliable foundation for subsequent processing. High-quality and complete data input enables machine learning models and prediction algorithms to more accurately capture trends and patterns, supporting more scientific and reasonable charging strategy formulation.

[0102] The purified data reduces the sources of errors, making the decisions based on the data more accurate and reliable, and improving the intelligence level of the automated operation platform. After removing outliers and smoothing, the automated operation platform can better respond to environmental changes and emergencies, and enhance overall robustness and stability. The complete data sequence helps to more accurately assess the battery status and environmental conditions, thereby optimizing the charging strategy and reducing unnecessary energy consumption and equipment wear. The processed data can more truly reflect the actual situation, help identify potential problems in advance and take preventive measures, and ensure the long-term performance and reliability of the automated operation platform.

[0103] Extract features from processed data packets and select predictive features to optimize the input of the adaptive learning model;

[0104] It should be noted that the automated operation platform conducts in-depth analysis of the data packets that have been low-pass filtered and smoothed, extracts key features such as current, voltage, temperature, internal resistance, charging rate, etc., and screens out the most representative and predictive features based on domain knowledge and statistical analysis to ensure that the data input into the adaptive learning model is highly relevant and representative, and combines the selected predictive features into the optimal input set to provide high-quality data support for training the adaptive learning model.

[0105] Through feature extraction, the automated operation platform can focus on the key parameters that best reflect battery status and environmental changes, enhancing the explanatory power of the data and the accuracy of the model. Selecting predictive features enables the model to more accurately capture the factors that affect charging demand and improves the accuracy of predictions for future charging demand. High-quality input data helps train more efficient adaptive learning models, reduce the risk of overfitting, and improve the generalization and robustness of the model.

[0106] By optimizing the feature input, the interference of irrelevant or redundant information is reduced, making the model output more accurate and reliable, and guiding the formulation of more scientific and reasonable charging strategies. Accurate feature selection and optimized input significantly improve the performance of the adaptive learning model, enabling the automated operation platform to dynamically adjust the charging strategy according to real-time environmental changes, improving the overall intelligence level. The optimized model can more accurately evaluate the battery status and environmental conditions, thereby optimizing the charging strategy, reducing unnecessary energy consumption and equipment wear, and improving operational efficiency. High-quality feature input and optimized models help to identify potential problems in advance and take preventive measures, ensuring the long-term performance and reliability of the automated operation platform.

[0107] Initialize the long short-term memory model parameters, set the initial learning rate and batch size, use historical data to train the adaptive learning model, and dynamically adjust the hyperparameters;

[0108] It should be noted that setting the initial weights and biases ensures that the model has a good starting point, avoids the problem of gradient disappearance or explosion, and selects an appropriate learning rate (such as 0.001), which determines the speed of parameter update. A higher learning rate can speed up convergence but may cause instability, while a lower learning rate ensures a more accurate gradient estimate but may prolong the training time. Setting an appropriate batch size (such as 64) balances memory usage and the accuracy of gradient estimation. Larger batches provide more stable gradient estimates but require more memory; smaller batches save memory, but the gradient estimate fluctuates more.

[0109] Reasonable initialization parameters help speed up the training process, avoid falling into local optimal solutions, and improve the efficiency of model training. Appropriate learning rate and batch size settings can ensure the accuracy of the model while ensuring the training speed, thereby improving the overall training effect.

[0110] By carefully setting the initial learning rate, the initial shock of training is reduced, the model enters a stable state faster, and the training time is significantly shortened. Reasonable learning rate and batch size configuration enable the model to better capture data features during training, thereby enhancing its generalization ability and robustness.

[0111] Use processed multi-dimensional historical data for training and dynamically adjust hyperparameters to optimize model performance:

[0112] The preprocessed and feature-extracted data packets are input into the long short-term memory model for training, so that it can learn the changing patterns of battery status and environmental conditions. During the training process, hyperparameters such as learning rate and batch size are dynamically adjusted according to the performance on the validation set to ensure that the model achieves optimal performance.

[0113] Through historical data training, the model can learn the changing patterns of battery status and environmental conditions, enhancing the ability to predict future charging needs. Dynamic adjustment of hyperparameters enables the model to continuously improve its performance and adapt to the needs under different working conditions, thereby improving the model's robustness and generalization capabilities.

[0114] The optimized long short-term memory model can more accurately predict future charging demand, reduce prediction errors, and guide the formulation of more scientific and reasonable charging strategies. The precise prediction capability enables the automated operation platform to dynamically adjust the charging strategy according to real-time environmental changes, significantly improving the intelligence level and response speed of the automated operation platform. The optimized model can more accurately evaluate the battery status and environmental conditions, thereby optimizing the charging strategy, reducing unnecessary energy consumption and equipment wear, and improving operational efficiency. The high-quality prediction and optimization mechanism helps to identify potential problems in advance and take preventive measures, ensuring the long-term performance and reliability of the automated operation platform.

[0115] S3, based on the trained adaptive learning model, automatically adjust the prediction parameters according to the real-time environmental changes to generate the charging demand prediction results. Charging strategy;

[0116] Furthermore, historical data and real-time sensor data are extracted from the distributed database and normalized and standardized;

[0117] Specifically, normalization and standardization processing enables data of different dimensions to be compared and analyzed under a unified standard. The processed data is more in line with the requirements of the machine learning model, reduces the impact of outliers and noise, and improves the quality of model input.

[0118] Consistent data input helps improve the stability of the model and prediction accuracy, reduces errors caused by data differences, and standardized data facilitates subsequent feature extraction and model training, reduces processing complexity, and improves overall efficiency.

[0119] Combining historical feature changes and current environmental impacts, the i-th predicted feature value f of the historical data and real-time sensor data that have been normalized and standardized over time is i (t), the jth environmental factor value e that affects charging demand in the real-time environment j and the weight w corresponding to the i-th prediction feature i The expression is introduced to calculate the dynamically adjusted hyperparameter α′, which is:

[0120]

[0121] Among them, λ is the attenuation factor, α is the initial hyperparameter, n is the total number of features, γ is the weight of the environmental factor, m is the total number of environmental factors, j is the accurate value in time change, i is the accurate value in charging demand, and t is the time variable;

[0122] It should be noted that the attenuation factor λ is used to control the influence of the historical hyperparameter α and balance the ratio of new and old information.

[0123] The initial hyperparameter α represents the predicted feature value that changes over time, reflecting the change in battery state.

[0124] The total number of features n is the predicted feature value that changes over time, reflecting the change in battery status.

[0125] The environmental factor weight γ is the importance or influence of the environmental factor on the charging demand prediction. Each environmental factor (such as temperature, humidity, light intensity, etc.) has a corresponding weight, which reflects the relative importance of the factor in the prediction model. By assigning different weights to different environmental factors, the actual impact of each environmental factor on the charging demand can be more accurately reflected. For example, temperature changes may have a greater impact on battery performance, so its weight is higher; while the weights of some minor factors are lower. Reasonable weight settings help improve the accuracy and reliability of the prediction model and ensure that the model can better adapt to real-time environmental changes.

[0126] The total number of environmental factors m represents the number of environmental factors involved in the calculation. It indicates how many different environmental variables the automated operation platform considers to adjust the hyperparameter α′. The total number of environmental factors m determines how many environmental variables the model considers. More environmental factors can provide more comprehensive information and make the model more intelligent and flexible. Although increasing the number of environmental factors can improve the accuracy of the model, too many factors will also increase the complexity of the calculation. Therefore, choosing an appropriate m value can ensure the prediction accuracy while maintaining the calculation efficiency.

[0127] The formula comprehensively considers the impact of historical data and real-time environment, so that the hyperparameter α′ can reflect the latest situation in real time, improves the adaptability and prediction accuracy of the model, and the attenuation factor λ ensures that historical information will not be completely ignored. At the same time, new environmental changes can also be taken into consideration in a timely manner, ensuring the stability of the model. The dynamically adjusted hyperparameter α′ can more accurately reflect the actual situation, reduce prediction errors, and improve the intelligence level of the automated operation platform. The model can be flexibly adjusted according to real-time environmental changes, enhancing the robustness and adaptability of the automated operation platform. Accurate hyperparameter adjustment enables the automated operation platform to maintain optimal performance in different environments, reducing unnecessary waste of resources and improving overall operational efficiency.

[0128] Using the initial hyperparameter α, the future charging demand is predicted through the long short-term memory model, and the dynamically adjusted hyperparameter α′ is optimized;

[0129] It should be noted that the dynamically adjusted hyperparameter α′ is used as the initial parameter of the optimization algorithm to ensure that the optimization process starts from a better starting point. The initial learning rate and batch size are dynamically adjusted according to the hyperparameter α′, so that the optimization algorithm can converge to the optimal solution faster. The obtained dynamically adjusted hyperparameter α′ is applied to the initial parameter setting of the optimization algorithm. The dynamically adjusted hyperparameter α′ is used as the initial parameter of the optimization algorithm to ensure the efficiency of the optimization process. According to the results of the dynamic adjustment, the population parameters are gradually optimized, so that the optimization algorithm can converge to the optimal solution faster.

[0130] The dynamically adjusted initial parameter α and population parameter speed up the convergence of the optimization algorithm and improve the optimization efficiency. The optimized charging strategy can more accurately evaluate the battery status and environmental conditions, reduce unnecessary energy consumption and equipment wear, and ensure the long-term performance and reliability of the automated operation platform. Through dynamic optimization, the automated operation platform can adjust the charging strategy according to real-time environmental changes, further improving the intelligence level and response speed of the automated operation platform.

[0131] It should be noted that according to the requirements of the optimization algorithm, the individual parameters in the population are updated to ensure that each individual can perform effective search under the new conditions. According to the dynamically adjusted hyperparameter α′, the initial learning rate and batch size of the optimization algorithm are adjusted to make the optimization process more flexible and efficient.

[0132] The dynamically adjusted hyperparameter α′ is used as the initial parameter of the optimization algorithm, ensuring that the optimization process starts from a better starting point. According to the results of the dynamic adjustment, the population parameters are gradually optimized so that the optimization algorithm can converge to the optimal solution faster. The dynamically adjusted initial learning rate and batch size accelerate the convergence speed of the optimization algorithm and improve the optimization efficiency. Through dynamic optimization, the automated operation platform can adjust the charging strategy according to real-time environmental changes, further improving the intelligence level and response speed of the automated operation platform.

[0133] The dynamically adjusted initial parameters and population parameters speed up the convergence of the optimization algorithm, reduce the number of unnecessary iterations, and improve the overall optimization efficiency. The optimized charging strategy can more accurately evaluate the battery status and environmental conditions, reduce unnecessary energy consumption and equipment wear, and ensure the long-term performance and reliability of the automated operation platform. The automated operation platform can flexibly adjust the charging strategy according to real-time environmental changes, enhance the robustness and adaptability of the automated operation platform, and ensure that it can maintain optimal performance under different working conditions. The optimized charging strategy provides a reliable basis for formulating scientific and reasonable charging plans, and improves the intelligence level and management efficiency of the automated operation platform.

[0134] Update the population parameters in the optimization algorithm and dynamically adjust the initial learning rate and batch size based on the hyperparameter α′.

[0135] It should be noted that the initial learning rate of the optimization algorithm is adjusted according to the dynamically adjusted hyperparameter α′ to ensure that the optimization process starts from a better starting point. According to the requirements of the optimization algorithm, the batch size is dynamically adjusted to ensure that each iteration can make full use of the latest data information and improve the optimization efficiency.

[0136] The dynamically adjusted initial learning rate and batch size speed up the convergence of the optimization algorithm, reduce the number of unnecessary iterations, and improve the overall optimization efficiency. The automated operation platform can flexibly adjust the charging strategy according to real-time environmental changes, enhance the robustness and adaptability of the automated operation platform, and ensure optimal performance under different working conditions. The optimized charging strategy can more accurately evaluate the battery status and environmental conditions, reduce unnecessary energy consumption and equipment wear, and ensure the long-term performance and reliability of the automated operation platform.

[0137] The optimized charging strategy provides a reliable basis for formulating scientific and reasonable charging plans, improves the intelligence level and management efficiency of the automated operation platform. Through dynamic optimization, the automated operation platform can adjust the charging strategy according to real-time environmental changes, further improving the intelligence level and response speed of the automated operation platform. The optimized charging strategy reduces unnecessary waste of resources, improves overall operational efficiency and resource utilization efficiency, and reduces operating costs.

[0138] S4. Based on the charging demand prediction results, an optimization algorithm is applied to determine the charging strategy.

[0139] Furthermore, the dynamically adjusted hyperparameter α′ is used to generate charging demand prediction results through the long short-term memory model to obtain the charging demand intensity Q and power level P at the time point;

[0140] Specifically, the demand characteristic value g that changes over time is extracted from multi-dimensional data i (t), and assign weight d i The total demand characteristic value is formed as follows:

[0141]

[0142] Where C is the total number of demand characteristics;

[0143] Extract the health feature value k that affects charging demand from the real-time environment j (t), and assign weight b j The total health characteristic value is formed as follows:

[0144]

[0145] Where R is the total number of health characteristics;

[0146] It should be noted that by using the dynamically adjusted hyperparameter α′, the long short-term memory model is used to predict future charging demand, and the charging demand intensity Q and power level P at a certain point in time are obtained. The long short-term memory model can capture the dependencies in long time series data and accurately predict future charging demand. The dynamically adjusted hyperparameter α′ ensures that the model is always in the best state, thereby improving the accuracy and stability of the prediction.

[0147] Accurate prediction results provide a reliable basis for formulating scientific and reasonable charging strategies and improve the intelligence level of the automated operation platform. Real-time prediction enables the automated operation platform to respond in the first time, improving the overall operating efficiency. The dynamically adjusted hyperparameter α′ reduces the prediction error, improves the prediction accuracy, and enhances the reliability of the automated operation platform.

[0148] Extracting demand characteristic values ​​g that change over time from multi-dimensional data i (t), such as current, voltage, temperature, etc., and assign corresponding weights di according to the importance of each demand feature to form the total demand feature value D. By introducing multiple demand feature values ​​and assigning different weights, the automated operation platform can more comprehensively evaluate the changes in charging demand. High-quality feature input helps to train more efficient adaptive learning models, reduce the risk of overfitting, and improve the generalization and robustness of the model. Comprehensive consideration of multiple demand characteristics makes the prediction more accurate and reduces the deviation that may be caused by a single feature. The weights are flexibly adjusted according to the importance of different features, which enhances the adaptability and flexibility of the model. Accurate demand assessment helps to optimize charging strategies, reduce unnecessary energy consumption and equipment wear, and improve operational efficiency.

[0149] Extract the health feature value k that affects charging demand from the real-time environment j (t), such as temperature, humidity, light intensity, etc. according to each health characteristic k j The importance of (t) is assigned the corresponding weight b j , forming the total health characteristic value H, by introducing the health characteristic value k j (t), the automated operation platform can monitor the health status of the battery and its environment in real time, identify potential problems in advance, and jBased on the evaluation results of (t), the automated operation platform can dynamically adjust the charging strategy to ensure optimal performance under different working conditions, monitor and evaluate the battery health status in real time, reduce the risk of failure due to environmental factors, ensure the long-term performance and reliability of the automated operation platform, identify potential problems in advance and take preventive measures, and improve the safety and stability of the automated operation platform. Accurate health assessment helps to discover and solve problems in a timely manner, reduce maintenance costs, and improve maintenance efficiency.

[0150] Furthermore, the charging demand intensity Q and power level P are introduced into the expression to calculate the optimal charging strategy S * , the expression is:

[0151]

[0152] Where η is the charging demand weight factor, u is the battery health weight factor, ω is the hyperparameter weight factor, ρ is the demand intensity weight factor, -u(P-P0) is the direction of change from power level P to initial power level P0, ξ is the battery health weight factor, e is the base of the natural logarithm, e -u(P-P0) Indicates the change direction of the current power level P relative to the initial power level P0.

[0153] It should be noted that the charging demand intensity Q is considered to ensure that the charging strategy can meet the actual demand. By introducing the current power level P and combining it with the initial power level P o , evaluate the direction and degree of power change, use the Sigmoid function to smooth the impact of power change, and ensure that the impact of power change on the charging strategy is gradual and controllable.

[0154] By introducing the charging demand intensity Q, the automated operation platform can more comprehensively evaluate changes in charging demand and ensure that the charging strategy is both satisfying and efficient. By introducing the Sigmoid function, the automated operation platform can smoothly handle the impact of power changes and avoid the impact of power mutations on the battery and the automated operation platform.

[0155] By comprehensively considering multiple factors, the charging strategy is made more scientific and reasonable, the charging efficiency and resource utilization are improved, and the impact of power mutations on the battery and the automated operation platform is reduced by smoothing power changes, thereby enhancing the stability and security of the automated operation platform. By introducing multiple factors for comprehensive evaluation, the automated operation platform can flexibly adjust the charging strategy according to real-time environmental changes, thereby ensuring the long-term performance and reliability of the automated operation platform.

[0156] Furthermore, the differential evolution algorithm is used to solve the charging strategy S * , initialize the population parameters and dynamically adjust them, and iteratively update the population individuals until they converge to the optimal solution.

[0157] It also includes the use of differential evolution algorithm to solve the charging strategy, initializing population parameters and dynamically adjusting them, and iteratively updating population individuals until they converge to the optimal solution. The specific steps are:

[0158] The automated operation platform generates multiple candidate charging strategies based on the charging demand forecast results;

[0159] The automated operation platform defines comprehensive evaluation criteria, analyzes each candidate charging strategy, evaluates whether it meets charging requirements, grid load balancing, and battery health impact, and draws evaluation results;

[0160] The automated operation platform uses the differential evolution algorithm to iteratively optimize the population, generates new individuals through selection, mutation, and crossover operations, and replaces the original individuals, gradually approaching the optimal solution;

[0161] Improve optimization efficiency and robustness according to the adaptive differential evolution algorithm parameters;

[0162] When the maximum number of iterations is reached, the adaptive differential evolution algorithm stops running and outputs the final optimal charging strategy.

[0163] Specifically, by generating multiple candidate strategies, it is ensured that the automated operation platform has appropriate solutions under different working conditions. Different candidate strategies can flexibly respond to real-time environmental changes and demand fluctuations, thus enhancing the adaptability of the automated operation platform.

[0164] The diverse candidate strategies enable the automated operation platform to respond quickly in different situations, improving overall flexibility. By having multiple options, it reduces the risks that may be brought about by a single strategy and enhances the reliability of the automated operation platform.

[0165] The automated operation platform defines comprehensive evaluation criteria, covering aspects such as charging demand, grid load balancing, and battery health impact. Based on the defined criteria, each candidate charging strategy is comprehensively analyzed, its performance is evaluated, and the evaluation results are obtained.

[0166] Comprehensive evaluation criteria ensure that candidate strategies are fully considered in multiple key aspects (such as charging demand, grid load balancing, and battery health). Through comprehensive performance evaluation, the automated operation platform can select the charging strategy that best meets actual needs, improving the scientificity and rationality of decision-making;

[0167] Comprehensive performance evaluation provides a reliable basis for selecting the best charging strategy and improves the intelligence level of the automated operation platform. By comprehensively considering multiple factors, the automated operation platform can more accurately evaluate battery status and environmental conditions, ensuring the long-term performance and reliability of the automated operation platform.

[0168] Set parameters such as the initial population size, mutation factor, and crossover probability to ensure that the optimization process starts from a reasonable starting point. Generate new individuals through selection, mutation, and crossover operations, and replace the original individuals to gradually approach the optimal solution.

[0169] The evolutionary algorithm is a global optimization algorithm that can effectively handle complex multi-dimensional optimization problems and is suitable for solving the best charging strategy. The process of iteratively updating population individuals can quickly find the global optimal solution, reducing unnecessary waste of computing resources. Through selection, mutation and crossover operations, it ensures a wide coverage of the search space and avoids falling into the local optimal solution.

[0170] The differential evolution algorithm can quickly find the global optimal solution, reduce optimization time, and improve overall efficiency. By dynamically adjusting population parameters, the automated operation platform can flexibly adjust according to real-time environmental changes, thereby improving the intelligence level of the automated operation platform. The optimized charging strategy can more accurately evaluate battery status and environmental conditions, reduce unnecessary energy consumption and equipment wear, and ensure the long-term performance and reliability of the automated operation platform.

[0171] According to the feedback from the optimization process, dynamically adjust population parameters such as population size, mutation factor and crossover probability to adapt to different working conditions;

[0172] By dynamically adjusting population parameters, the algorithm can be flexibly adjusted under different working conditions, which improves the robustness and adaptability of optimization. Dynamic adjustment of parameters enables the algorithm to maintain optimal performance at different stages and improves the quality of optimization results.

[0173] Dynamic adjustment of population parameters enables the automated operation platform to flexibly adjust according to real-time environmental changes, enhancing the robustness and adaptability of the automated operation platform. By dynamically adjusting parameters, the optimization process can maintain optimal performance at different stages, improving the quality and reliability of the optimization results.

[0174] When the maximum number of iterations is reached or the convergence condition is met, the optimization process is stopped and the final optimal charging strategy is output for execution by the automated operation platform;

[0175] The termination condition is determined to ensure that the optimization process is completed within a reasonable time, avoiding unnecessary waste of computing resources, and outputting the final optimal charging strategy to provide clear guidance for practical applications;

[0176] Automatic termination conditions ensure that the optimization process is completed within a reasonable time, improving overall efficiency. The output of the final optimal charging strategy provides clear guidance for practical applications and improves the practicality and operability of the automated operation platform.

[0177] S5. Convert the charging strategy into a control instruction and send it to the controller in the automated operation platform. After receiving the instruction, the controller adjusts the working mode of the charging device.

[0178] Furthermore, according to the final optimal charging strategy, the control instructions are organized into JSON format;

[0179] Specifically, the control instructions are sent from the automated operation platform to the controller via the POST request method of the HTTP protocol using the TLS / SSL encryption protocol, and the controller receives and parses the instructions and extracts parameters;

[0180] The controller adjusts the working mode of the charging device according to the extracted parameters;

[0181] The controller executes the adjusted working mode and monitors the charging effect through sensors;

[0182] The controller continuously monitors the charging process through sensors. When an abnormal situation is detected, the controller information is fed back to the automated operation platform, and the automated operation platform records the charging data.

[0183] Furthermore, the controller uses sensors to continuously collect the operating parameters of the battery (such as voltage, current, temperature, etc.), and transmits the collected data to the automated operation platform in real time to ensure that the automated operation platform can obtain the latest battery status information;

[0184] Through continuous monitoring, the automated operation platform can timely grasp the operating status of the battery, ensuring that the health status of the battery can be understood at any time. Real-time data transmission enables the automated operation platform to receive the latest information at the first time, improving the response speed and decision-making efficiency of the automated operation platform.

[0185] Real-time monitoring and data transmission ensure the transparency of battery status, reduce the risk of potential failures, and enhance the security of the automated operation platform. Continuous data collection and transmission provide a reliable foundation for subsequent analysis and improve the overall reliability of the automated operation platform.

[0186] The automated operation platform compares and analyzes historical data and real-time data to identify possible abnormal patterns. Through data analysis, the automated operation platform can identify abnormal situations that do not conform to normal patterns and discover potential problems in advance.

[0187] By comparing and analyzing historical data and real-time data, the automated operation platform can detect potential problems in advance and avoid the expansion of failures. Based on the results of data analysis, the automated operation platform can make more scientific and reasonable decisions, thereby improving management efficiency.

[0188] Early identification of abnormal patterns enables the automated operation platform to take preventive measures before problems occur, reducing the possibility of failures. Discovering potential problems in advance can reduce the need for emergency repairs and reduce maintenance costs and time.

[0189] Once a potential failure is identified, the automated operation platform immediately triggers an early warning mechanism and notifies relevant personnel through multiple channels (such as email, SMS, and application push) to ensure that they can take prompt action;

[0190] The triggering warning mechanism enables relevant personnel to be informed of potential risks at the first time and take timely response measures. Through various notification methods, the timeliness and accuracy of information transmission are ensured, thus improving the efficiency of emergency response.

[0191] Timely response to potential risks reduces the possibility of fault spread, enhances the emergency handling capabilities of the automated operation platform, and ensures the safety of personnel by notifying relevant personnel in a timely manner to ensure that they take appropriate measures in the shortest possible time.

[0192] If no potential fault is detected, the automated operation platform will continue to maintain the current charging status to ensure the continuity of the charging process. Even if no fault is detected, the automated operation platform will continue to monitor the operating status of the battery to ensure that any changes can be captured in time;

[0193] Maintaining the current charging status ensures the stable operation of the automated operation platform and avoids unnecessary interruptions or adjustments. Even under normal circumstances, the automated operation platform constantly monitors the battery status to ensure that any potential problems can be discovered in a timely manner;

[0194] Maintaining the current charging status ensures the continuous operation of the automated operation platform, reduces unnecessary interference and adjustments, and continuous monitoring enables the automated operation platform to detect problems at any time, enhancing the overall reliability of the automated operation platform.

[0195] When the early warning mechanism is triggered, the automated operation platform notifies relevant personnel through various means such as email, SMS and application push, ensuring that different types of users can receive early warning information in a timely manner, thus improving the coverage and efficiency of information transmission;

[0196] Through a variety of notification methods, it is ensured that the early warning information can be quickly delivered to the relevant personnel, which improves the response speed. The multi-channel notification method ensures that different types of users can receive the information, which improves the comprehensiveness and accuracy of information transmission;

[0197] The rapid transmission of early warning information enables relevant personnel to take action in the shortest possible time, reducing the possibility of fault expansion. Through a variety of notification methods, the timeliness and accuracy of information transmission are ensured, improving communication efficiency and emergency response capabilities.

[0198] This embodiment also provides a computer device, which is suitable for the case of an unattended hydrogen ion battery charging station method, including: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions to implement the unattended hydrogen ion battery charging station method proposed in the above embodiment.

[0199] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0200] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for realizing an unattended hydrogen ion battery charging station as proposed in the above embodiment is implemented; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0201] In summary, the present invention deploys high-precision sensors to monitor battery status and environmental conditions in real time, collects multi-dimensional data, ensures the quality and reliability of input data, enhances the adaptability of the automated operation platform to complex environments, and realizes instant response to changes to optimize charging strategies. The data is securely transmitted to the automated operation platform using an encrypted communication protocol, cleaned and feature extracted on the platform, and the adaptive learning model is trained using a long short-term memory model, thereby improving the accuracy and security of data analysis. Based on the trained model, the automated operation platform automatically adjusts the prediction parameters according to real-time environmental changes, generates accurate charging demand prediction results, reduces prediction errors and enhances the flexibility and reliability of the automated operation platform. Finally, the optimal charging strategy is determined by the differential evolution algorithm, and multiple factors such as charging demand intensity, power level and battery health are comprehensively considered to quickly find the global optimal solution, improve charging efficiency and resource utilization, and reduce operating costs.

[0202] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An unattended hydrogen ion battery charging station, characterized in that: include, Monitor battery status parameters and environmental conditions through sensors to collect multi-dimensional data; Transmit multi-dimensional data to the automated operation platform for cleaning and feature extraction, and train the adaptive learning model; Based on the trained adaptive learning model, the prediction parameters are automatically adjusted according to real-time environmental changes to generate charging demand prediction results; Based on the charging demand forecast results, the optimization algorithm is applied to determine the charging strategy; The charging strategy is converted into control instructions and sent to the controller in the automated operation platform. After receiving the instructions, the controller adjusts the working mode of the charging equipment; Detect the battery's operating status in real time and trigger an early warning mechanism when a potential fault is detected.

2. The unattended hydrogen ion battery charging station as claimed in claim 1, characterized in that: The method monitors battery status parameters and environmental conditions through sensors to collect multi-dimensional data, and the specific steps are: Select and install current, voltage, temperature, humidity, light intensity, and grid load sensors; Configure the sensor's operating mode, set the measurement range and accuracy; Connect the sensor to the automated operation platform, configure the communication protocol and set the sampling frequency; The automated operation platform triggers data collection at regular intervals to obtain multi-dimensional data on current, voltage, temperature, internal resistance, charging rate, discharge depth, and grid load conditions from sensors.

3. The unattended hydrogen ion battery charging station as claimed in claim 2, characterized in that: The multi-dimensional data is transmitted to the automated operation platform for cleaning and feature extraction, and the adaptive learning model is trained. The specific steps are: The sensor data is transmitted to the automated operation platform in real time using an encrypted communication protocol. The automated operation platform receives and stores data packets through HTTP protocol POST requests. Apply low-pass filtering and smoothing, and store the packets in a distributed database; The automated operation platform identifies data packets in the distributed database and removes outliers and fills in missing values; Extract features from processed data packets and select predictive features to optimize the input of the adaptive learning model; Initialize the long short-term memory model parameters, set the initial learning rate and batch size, use historical data to train the adaptive learning model, and dynamically adjust the hyperparameters.

4. The unattended hydrogen ion battery charging station as claimed in claim 3, characterized in that: The adaptive learning model based on training automatically adjusts the prediction parameters according to the real-time environmental changes. The specific steps are: Extract historical data and real-time sensor data from distributed databases and perform normalization and standardization; Combining historical feature changes and current environmental impacts, the i-th predicted feature value f of the historical data and real-time sensor data that have been normalized and standardized over time is i (t), the jth environmental factor value e that affects charging demand in the real-time environment j and the weight w corresponding to the i-th prediction feature i , introduce the expression to calculate the dynamically adjusted hyperparameter α'; Using the initial hyperparameter α, the future charging demand is predicted through the long short-term memory model, and the dynamically adjusted hyperparameter α' is optimized; Apply the obtained dynamically adjusted hyperparameter α' to the initial parameter setting of the optimization algorithm; Update the population parameters in the optimization algorithm, and dynamically adjust the initial learning rate and batch size based on the dynamic adjustment hyperparameter α'.

5. The unattended hydrogen ion battery charging station as claimed in claim 4, characterized in that: The charging strategy is determined by applying an optimization algorithm according to the charging demand prediction result. The specific steps are: Using the dynamically adjusted hyperparameter α', the charging demand forecast results are generated through the long short-term memory model to obtain the charging demand intensity Q and power level P at the time point; Combined with extracting demand characteristic values ​​g that change over time from multi-dimensional data i (t), and assign weight d i Form the total demand characteristic value D; Extract the health feature value k that affects charging demand from the real-time environment j (t), and assign weight b j Form the total health characteristic value H; The charging demand intensity Q and power level P are introduced into the expression to calculate the optimal charging strategy S * ; Differential evolution algorithm is used to solve the charging strategy S * , initialize the population parameters and dynamically adjust them, and iteratively update the population individuals until they converge to the optimal solution.

6. The unattended hydrogen ion battery charging station as claimed in claim 5, characterized in that: The differential evolution algorithm is used to solve the charging strategy S * , initialize the population parameters and dynamically adjust them, iteratively update the population individuals until they converge to the optimal solution. The specific steps are: The automated operation platform generates multiple candidate charging strategies based on the charging demand forecast results; The automated operation platform defines comprehensive evaluation criteria, analyzes each candidate charging strategy, evaluates whether it meets charging requirements, grid load balancing, and battery health impact, and draws evaluation results; The automated operation platform uses the differential evolution algorithm to iteratively optimize the population, generates new individuals through selection, mutation, and crossover operations, and replaces the original individuals, gradually approaching the optimal solution; Improve optimization efficiency and robustness according to the adaptive differential evolution algorithm parameters; When the maximum number of iterations is reached, the adaptive differential evolution algorithm stops running and outputs the final optimal charging strategy.

7. The unattended hydrogen ion battery charging station as claimed in claim 6, characterized in that: The charging strategy is converted into a control instruction and sent to the controller in the automated operation platform. After receiving the instruction, the controller adjusts the working mode of the charging device. The specific steps are: According to the final optimal charging strategy, it is organized into control instructions in JSON format; Use TLS / SSL encryption protocol to send control instructions from the automated operation platform to the controller through the HTTP protocol POST request method. The controller receives and parses the instructions and extracts parameters. The controller adjusts the working mode of the charging device according to the extracted parameters; The controller executes the adjusted working mode and monitors the charging effect through sensors; The controller continuously monitors the charging process through sensors. When an abnormal situation is detected, the controller information is fed back to the automated operation platform, and the automated operation platform records the charging data.

8. The unattended hydrogen ion battery charging station as claimed in claim 7, characterized in that: The real-time detection of the battery's operating status triggers an early warning mechanism when a potential fault is detected. The specific steps are: The controller continuously monitors the operating parameters of the battery through sensors and transmits the data to the automated operation platform; The automated operations platform conducts comparative analysis based on historical data and real-time data to identify abnormal patterns; If a potential fault is displayed, the early warning mechanism is triggered; If no potential fault is detected, the current charging state is maintained and the battery operating status continues to be monitored; When the early warning mechanism is triggered, the automated operation platform pushes it to relevant personnel via email, text message, and application.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the unattended hydrogen ion battery charging station according to any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the unattended hydrogen ion battery charging station according to any one of claims 1 to 8 are implemented.

Images