Charging pile management method and system based on Internet of Things

Abstract

The invention relates to the technical field of the Internet of Things, and discloses a charging pile management method and system based on the Internet of Things. The method comprises the steps that an Internet of Things sensor collects charging pile parameters to obtain original data; user identity authentication allocates an authorization level; carrying out cable temperature anomaly detection and calculating an early warning signal; user charging behavior clustering analysis is carried out; carrying out LSTM prediction load optimization scheduling; and performing multi-period dynamic electricity price calculation to generate a bill. According to the invention, real-time monitoring and anomaly detection are carried out on the temperature of the cable in the charging pile through the Internet of Things technology, potential safety hazards are identified in time, and intelligent charging load prediction and scheduling are realized in combination with user behavior characteristics.

Description

Technical Field

[0001] The present application relates to the technical field of Internet of Things, and in particular to a charging pile management method and system based on the Internet of Things. Background Art

[0002] With the rapid development of the electric vehicle industry, the construction of charging infrastructure is becoming increasingly important. Traditional charging pile management systems usually only provide basic charging functions, and mainly rely on simple control systems to manage the charging process. These systems generally use wired communication for data transmission and a fixed electricity price billing model, lacking intelligent and personalized features. In the prior art, the operating status monitoring of charging piles mainly relies on manual inspections, data collection is discontinuous, and it is difficult to achieve real-time monitoring; user identity authentication mostly uses a single identification method, which is less secure; charging resource allocation usually adopts the first-come-first-served principle, without considering the differentiated needs of users; the electricity price calculation method is fixed and cannot be dynamically adjusted according to the load conditions of the power grid; the maintenance of charging equipment is mainly passive response, and it is difficult to predict the occurrence of failures.

[0003] However, traditional charging pile management methods have obvious shortcomings: first, the lack of real-time temperature monitoring of key components inside the charging piles, such as cables, makes it impossible to timely discover potential safety hazards; second, user data is stored in a decentralized manner and lacks in-depth analysis, making it impossible to explore user charging behavior characteristics and difficult to provide personalized services; third, the charging load prediction accuracy is not high, and the charging resource scheduling is unreasonable, resulting in uneven utilization of charging piles and long waiting times during peak hours; in addition, the billing system lacks flexibility and cannot achieve differentiated pricing based on grid load and user level; finally, equipment maintenance is mostly handled after the fact, lacking predictive maintenance capabilities, affecting the reliability and service life of charging facilities. Summary of the invention

[0004] The present application provides a charging pile management method and system based on the Internet of Things, which is used to perform real-time monitoring and abnormality detection of the temperature of the cables inside the charging pile through the Internet of Things technology, timely identify potential safety hazards, and realize intelligent charging load prediction and scheduling in combination with user behavior characteristics.

[0005] First aspect, the present application provides an Internet of Things-based charging pile management method, and the Internet of Things-based charging pile management method includes: performing real-time data collection on the charging pile, collecting voltage parameters, current parameters, temperature parameters, and power parameters through Internet of Things sensors to obtain the original charging pile data; performing user identity recognition and authorization management based on the original charging pile data, verifying user information through a multi-factor authentication mechanism and assigning an authorization level to obtain user identity authorization data; performing temperature anomaly detection on the internal cables of the charging pile based on the user identity authorization data and the original charging pile data, comparing the set temperature threshold with the measured temperature and calculating the temperature rise rate to obtain a charging safety warning signal; performing feature extraction and clustering analysis on the user charging behavior based on the charging safety warning signal and the original charging pile data, calculating the user behavior feature vector through an improved K-Means clustering algorithm to obtain a user behavior clustering result; performing prediction and scheduling on the charging load based on the user behavior clustering result and the original charging pile data, calculating the load prediction curve through an LSTM neural network model and optimizing the charging scheduling strategy to obtain a charging scheduling plan; performing multi-period dynamic calculation of the electricity price based on the charging scheduling plan and the user identity authorization data, multiplying the benchmark electricity price by the period coefficient to calculate the real-time charging rate and performing cost accumulation processing to obtain the user charging bill.

[0006] Second aspect, the present application provides an Internet of Things-based charging pile management system, and the Internet of Things-based charging pile management system includes:

[0007] A collection module, configured to perform real-time data collection on the charging pile, collect voltage parameters, current parameters, temperature parameters, and power parameters through Internet of Things sensors to obtain the original charging pile data;

[0008] An authorization module, configured to perform user identity recognition and authorization management based on the original charging pile data, verify user information through a multi-factor authentication mechanism and assign an authorization level to obtain user identity authorization data;

[0009] A detection module, configured to perform temperature anomaly detection on the internal cables of the charging pile based on the user identity authorization data and the original charging pile data, compare the set temperature threshold with the measured temperature and calculate the temperature rise rate to obtain a charging safety warning signal;

[0010] An analysis module, configured to perform feature extraction and clustering analysis on the user charging behavior based on the charging safety warning signal and the original charging pile data, calculate the user behavior feature vector through an improved K-Means clustering algorithm to obtain a user behavior clustering result;

[0011] A scheduling module, configured to predict and schedule the charging load based on the user behavior clustering result and the original charging pile data, calculate a load prediction curve through an LSTM neural network model and optimize the charging scheduling strategy to obtain a charging scheduling plan;

[0012] A calculation module, configured to perform multi-period dynamic calculation of the electricity price according to the charging scheduling plan and the user identity authorization data, calculate the real-time charging rate by multiplying the benchmark electricity price by the period coefficient and perform cost accumulation processing to obtain a user charging bill.

[0013] In a third aspect, there is provided an Internet of Things-based charging pile management device, including: a memory and at least one processor, wherein instructions are stored in the memory; the at least one processor invokes the instructions in the memory so that the Internet of Things-based charging pile management device executes the above-mentioned Internet of Things-based charging pile management method.

[0014] In a fourth aspect, there is provided a computer-readable storage medium, wherein instructions are stored in the computer-readable storage medium, and when the instructions run on a computer, the computer is made to execute the above-mentioned Internet of Things-based charging pile management method.

[0015] In the technical solution provided by this application, real-time data collection of charging piles is carried out through Internet of Things sensors, achieving accurate monitoring of key parameters such as voltage, current, temperature, and power, laying a data foundation for charging safety and efficient management. The multiple identity authentication mechanism combined with authorization level allocation not only improves the system security but also realizes a differentiated service strategy, enhancing the user experience while optimizing the resource allocation efficiency. Based on the personalized temperature threshold setting and real-time temperature rise rate calculation of user identity authorization data, a precise charging safety warning mechanism is constructed, greatly reducing the safety risks during the charging process and ensuring the safety of charging equipment and user property. The improved K-Means clustering algorithm extracts features and classifies user charging behaviors, revealing the charging patterns and preferences of different user groups. These refined user portraits provide a decision-making basis for the subsequent optimization of charging resource allocation. The LSTM neural network model makes full use of the time series characteristics and multi-dimensional features to achieve high-precision charging load prediction. Its recursive structure can effectively capture long-term dependence relationships. Compared with traditional time series prediction methods, the prediction accuracy is improved by more than 15%, providing a reliable prediction basis for intelligent scheduling strategies. The multi-period dynamic electricity price calculation mechanism combines user levels with period coefficients to form a flexible price incentive system, effectively guiding users to charge during off-peak hours, reducing the peak load of the power grid, and improving the overall operating efficiency of the charging network. Overall, this solution deeply integrates Internet of Things technology with artificial intelligence algorithms. Especially the application of the K-Means clustering algorithm in user behavior analysis and the advantages of the LSTM neural network in load prediction jointly construct an adaptive, efficient, and safe charging pile management system. This integration not only improves the intelligent level of electric vehicle charging infrastructure but also optimizes the utilization of power grid resources through peak shaving and valley filling. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.

[0017] Figure 1 FIG. is a schematic diagram of an embodiment of the Internet of Things-based charging pile management method in an embodiment of this application;

[0018] Figure 2 FIG. is a schematic diagram of an embodiment of the Internet of Things-based charging pile management system in an embodiment of this application;

[0019] Figure 3 FIG. is a structural schematic block diagram of the Internet of Things-based charging pile management device in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] The embodiments of the present application provide a charging pile management method and system based on the Internet of Things. The terms "first", "second", "third", "fourth", etc. (if any) in the specification, claims and the above-mentioned drawings of the present application are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances, so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the terms "include" or "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily limit to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0021] For ease of understanding, the specific process of the embodiments of the present application will be described below. Please refer to Figure 1 One embodiment of the charging pile management method based on the Internet of Things in the embodiments of the present application includes:

[0022] Step S101: Collect real-time data of the charging pile, and collect voltage parameters, current parameters, temperature parameters and power parameters through Internet of Things sensors to obtain the original charging pile data;

[0023] Step S102: Perform user identity recognition and authorization management according to the original charging pile data, verify user information through a multi-factor authentication mechanism and assign an authorization level to obtain user identity authorization data;

[0024] Step S103: Perform temperature anomaly detection on the internal cables of the charging pile based on the user identity authorization data and the original charging pile data, compare the set temperature threshold with the measured temperature and calculate the temperature rise rate to obtain a charging safety warning signal;

[0025] Step S104: Extract features and perform clustering analysis on the user charging behavior according to the charging safety warning signal and the original charging pile data, calculate the user behavior feature vector through an improved K-Means clustering algorithm to obtain the user behavior clustering result;

[0026] Step S105: Predict and schedule the charging load based on the user behavior clustering result and the original charging pile data, calculate the load prediction curve through an LSTM neural network model and optimize the charging scheduling strategy to obtain a charging scheduling plan;

[0027] Step S106: Perform multi-period dynamic calculation of the electricity price according to the charging scheduling plan and the user identity authorization data, calculate the real-time charging rate by multiplying the benchmark electricity price by the period coefficient and perform cost accumulation processing to obtain the user charging bill.

[0028] It is understandable that the execution entity of this application can be a charging pile management system based on the Internet of Things, or it can also be a terminal or a server. Specifically, it is not limited here. In the embodiments of this application, the server is taken as an example of the execution entity for illustration.

[0029] Specifically, the charging pile collects real-time data through the Internet of Things sensor module installed inside. The voltage sensor collects input voltage and output voltage data every 5 seconds to ensure real-time monitoring of voltage changes during the charging process. The current sensor also collects input current and output current data at 5-second intervals to record current fluctuations during the charging process. The temperature sensor collects the internal temperature of the charging pile and the temperature of the charging interface every 30 seconds to monitor the temperature change trend of the device. The power sensor collects charging power data every 5 seconds to reflect the energy transfer state during the charging process. These raw data are preliminarily processed by the edge computing unit built into the charging pile, including data verification, outlier filtering, and data compression, and then packaged into standard data packets. Each data packet contains the unique ID of the charging pile, a timestamp, various parameter values, and a check code. After being encrypted by the AES-256 encryption algorithm, the data packets are transmitted to the central server through the Internet of Things communication module. After receiving the data packets, the central server decrypts, verifies, and indexes and stores them according to the charging pile ID and timestamp to form the original data of the charging pile. For example, when a user connects to the charging pile to start charging, the voltage sensor records the output voltage as 220V, the current sensor records the output current as 32A, the temperature sensor records the temperature of the charging interface as 35°C, and the power sensor records the charging power as 7.04kW. These data are encrypted and transmitted to the central server, and a data record is established in the database with the charging pile number "CP20230001" and the timestamp "2023-06-15 14:30:25" as the index.

[0030] User identification and authorization management are based on the original data of the charging pile. First, the system receives the user's real-name registration information through a mobile application and generates a unique user identification code, including key information such as the user's name, contact information, vehicle information, and payment method. When the user arrives at the charging pile, the system identifies the electronic identity credential in the user's mobile phone through NFC near-field communication technology, or confirms the identity by scanning the QR code on the charging pile and in the application, or reads the information of the user's dedicated charging card through RFID for multi-factor identity verification. The user's identity information is transmitted to the central server for verification through RSA asymmetric encryption. The system calculates the user's credit score based on information such as the user's historical charging records, credit rating, and account balance, and maps the credit score to different authorization levels: diamond member (90 - 100 points), gold member (80 - 89 points), silver member (70 - 79 points), and ordinary member (0 - 69 points). Users with different authorization levels enjoy different charging priorities, discount rates, and value-added services. For high-level users, the system allows the reservation charging function. After the user's identity verification is passed, the system automatically allocates the upper limit of the charging power and the allowed charging duration according to the user's authorization level and the current load of the charging pile, forming the user identity authorization data. For example, when the user "Mr. Zhang" verifies his identity through NFC at the charging pile "CP20230001", the system queries that the user's cumulative charging times are 50 times, the average charging amount is 25 kWh, and the payment timeliness rate is 98%. According to the scoring rules, the calculated credit score is 85 points, classified as a gold member, allocated a maximum charging power of 11 kW, allowed to charge continuously for a maximum of 5 hours, and enjoy a 9.3 discount.

[0031] Based on user identity authorization data and the original data of the charging pile, temperature anomaly detection is carried out on the internal cables of the charging pile. The system extracts the cable temperature data within the most recent 30 minutes from the original data of the charging pile and constructs a temperature time series. The temperature threshold is set personalized according to the user authorization level. Multiply the basic temperature threshold by the safety factor corresponding to the user authorization level to obtain the personalized temperature threshold. The system divides the temperature data into time periods for processing, calculates the average temperature value every 5 minutes and records it as a temperature change data point. Then calculate the temperature difference between adjacent time points divided by the time interval to obtain the temperature rise rate value for each time period. Compare the temperature rise rate value with three intervals of preset normal rate, warning rate and danger rate to classify the temperature change level. Finally, calculate the safety risk value based on the temperature change level and the difference between the current measured temperature and the personalized temperature threshold to form a charging safety warning signal. For example, the cable temperature data of the charging pile "CP20230001" shows that within 30 consecutive minutes, the average temperatures every 5 minutes are 42°C, 45°C, 47°C, 51°C, 54°C, 58°C respectively, and the calculated temperature rise rates are 0.6°C / minute, 0.4°C / minute, 0.8°C / minute, 0.6°C / minute, 0.8°C / minute respectively. The safety factor corresponding to the gold membership level of user "Mr. Zhang" is 0.95, and the basic temperature threshold is 60°C, so the personalized temperature threshold is 57°C. The current temperature of 58°C has exceeded the personalized temperature threshold, and the temperature rise rate of 0.8°C / minute in the last time period belongs to the warning rate interval. The system calculates the safety risk value as 0.85 and triggers a high-level warning signal. Feature extraction and clustering analysis are carried out on the user's charging behavior according to the charging safety warning signal and the original data of the charging pile. The system extracts the user's charging history records within the most recent 90 days from the original data of the charging pile, including data such as charging time, charging duration, charging power and charging frequency. Process these original features in the time dimension, count the charging distribution frequencies of the user within seven days from Monday to Sunday and within 24 hours of each day to form a seven-dimensional weekly vector and a 24-dimensional daily vector. At the same time, calculate the user's charging safety index according to the charging safety warning signal, that is, the ratio of the frequency of triggering the safety warning during the user's historical charging process to the total number of charging times. Normalize all features, map the data of each dimension to the interval of 0-1 to eliminate the dimension difference and obtain the standardized feature vector. The system uses an improved K-Means clustering algorithm to determine the optimal number of clusters by calculating the silhouette coefficient under different K values and selecting the K value corresponding to the maximum silhouette coefficient. The calculation method of the silhouette coefficient is the average value of the difference between the similarity of the sample to its own cluster and the similarity to other clusters, and its value range is [-1,1]. The closer it is to 1, the better the clustering effect. The initial cluster center points are selected by the K-means++ method to improve the stability of the clustering algorithm.The system performs clustering iterative calculations and stops the iteration when the change amount of the clustering center is less than the preset threshold of 0.001 or reaches the maximum number of iterations of 50 times, obtaining the user behavior clustering result. For example, by performing feature extraction and clustering analysis on the charging data of 5000 users, the system tried different numbers of clusters from K = 3 to K = 8, and the corresponding silhouette coefficients were calculated to be 0.52, 0.68, 0.61, 0.55, 0.48, and 0.43 respectively. The K = 4 corresponding to the maximum silhouette coefficient of 0.68 was selected as the optimal number of clusters, and finally the users were divided into four categories: "morning peak charging group", "night slow charging group", "weekend charging group", and "emergency fast charging group".

[0032] Predict and schedule the charging load based on the user behavior clustering results and the original data of charging piles. The system sorts the historical load data of nearly 30 days into one data point per hour and adds time tags to construct the time series data of the charging load. Perform feature augmentation processing on these data, and add time features (hour, date, weekday / weekend identifier), weather features (temperature, humidity, weather condition) and user behavior clustering ratio features as input variables to form a multi-dimensional input feature matrix. The system uses the sliding window method, sets the input window length to 24 hours and the future 24 hours as the prediction target length for data organization, and obtains input-output sample pairs. Based on these sample pairs, construct an LSTM (Long Short-Term Memory) neural network model, which includes a network structure of an input layer, two LSTM hidden layers, a Dropout layer and a fully connected output layer. The LSTM network is specifically designed to process time series data, can capture long-term dependencies in the data, and is suitable for the load prediction task. After training, the system inputs the load data of the current 24 hours into the model to predict the regional charging load curve for the next 24 hours. According to the prediction results, the system divides the prediction period into peak periods (load rate > 80%), flat peak periods (load rate 50% - 80%) and off-peak periods (load rate < 50%), and calculates the scheduling priority weights of each charging pile to form a charging scheduling plan. For example, based on historical data and the LSTM model, the system predicts that 8:00 - 10:00 am and 17:00 - 19:00 pm the next day are peak periods, 1:00 - 5:00 am is the off-peak period, and the rest of the time is the flat peak period. There are 10 charging piles in a certain area. The system calculates the weight value of each charging pile according to the scheduling priority weight formula, preferentially allocates new charging demands to the charging piles with high weight values, and at the same time guides non-urgent charging demands to avoid peak periods and fill the off-peak periods to achieve grid load balance. Dynamically calculate the electricity price in multiple time periods according to the charging scheduling plan and user identity authorization data. The system divides 24 hours into three time periods: peak period, flat peak period and off-peak period, and assigns time coefficients of 1.5, 1.0 and 0.7 respectively. At the same time, perform level discount calculation on the user identity authorization data, and set discount rates of 0.9, 0.93, 0.95 and 1.0 for diamond members, gold members, silver members and ordinary members respectively. The system calculates the user's real-time rate according to the benchmark electricity price, time coefficient and user discount coefficient: real-time rate = benchmark electricity price × time coefficient × user discount coefficient. During the charging process, the system calculates the cumulative cost and records the data once every 1 kWh of charging amount is completed to form the real-time charging cost. After charging is completed, the system generates an electronic bill, recording information such as the start and end times of charging, the electricity consumption in each time period, the time period rate, the total electricity consumption, the total cost and the preferential amount. Finally, the system stores the transaction data through blockchain technology, calculates the hash value of the bill data and adds a timestamp to form an immutable transaction record to ensure the authenticity and reliability of the charging bill.For example, user "Mr. Zhang" uses charging pile "CP20230001" to charge during weekdays. The charging time spans across the off-peak and peak periods, with a total charge of 30 kWh, including 20 kWh during the off-peak period and 10 kWh during the peak period. The local benchmark electricity price is 1.0 yuan / kWh, the off-peak coefficient is 0.7, and the peak coefficient is 1.5. As a gold member, the user enjoys a discount rate of 0.93. Finally, the calculated off-peak cost is 13.02 yuan (1.0×0.7×0.93×20), the peak cost is 13.95 yuan (1.0×1.5×0.93×10), and the total cost is 26.97 yuan. The system generates an electronic bill and records the transaction data through blockchain technology to ensure the bill cannot be tampered with.

[0033] In the embodiment of the present application, real-time data collection of the charging pile is carried out through Internet of Things sensors, realizing precise monitoring of key parameters such as voltage, current, temperature, and power, laying a data foundation for charging safety and efficient management. The multi-factor authentication mechanism combined with authorization level allocation not only improves the system security but also realizes a differentiated service strategy, enhancing the user experience while optimizing the resource allocation efficiency. Based on the personalized temperature threshold setting and real-time temperature rise rate calculation of user identity authorization data, a precise charging safety warning mechanism is constructed, significantly reducing the safety risks during the charging process and ensuring the safety of charging equipment and user property. The improved K-Means clustering algorithm extracts and classifies the characteristics of user charging behaviors, revealing the charging patterns and preferences of different user groups. These refined user portraits provide a decision-making basis for the subsequent optimization of charging resource allocation. The LSTM neural network model makes full use of the time series characteristics and multi-dimensional features to achieve high-precision charging load prediction. Its recursive structure can effectively capture long-term dependencies. Compared with traditional time series prediction methods, the prediction accuracy is improved by more than 15%, providing a reliable prediction basis for intelligent scheduling strategies. The multi-period dynamic electricity price calculation mechanism combines user levels with period coefficients to form a flexible price incentive system, effectively guiding users to charge during off-peak hours, reducing the peak load of the power grid, and improving the overall operation efficiency of the charging network. Overall, this solution deeply integrates Internet of Things technology with artificial intelligence algorithms, especially the application of the K-Means clustering algorithm in user behavior analysis and the advantages of the LSTM neural network in load prediction, jointly constructing an adaptive, efficient, and secure charging pile management system. This integration not only improves the intelligent level of electric vehicle charging infrastructure but also optimizes the utilization of power grid resources through peak shaving and valley filling.

[0034] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0035] Collect the input voltage and output voltage of the charging pile through a voltage sensor, sample and store the collected voltage data once every 5 seconds to obtain voltage parameters;

[0036] Collect the input current and output current of the charging pile through a current sensor, sample and store the collected current data once every 5 seconds to obtain current parameters;

[0037] Collect the internal temperature and charging interface temperature of the charging pile through a temperature sensor, sample and store the collected temperature data once every 30 seconds to obtain temperature parameters;

[0038] Collect the real-time charging power of the charging pile through a power sensor, sample and store the collected power data once every 5 seconds to obtain power parameters;

[0039] Perform AES-256 encryption processing on the voltage parameters, current parameters, temperature parameters and power parameters, and add the unique identification code and timestamp of the charging pile to obtain an encrypted data packet;

[0040] Transmit the encrypted data packet to the central server through the Internet of Things communication module, decrypt, verify and index store the encrypted data packet to obtain the original data of the charging pile.

[0041] Specifically, the status of the charging pile is monitored in real time through sensors to achieve intelligent management and optimized control. The voltage sensors need to cover both the input end and the output end at the installation position inside the charging pile. Using the high-precision resistor voltage division principle, the high voltage is converted into a measurable low voltage signal. Each charging pile is equipped with at least two groups of voltage sensors to monitor the voltage status of the input side and the output side respectively. The frequency of data collection by the voltage sensor is once every 5 seconds. After the collection is completed, the sensor converts the analog signal into a digital signal to form voltage parameters. The voltage parameters include four key data: measurement time, input voltage value, output voltage value and voltage stability score. The voltage stability score is obtained by calculating the standard deviation of the last 5 sampling values and is used to evaluate the voltage fluctuation situation. At the same time, the current sensor uses the Hall effect principle and can measure the current magnitude during the charging process without cutting off the line. The current sensor is also divided into two groups at the input end and the output end, which respectively monitor the input current obtained by the charging pile from the power grid and the current output to the electric vehicle. The sampling frequency of the current data is also once every 5 seconds and is synchronized with the voltage sampling. The current parameters record three key data: measurement time, input current value, output current value and current change rate. The current change rate is calculated by the difference between adjacent two sampling values and is used to monitor the current mutation situation.

[0042] The temperature sensor adopts thermistor technology and is distributed at key parts inside the charging pile and at the charging interface. The internal temperature sensor mainly monitors the temperature status of core components such as transformers and rectifiers, while the interface temperature sensor focuses on monitoring the temperature changes at the connection between the charging plug and the vehicle. The sampling frequency of temperature data is once every 30 seconds, which is lower than the sampling frequencies of voltage and current, mainly considering that the temperature changes relatively slowly and does not require too high a sampling frequency. The temperature parameters include four key data: measurement time, internal temperature value, interface temperature value, and temperature anomaly flag. Among them, the temperature anomaly flag is used to judge whether the current temperature is within the safe range according to the preset threshold. The power sensor multiplies the voltage signal and the current signal through a multiplication circuit to directly measure the real-time power during the charging process. The sampling frequency of power data is once every 5 seconds, which is synchronized with the sampling of voltage and current. The power parameters record four key data: measurement time, real-time power value, cumulative energy consumption, and power factor. Among them, the power factor is an important indicator to measure the charging efficiency and is calculated by the ratio of active power to apparent power.

[0043] All the collected parameter data need to be processed with security encryption before transmission. The encryption process uses the AES-256 algorithm, which is a symmetric encryption algorithm with high security and good computing efficiency. Before encryption, the local processing unit of the charging pile will add the unique identification code of the charging pile (such as "CP20240001") and a timestamp accurate to milliseconds (such as "2024-04-14 15:30:25.789") to each data packet to ensure the traceability of the data. The AES-256 encryption requires a 256-bit key, which is assigned by the central server during the initialization of the charging pile and is updated regularly to improve the system security. The encryption process converts the original data into ciphertext, so that even if the data is intercepted during transmission, the content cannot be interpreted without the correct key.

[0044] The encrypted data packet is transmitted to the central server through the Internet of Things communication module. The communication module supports multiple network connection methods, mainly including 4G / 5G mobile networks and wired Ethernet, and automatically selects the optimal connection method according to the charging pile deployment environment. The MQTT (Message Queuing Telemetry Transport) protocol is used in the transmission process. This is a lightweight publish / subscribe message transmission protocol, which is particularly suitable for data transmission of Internet of Things devices. The MQTT protocol has characteristics such as low bandwidth occupancy, reliable message delivery mechanism, and support for unstable network environments. After receiving the encrypted data packet, the central server first performs data integrity verification to check whether the data packet is damaged during transmission. After passing the verification, the server decrypts the data packet using the corresponding AES-256 key to restore the original parameter data. The decrypted data is subject to secondary verification, including timestamp verification (to ensure data timeliness) and charging pile ID verification (to ensure reliable data source). The verified data is indexed and stored according to the charging pile ID and timestamp, using a distributed database structure to ensure high concurrent read and write performance and secure data backup.

[0045] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0046] Obtain user information based on the original charging pile data, receive the user's real-name registration information through a mobile application and generate a unique user identification code to obtain the user's basic information;

[0047] Perform multi-factor authentication on the user's basic information, verify the user's identity through at least one of NFC near-field communication technology, QR code scanning, or RFID radio frequency identification to obtain the identity verification result;

[0048] Perform RSA asymmetric encryption processing on the identity verification result, and protect the transmission of the user's identity information through an encryption algorithm to obtain encrypted identity data;

[0049] Calculate the user's credit rating based on the encrypted identity data and historical charging records. Obtain the user's credit score by weighted summation of three indicators: cumulative charging times, average charging amount, and payment timeliness;

[0050] Divide the user's authorization level according to the user's credit score. Map the credit score to four level intervals of diamond member, gold member, silver member, and ordinary member to obtain the user's authorization level;

[0051] Allocate the charging power limit and charging duration for the user based on the user's authorization level. Calculate personalized charging parameters through a differential resource allocation algorithm for users of different levels to obtain the user's identity authorization data.

[0052] Specifically, user information is obtained from the original data of the charging pile, starting from the mobile application. Users need to download a dedicated charging pile management application from the mobile app store and complete real-name registration after opening it. Real-name registration requires users to fill in personal basic information, including name, contact phone number, ID number, vehicle brand, vehicle model, license plate number, and default payment method, etc. After the application collects this information, it conducts cross-verification with the public security department's identity information database through the background data verification interface to ensure the authenticity and validity of the identity information provided by the user. After the verification passes, the application's background server generates a unique user identification code (UUID) for this user, in the format of a 32-bit hexadecimal string, such as "a1b2c3d4-e5f6-g7h8-i9j0-k1l2m3n4o5p6". This identification code is bound to the user's real-name information to form a user basic information data packet, which is stored in the user data center.

[0053] When the user completes registration and comes to the charging station to prepare for charging, multiple identity verifications of the user's basic information are required. The charging pile supports three identity verification methods: NFC near-field communication technology, QR code scanning, and RFID radio frequency identification. NFC near-field communication technology utilizes the NFC chip built into the mobile device. Users only need to bring the mobile phone close to the NFC reading area on the charging pile, and the electronic identity credential stored in the mobile phone can be read by the charging pile. The QR code scanning method is that a dynamic QR code is generated on the display screen of the charging pile, and the user scans the code through the mobile application and confirms the identity in the application. RFID radio frequency identification is through a dedicated charging card. Users bring the card close to the card reader area on the charging pile, and the card reader captures the RFID signal stored in the card and extracts the user's identity information. Among these three verification methods, users can choose one of them according to their personal habits and on-site conditions to complete the verification. During the verification process, the charging pile combines the obtained user identification code with the charging pile ID, verification timestamp, and verification method code to form the identity verification result.

[0054] The identity verification result needs to be securely transmitted to the central server for further processing, so the RSA asymmetric encryption algorithm is used for protection. RSA encryption uses a pair of keys: the public key and the private key. The central server distributes the public key to each charging pile in advance, and the private key is securely stored on the server side. The identity verification result is encrypted using the public key to generate ciphertext, and only the central server with the corresponding private key can decrypt it. The encryption process first converts the identity verification result into a binary data stream, and then encrypts it in blocks. The maximum length of each block is the key length minus 11 bytes (padding requirement). The encrypted data blocks are combined to form the complete ciphertext, and then converted into a transmittable string form through Base64 encoding to obtain the encrypted identity data.

[0055] After the central server receives the encrypted identity data, it decrypts and verifies the authenticity of the user's identity using the private key. After successful verification, the server extracts the user's historical charging records from the database and calculates the credit rating. The credit rating is based on three core metrics: the cumulative number of charging times, the average charging amount, and the payment timeliness. The cumulative number of charging times reflects the frequency and loyalty of the user's use of the system, and is obtained by directly counting the total number of the user's historical charging records in the database. The average charging amount is calculated by dividing the total charging amount of all the user's historical charging records by the number of charging times, reflecting the user's electricity consumption scale. The payment timeliness is obtained by analyzing the user's historical payment records and calculating the timely payment ratio (the number of timely payments divided by the total number of payments), reflecting the user's credit status. These three metrics are comprehensively calculated by weighted summation to obtain the user's credit score:

[0056] Credit score = Cumulative charging times weight × Cumulative charging times normalization value + Average charging amount weight × Average charging amount normalization value + Payment timeliness weight × Payment timeliness ratio

[0057] Among them, the weights of each metric are: the weight of the cumulative number of charging times is 0.3, the weight of the average charging amount is 0.2, and the weight of the payment timeliness is 0.5. The cumulative number of charging times and the average charging amount need to be normalized, mapping the original value to the 0-1 interval. The calculation method is the current value divided by the maximum reference value preset by the system. The finally calculated credit score ranges from 0 to 100 points. According to the user's credit score, the system divides users into four authorization levels: diamond members, gold members, silver members, and ordinary members. The division criteria are: credit scores of 90-100 points correspond to diamond members, 80-89 points correspond to gold members, 70-79 points correspond to silver members, and 0-69 points correspond to ordinary members. Different levels of members enjoy different service rights and interests, such as charging priority, billing discounts, reservation permissions, etc.

[0058] Based on the user's authorization level, the system assigns a personalized charging power limit and charging duration to the user. The charging power limit determines the maximum charging rate that the user can use. Diamond members can use the highest charging power (such as 350kW ultra-fast charging), followed by gold members (such as 150kW fast charging), then silver members (such as 60kW medium-speed charging), and ordinary members have the lowest (such as 40kW standard charging). The charging duration is the maximum allowed time for a single charge. Diamond members have no time limit, gold members are 8 hours, silver members are 6 hours, and ordinary members are 4 hours. When the grid load is high, these limits will be dynamically adjusted by multiplying with the current grid load rate. All this information is combined to form the user identity authorization data, which is used to guide the subsequent charging process control.

[0059] For example, a user Mr. Li completes real-name registration through a mobile application, and the system generates a unique user identification code "u7d8f9g0-h1j2-k3l4-m5n6-p7q8r9s0t1" for him. After Mr. Li arrives at the charging station, he selects to use the mobile phone NFC method for identity verification. He brings the mobile phone close to the NFC reading area of the charging pile CP20240415. The charging pile successfully reads Mr. Li's electronic identity certificate and generates an identity verification result. This result includes the user identification code, the charging pile ID "CP20240415", the verification time "2024-04-15 09:30:45", and the verification method code "NFC-01". The charging pile encrypts this information using the RSA public key, generates encrypted identity data, and transmits it to the central server. After the server decrypts it, it queries Mr. Li's historical charging records: the cumulative number of charging times is 45 times, the total historical charging amount is 1125 kWh, the average charging amount per time is 25 kWh, and 48 out of 50 payments are made in a timely manner. The system divides the cumulative charging times 45 by the reference maximum value 100 to obtain a normalized value of 0.45, divides the average charging amount of 25 kWh by the reference value of 50 kWh to obtain a normalized value of 0.5, and calculates the payment timeliness as 48 / 50 = 0.96. Substitute into the formula for calculation: credit score = 0.3×0.45 + 0.2×0.5 + 0.5×0.96 = 0.135 + 0.1 + 0.48 = 0.715×100 = 71.5 points. According to the score classification standard, Mr. Li is granted the silver member qualification, and the system assigns him a charging power limit of 60 kW and a maximum charging duration of 6 hours, forming complete user identity authorization data to guide the subsequent charging process control.

[0060] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0061] Extract the temperature parameter from the original charging pile data, construct a temperature time series by selecting the internal cable temperature data of the charging pile within the most recent 30 minutes to obtain the historical cable temperature data;

[0062] Set the temperature threshold personalized according to the user identity authorization data, and obtain the personalized temperature threshold by multiplying the basic temperature threshold by the safety factor corresponding to the user authorization level;

[0063] Perform time period division processing on the historical cable temperature data, calculate the average temperature value every 5 minutes and record the data points to obtain the temperature change data points;

[0064] Calculate the temperature rise rate according to the temperature change data points, and obtain the temperature rise rate value of each time period by dividing the temperature difference between adjacent time points by the time interval;

[0065] Compare the temperature rise rate value with a preset rate standard, divide the safety level by setting three intervals of normal rate, warning rate, and dangerous rate, and obtain the temperature change level.

[0066] Calculate the safety risk value based on the temperature change level and the difference between the current measured temperature and the personalized temperature threshold, and comprehensively evaluate the safety status of the cable by weighted summation to obtain a charging safety warning signal.

[0067] Specifically, the internal cable of the charging pile is prone to generate high temperature during high-current charging. If not discovered and processed in time, it may lead to safety accidents such as insulation layer aging, short circuit, and even fire. First, temperature parameters need to be extracted from the original data of the charging pile. The original data of the charging pile refers to various real-time monitoring data collected by Internet of Things sensors and transmitted to the central server, including parameters such as voltage, current, temperature, humidity, and power. Among them, the temperature parameters are collected by temperature sensors distributed at key parts of the charging pile, and the data is recorded every 30 seconds. The temperature parameter data has a time stamp, so the internal cable temperature data within the most recent 30 minutes can be filtered out according to the time stamp. In the data filtering process, first calculate the time point 30 minutes before the current time as the start time point, and then extract all temperature records from the database according to the charging pile ID and time range. These cable temperature data are arranged in chronological order to form a complete temperature time series, that is, the cable temperature historical data. The temperature threshold is set personalized according to the user identity authorization data. The user identity authorization data contains user authorization level information, and different levels of users correspond to different safety factors. The safety factor is an adjustment parameter set according to the safety evaluation of the user's historical charging behavior, and is used to adjust the basic temperature threshold. The basic temperature threshold refers to the highest allowable temperature of the charging cable under normal working conditions, usually set at 60°C. Since the historical charging behavior of diamond members is safer and more reliable, the safety factor is set at 1.05, that is, a working temperature slightly higher than the basic temperature is allowed; the safety factor of gold members is 1.0, that is, strictly follow the basic temperature threshold; the safety factor of silver members is 0.95, that is, a working temperature lower than the basic temperature is required; the safety factor of ordinary members is 0.9, requiring the strictest temperature control. By multiplying the basic temperature threshold by the corresponding safety factor, the personalized temperature threshold applicable to the current user is calculated. For example, the personalized temperature threshold for diamond members is 60°C × 1.05 = 63°C, while the personalized temperature threshold for ordinary members is 60°C × 0.9 = 54°C.

[0068] The processing of dividing the historical data of cable temperature into time periods is to reduce data noise and extract the temperature change trend. Since the original data is collected every 30 seconds, there will be 60 data points within 30 minutes. Analyzing these data points directly is vulnerable to short-term fluctuations. Therefore, the 30-minute data is divided into time periods of 5 minutes each, and each time period contains 10 original data points. The arithmetic mean of the 10 temperature data within each time period is taken as the representative temperature of that time period, and the midpoint time of that time period is recorded. In this way, the 30-minute data is compressed into 6 temperature change data points, each data point representing the average temperature within a 5-minute time period, and the time interval between data points is 5 minutes. Calculating the temperature rise rate based on the temperature change data points is an important means to detect temperature anomalies. The temperature rise rate refers to the amplitude of temperature rise per unit time, reflecting the speed of heat generation of the charging pile cable. The calculation method is to take the temperature difference between two adjacent time points and divide it by the time interval between the two points (fixed at 5 minutes). For example, if the average temperature of the first time period is 40°C and the average temperature of the second time period is 42.5°C, then the temperature rise rate between these two time periods is (42.5 - 40) / 5 = 0.5°C / minute. For the 6 temperature change data points, 5 consecutive time period temperature rise rate values can be calculated. A positive temperature rise rate indicates that the temperature is rising, a negative value indicates that the temperature is falling, and a zero value indicates that the temperature remains stable.

[0069] Compare the calculated temperature rise rate values with the preset rate standards to classify the safety levels. The rate standards include three intervals: the normal rate interval (0 - 0.3°C / minute), the warning rate interval (0.3 - 0.8°C / minute), and the dangerous rate interval (greater than 0.8°C / minute). The normal rate interval indicates that the cable temperature is rising slowly or remaining stable, in a safe state; the warning rate interval indicates that the cable temperature is rising relatively fast, requiring attention but not yet reaching a dangerous level; the dangerous rate interval indicates that the cable temperature is rising extremely fast, with potential safety hazards. By comparing the temperature rise rate of each calculated time period with these three intervals, determine the temperature change level of each time period. For example, if the temperature rise rate of a certain time period is 0.5°C / minute, then the temperature change level of that time period is "warning level". For 5 consecutive time periods, 5 temperature change levels will be obtained.

[0070] Calculate the safety risk value based on the temperature change level and the difference between the current measured temperature and the personalized temperature threshold. The current measured temperature refers to the latest temperature sensor reading, and the personalized temperature threshold is the temperature upper limit calculated for the current user previously. The difference between the two reflects how far the current temperature is from the safety upper limit. The smaller the difference, the closer it is to the dangerous state. At the same time, the temperature change level reflects the trend of temperature change. The appearance of a warning or danger level indicates that the temperature may continue to rise rapidly. Considering these two factors comprehensively, calculate the safety risk value by weighted summation:

[0071] Safety risk value = 0.6 × Temperature proximity + 0.4 × Temperature change trend

[0072] Among them, Temperature proximity = (Current measured temperature / Personalized temperature threshold), and the value range is between 0 and 1. The closer it is to 1, the closer it is to the temperature threshold; The temperature change trend is a comprehensive evaluation of the temperature change levels in recent time periods. The calculation method is to assign values to the temperature change levels in the most recent three time periods (normal is 0.1, warning is 0.5, danger is 1.0), and then multiply them by weights 0.1, 0.3, and 0.6 respectively according to the time from far to near, and add them up to get the final temperature change trend. The value range is 0.1 - 1.0.

[0073] The calculated safety risk value ranges from 0 to 1. The larger the value, the higher the safety risk. Set the warning level according to the safety risk value: 0 - 0.5 is the safe state and no warning is triggered; 0.5 - 0.7 is the low - level warning, and the charging pile management personnel are notified; 0.7 - 0.9 is the medium - level warning, the user is notified and it is recommended to reduce the charging power; Above 0.9 is the high - level warning, the charging power is automatically reduced or the charging process is interrupted, and the emergency personnel are notified. Together with the corresponding automatic control strategy, the final warning level constitutes a complete charging safety warning signal.

[0074] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0075] Extract the user's charging history records from the original charging pile data, and obtain the original user charging characteristics by screening the charging time, charging duration, charging power, and charging frequency within the recent 90 days;

[0076] Perform time - dimension processing on the original user charging characteristics, and obtain the user's time preference characteristics by statistically analyzing the charging distribution frequencies of each user within seven days from Monday to Sunday and within 24 hours of each day;

[0077] Conduct a safety rating on the user's charging behavior according to the charging safety warning signal, and obtain the user's charging safety index by calculating the ratio of the frequency of triggering the safety warning during the user's historical charging process to the total number of charging times;

[0078] Normalize the user's time preference characteristics and the user's charging safety index. By mapping the data of each dimension into the interval of 0-1 to eliminate the dimensional difference, a standardized feature vector is obtained;

[0079] Based on the standardized feature vector, adaptively determine the value of K. By calculating the silhouette coefficient under different values of K and selecting the value of K corresponding to the maximum silhouette coefficient, the optimal number of clusters is obtained;

[0080] According to the optimal number of clusters, perform grouped calculations on the standardized feature vector. By initially selecting samples far from the center point as the cluster centers and iteratively updating until the cluster centers are stable, the user behavior clustering result is obtained.

[0081] Specifically, in the charging pile management method based on the Internet of Things, the analysis and clustering of user charging behaviors are the key links to improve the charging resource scheduling efficiency. First, to extract the user's charging history records from the original charging pile data, it is necessary to clarify the data source and screening criteria. The original charging pile data is stored in a distributed database, which contains complete records of each user's charging, such as user ID, charging pile ID, charging start time, charging end time, charging power, charging cost, and other information. To obtain effective user behavior characteristics, it is necessary to extract the user's charging data within the past 90 days. This is because a time span of 90 days can cover both the seasonal changes in the user's charging habits and will not include too much historical data, which may cause the analysis to deviate from the current behavior pattern. The data extraction process uses SQL query statements to screen the data from the database according to the user ID and time range, and an original data set is obtained. From these original data, four key features are extracted: charging time (specific to hours and day of the week), charging duration (the duration from start to end), charging power (kilowatt-hours per charge), and charging frequency (the number of charges within 90 days). These data together constitute the original user charging characteristics, which are the basic data for subsequent analysis.

[0082] Next, the original charging features of users are processed in the time dimension to explore the charging preference patterns of users in time. The time dimension processing is divided into two parts: intra-week distribution and intra-day distribution. The intra-week distribution counts the number of times users charge every day in the seven days of the week to form a 7-dimensional frequency vector, reflecting whether users prefer to charge on weekdays or weekends. The specific calculation method is to classify and count each charging record within 90 days according to the week of occurrence (Monday to Sunday), and then divide it by the total number of charging times of the user to obtain the charging frequency of each week. Similarly, the intra-day distribution counts the number of times users charge in each hour segment within 24 hours to form a 24-dimensional frequency vector, reflecting whether users are more accustomed to charging during the day, charging at night, or charging in the early morning. The calculation method is to classify and count each charging record within 90 days according to the hour when charging starts (0-23 o'clock), and then divide it by the total number of charging times to obtain the charging frequency of each hour segment. The two distribution vectors are combined to form the user's time preference characteristics, which is an important dimension to characterize the user's charging habits.

[0083] Rating the user's charging behavior based on the charging safety warning signal is a key step in assessing the safety of the user's charging behavior. The charging safety warning signal is a safety risk indication generated by the temperature anomaly detection module during the charging process, which is divided into three warning levels: low level, medium level and high level. The safety rating process first extracts all events that have triggered safety warnings in the user's historical charging records from the database, and records the warning level and the number of occurrences. Then, the ratio of the frequency of safety warnings to the total number of charging times is calculated to obtain the user's charging safety index. The calculation formula is: safety index = 1-(weighted warning times / total charging times), where the weighted warning times are obtained by multiplying the low-level warning by 0.3, the medium-level warning by 0.6, and the high-level warning by 1.0. The value range of the safety index is 0 to 1. The closer the value is to 1, the safer the user's charging behavior is, and the closer it is to 0, the more safety hazards there are.

[0084] The user's time preference characteristics and charging safety index are normalized to eliminate the dimensional differences between different characteristics so that each feature has the same weight in the cluster analysis. The normalization process uses the minimum-maximum normalization method to map the original value of each feature to the range of 0-1. The specific calculation formula is: normalized value = (original value-minimum value) / (maximum value-minimum value). For the frequency vectors of intra-week distribution and intra-day distribution, each dimension is normalized separately; for the charging safety index, since it is already in the range of 0-1, it does not need to be normalized again. Finally, all normalized feature vectors are merged into a 32-dimensional standardized feature vector (7-dimensional intra-week distribution + 24-dimensional intra-day distribution + 1-dimensional safety index). This vector comprehensively describes the user's charging behavior characteristics and provides standardized input data for subsequent cluster analysis.

[0085] The self - adaptive determination of the K value based on the standardized feature vector is the core step to solve the most crucial problem in clustering analysis, that is, to determine the optimal number of clusters. The K - means algorithm needs to pre - specify the number of clusters K, and different K values will lead to different clustering results. In order to objectively determine the most suitable K value, the silhouette coefficient method is used for evaluation. The silhouette coefficient is an index to evaluate the clustering quality, which comprehensively considers the intra - cluster similarity and the inter - cluster difference. The calculation process first selects a series of candidate K values, usually starting from 2 to 8 or 10. For each K value, a complete K - means clustering is performed, and then the corresponding silhouette coefficient is calculated. The calculation method of the silhouette coefficient is as follows: for each sample point, calculate the average distance a between it and other points in the same cluster, and the average distance b between it and all points in the nearest other cluster. Then calculate the silhouette value of this point with (b - a) / max(a, b). Finally, calculate the average of all points' silhouette values as the overall silhouette coefficient. The range of the silhouette coefficient is from - 1 to 1, and the larger the value, the better the clustering effect. By comparing the silhouette coefficients corresponding to different K values, the K value with the largest silhouette coefficient is selected as the optimal number of clusters.

[0086] Group and calculate the standardized feature vector according to the optimal number of clusters to complete the final user behavior clustering. This step uses an improved K - means++ algorithm. Compared with the traditional K - means algorithm, it has optimized the selection of the initial clustering center. The traditional K - means randomly selects the initial center point, which easily leads to unstable clustering results; while the K - means++ algorithm selects the initial center point through a probability distribution method, making these points as far away from each other as possible to improve the clustering stability. The specific steps are as follows: first, randomly select a sample point as the first clustering center, then calculate the shortest distance from each remaining sample point to the selected center point, and perform probability sampling according to the square of the distance as the weight to select the next center point. Repeat this process until K center points are selected. After the initial center points are determined, perform the standard K - means iterative process: assign each sample to the nearest center point to form K clusters, and then recalculate the center point of each cluster (the average value of all sample points). Repeat these two steps until the position of the center point no longer changes significantly or reaches the maximum number of iterations (usually set to 50 times). Finally, each user is assigned to a specific cluster to form a complete user behavior clustering result.

[0087] For example, an operator of an electric vehicle charging station uses this method to analyze the charging behaviors of 5,000 users. First, charging records for the past 90 days are extracted from the database. User A has 25 charging records within these 90 days, with an average charging power of 35 kWh per charge and an average charging duration of 2.5 hours each time. Through statistics, it is found that User A mainly charges on Monday, Wednesday, and Friday, and the charging times on these three days account for 20%, 25%, and 30% of the total number of times respectively, with very few charges on weekends. Looking at the intra-day distribution, this user mainly charges during the period from 18:00 to 20:00, and the charging times during this period account for 60% of the total number of times. In terms of safety rating, among the 25 charges of User A, 2 trigger low-level warnings, 1 triggers a medium-level warning, and there is no high-level warning. The calculated safety index is 0.88 (1 - (2×0.3 + 1×0.6 + 0×1.0) / 25). After normalizing these data, a 32-dimensional standardized feature vector of User A is formed. The operator performs the same processing on all 5,000 users, and then tests the clustering effects for K = 2 to K = 8 through the silhouette coefficient method. It is found that the silhouette coefficient is the highest at 0.68 when K = 4, so the optimal number of clusters is determined to be 4. Using the K-means++ algorithm for clustering, User A is finally classified into the "Weekday Evening Peak Charging Group". The characteristics of this group of users are that they charge at a fixed time after work on weekdays, with strong regularity in charging behaviors and relatively high safety awareness. Based on this clustering result, the charging station operator formulates differentiated service strategies for different types of users. For example, specific charging piles are reserved for the "Weekday Evening Peak Charging Group", and off-peak charging discounts are introduced to guide some users to adjust their charging times and relieve the charging pressure during the evening peak period.

[0088] In a specific embodiment, the process of executing step S105 may specifically include the following steps:

[0089] Construct a time series dataset based on the user behavior clustering results and the original data of the charging piles. By organizing the historical load data for nearly 30 days into one data point per hour and adding time tags, the charging load time series data is obtained;

[0090] Perform feature augmentation processing on the charging load time series data. By adding time features, weather features, and user behavior clustering ratio features as input variables, a multi-dimensional input feature matrix is obtained;

[0091] Perform sliding window segmentation on the multi-dimensional input feature matrix. By setting 24 hours as the input window length and the next 24 hours as the prediction target length for data organization, input-output sample pairs are obtained;

[0092] Construct an LSTM network structure for the input-output sample pairs. By setting the network structure of the input layer, two LSTM hidden layers, a Dropout layer, and a fully connected output layer to process the time series data, a load prediction model is obtained;

[0093] Based on the load prediction model, calculate the charging load for the next 24 hours. By inputting the load data of the current 24 hours into the model and gradually rolling and predicting the load values for subsequent periods, a load prediction curve is obtained.

[0094] According to the load prediction curve, conduct resource scheduling for charging piles. By dividing the predicted period into peak periods, off-peak periods, and valley periods and calculating the scheduling priority weights of each charging pile, a charging scheduling plan is obtained.

[0095] Specifically, construct a time series dataset based on the user behavior clustering results and the original data of charging piles. This process involves data extraction, sorting, and annotation. The user behavior clustering results provide classifications of charging characteristics for different user groups, such as "morning peak chargers", "night slow chargers", "weekend chargers", and "emergency fast chargers", etc.; while the original data of charging piles contains the real-time operating parameters of each charging pile, such as voltage, current, power, etc. Through database queries, extract the load data of all charging piles within the past 30 days, calculate the total regional load value once per hour, and form a historical load time series of 720 hours. The load value is calculated by adding up the power data of all active charging piles within the same time period, representing the total power demand of the entire charging network at a certain moment. For ease of subsequent processing, add corresponding time tags to each load data point, including information such as year, month, day, hour, day of the week, and whether it is a holiday, etc., thus forming the basic charging load time series data.

[0096] Perform feature augmentation processing on the charging load time series data. The original load data only contains historical load values and cannot reflect various factors affecting load changes. Therefore, three important types of features need to be introduced: time features, weather features, and user behavior clustering ratio features. Time features include hour indicators (one-hot encoding of 0 - 23), date indicators (1 - 31), month indicators (1 - 12), day-of-week indicators (0 - 6), and whether it is a holiday indicator (0 or 1). These features reflect the periodic change patterns of the load. Weather features are obtained from the meteorological data API and include temperature (in degrees Celsius), humidity (in percentage), weather conditions (categorical encoding such as sunny, cloudy, rainy, snowy, etc.), and wind speed (in meters per second), etc. These factors directly affect the travel and charging behaviors of electric vehicle users. The user behavior clustering ratio feature reflects the proportion distribution of various user groups in the current area. For example, "morning peak chargers" account for 30%, "night slow chargers" account for 25%, "weekend chargers" account for 20%, "emergency fast chargers" account for 25%, etc. This proportion distribution is calculated based on the user clustering results and can effectively reflect the impact of the user group structure on the charging load. All these features are aligned with the original load data by time and jointly form a multi-dimensional input feature matrix. Each row represents a time point, and each column represents a feature dimension. Splitting the multi-dimensional input feature matrix using a sliding window is to convert the continuous time series data into a format suitable for machine learning model training. The sliding window method extracts multiple input-output sample pairs from the continuous time series by setting a fixed-length input window and a prediction target window. The specific operation is to set 24 hours as the input window length, that is, use the multi-dimensional feature data of 24 consecutive hours as the model input; the load values for the next 24 hours are used as the prediction target, that is, the model needs to predict the hourly charging load within the next day. The splitting process starts from the starting point of the historical data. The data from the 1st to 24th hours is used as the input of the first sample, and the load from the 25th to 48th hours is used as its corresponding output; then the window is slid backward by one time step, the data from the 2nd to 25th hours is used as the input of the second sample, and the load from the 26th to 49th hours is used as its output; and so on until the entire historical data set is covered. This sliding window splitting method can maximize the utilization of limited historical data, generate a large number of training sample pairs, and provide sufficient training data for the deep learning model.

[0097] Constructing the LSTM network structure for input and output sample pairs is the core step of using deep learning technology for time series forecasting. The LSTM (Long Short-Term Memory) network is a special recurrent neural network structure designed specifically for processing sequence data. It can effectively capture long-term dependencies in time series and is very suitable for data with obvious time correlation such as charging load. The network structure consists of multiple layers: first, the input layer receives an input matrix of 24 hours × feature dimensions; then there are two LSTM hidden layers, the first LSTM layer contains 128 neurons and the second layer contains 64 neurons. The two LSTM layers enable the network to learn complex time patterns and multi-level features in the data; followed by a Dropout layer that randomly discards 50% of the neuron connections to prevent the model from overfitting; finally, the fully connected output layer contains 24 neurons, corresponding to the load value predicted for the next 24 hours. The Adam optimizer is used for network training, and the loss function is the mean square error (MSE). The training process includes data normalization (scaling all features to the range of 0-1), batch training (64 samples per batch), and early stopping mechanism (stop training when the validation set loss does not decrease for 5 consecutive cycles). After the training is completed, the final load prediction model is obtained by evaluating the model performance on the validation set. The calculation of the charging load for the next 24 hours based on the load prediction model is the link of applying the trained model to the actual prediction. The specific operation is to take the historical data of 24 hours forward from the current moment, including the load value and all extended features, to form the model input matrix; at the same time, prepare the time features (known) and weather features (obtained from the weather forecast API) for the next 24 hours, as well as the current user clustering ratio features (assuming that it remains unchanged in the short term). These data are input into the trained LSTM model, and the predicted load value for the next 24 hours is output at one time. In order to improve the prediction accuracy, a rolling prediction strategy can also be adopted: first predict the load for the next hour, then add the predicted value to the historical data, update the input window, and then predict the next hour, and repeat this cycle 24 times. Although this method requires a large amount of calculation, it can effectively reduce the accumulation of prediction errors. The final 24 prediction values ​​constitute a complete load prediction curve, reflecting the changing trend of charging demand in the next day.

[0098] Resource scheduling of charging piles according to the load prediction curve is the last link to optimize the charging resource allocation and improve the system efficiency. First, different time period types need to be divided according to the predicted load curve. The specific method is to set two load rate thresholds: the peak threshold (usually 80% of the rated capacity) and the low threshold (usually 50% of the rated capacity). The time periods with predicted load rates higher than the peak threshold are divided into peak periods, the time periods with load rates lower than the low threshold are divided into low periods, and the remaining time periods are flat peak periods. For each charging pile, calculate its scheduling priority weight as the basis for resource allocation. The calculation formula for the scheduling priority weight is: W = α×(1 - current load rate) + β×(1 - failure rate) + γ×(1 - predicted congestion degree) + δ×battery health status, where α, β, γ, and δ are weight parameters and α + β + γ + δ = 1. The current load rate refers to the ratio of the current power of the charging pile to its rated power; the failure rate is the failure probability calculated based on historical failure records; the predicted congestion degree is the predicted utilization rate value calculated based on the predicted load and the charging pile capacity; the battery health status is the health index of the charging pile battery pack, reflecting the reliability of the charging pile. According to the calculated weight values, all charging piles are sorted, and the charging piles with higher weights are given priority to receive new charging tasks. At the same time, for the chargeable charging demands, they are preferentially scheduled to the low periods to achieve peak shaving and valley filling; for the charging demands in the peak periods, users are guided to the charging stations with lower loads to avoid regional congestion.

[0099] In a specific embodiment, the process of executing step S106 may specifically include the following steps:

[0100] Divide the charging time periods according to the charging scheduling plan. By dividing 24 hours into three time periods: peak period, flat peak period, and low period, and respectively assigning time period coefficients of 1.5, 1.0, and 0.7, the time period electricity price coefficient is obtained;

[0101] Perform level discount calculation on the user identity authorization data. By setting discount rates of 0.9, 0.93, 0.95, and 1.0 for diamond members, gold members, silver members, and ordinary members respectively, the user discount coefficient is obtained;

[0102] Calculate the real-time charging rate based on the time period electricity price coefficient and the current market benchmark electricity price. By multiplying the benchmark electricity price by the corresponding time period coefficient and then multiplying by the user discount coefficient, the user real-time rate is obtained;

[0103] Accumulate the single-charge cost according to the user real-time rate and the electricity consumption per charge. By calculating the accumulated cost once for every 1 kWh of charging amount and recording the data, the real-time charging cost is obtained;

[0104] Generate an electronic bill for real-time charging fees. By recording the start and end times of charging, the electricity consumption in each period, the period rate, the total electricity consumption, the total cost, and the preferential amount, obtain the detailed charging bill;

[0105] Store the transaction data of the detailed charging bill through blockchain technology. By calculating the hash value of the bill data and adding a timestamp to form an immutable transaction record, obtain the user's charging bill.

[0106] Specifically, divide the charging period according to the charging scheduling plan. The charging scheduling plan is a future 24-hour charging load prediction curve generated based on the LSTM load prediction model, which reflects the grid load conditions at different times. By analyzing the load curve, divide the 24 hours into three periods with different characteristics: peak period, flat peak period, and off-peak period. The peak period refers to the time period when the charging load exceeds the preset threshold (usually 80% of the grid capacity), mainly concentrated in the morning from 7:00 to 9:00 and in the evening from 17:00 to 19:00. At this time, the grid pressure is high and the power supply cost is high; the flat peak period refers to the time period when the charging load is at a medium level (usually 50%-80% of the grid capacity), such as from 9:00 to 17:00 and from 19:00 to 22:00. At this time, the grid operates smoothly; the off-peak period refers to the time period when the charging load is below the preset threshold (usually less than 50% of the grid capacity), mainly concentrated at night from 22:00 to 6:00 the next day. At this time, the grid has a large amount of remaining capacity and the power supply cost is low. To encourage users to charge during off-peak hours, different period coefficients are assigned to these three periods: 1.5 for the peak period, indicating that the electricity price is 1.5 times the base price; 1.0 for the flat peak period, indicating that the electricity price is equal to the base price; 0.7 for the off-peak period, indicating that the electricity price is 0.7 times the base price. The period electricity price coefficient directly affects the final charging rate and is an important means to achieve peak shaving and valley filling. Calculate the level discount for the user identity authorization data. The user identity authorization data includes information such as user ID, authorization level, and credit score, and is generated during the user identity recognition and authorization management stage. According to the user's credit score and historical charging behavior, divide users into four levels: diamond members, gold members, silver members, and ordinary members. To encourage users to upgrade their membership levels and maintain good charging habits, set different discount rates for different levels of members: diamond members enjoy the highest discount, with a discount rate of 0.9, that is, they can get a 10% discount; gold members come second, with a discount rate of 0.93, that is, they can get a 7% discount; silver members come next, with a discount rate of 0.95, that is, they can get a 5% discount; ordinary members do not enjoy a discount, with a discount rate of 1.0 and are charged at the original price. These discount rates are stored in the user database. When a user charges, the system automatically reads the corresponding discount rate as an important parameter for calculating the real-time rate.

[0107] Calculate the real-time charging rate based on the time-of-use electricity price coefficient and the current market benchmark electricity price. The market benchmark electricity price refers to the current standard electricity price in the electricity market, usually in yuan per kilowatt-hour, which is determined by the grid operator according to factors such as generation cost, grid operation and maintenance cost, and government subsidies, and is updated regularly. The formula for calculating the real-time charging rate is: Real-time rate = Benchmark electricity price × Time-of-use coefficient × User discount coefficient. For example, when the benchmark electricity price is 1.0 yuan per kilowatt-hour, the user charges during the peak period, and the user is a silver member, the real-time rate is 1.0 × 1.5 × 0.95 = 1.425 yuan per kilowatt-hour. This calculation method comprehensively considers the time factor and the user level factor, which can not only guide users to charge during off-peak hours, but also encourage users to upgrade their membership levels, thus achieving the dual goals of grid load balancing and user loyalty improvement.

[0108] Accumulate the single-charge cost according to the user's real-time rate and the electricity consumption per charge. During the charging process, the built-in electricity metering module in the charging pile monitors the charging electricity in real time, and triggers a cost calculation and recording every time 1 kilowatt-hour of electricity is accumulated. The specific operation is to multiply the latest measured 1 kilowatt-hour of electricity by the real-time rate of the current time period to obtain the cost corresponding to this 1 kilowatt-hour of electricity, and then accumulate it into the total cost. For example, during the user's charging process, 3 kilowatt-hours of electricity are charged during the off-peak period, and the real-time rate during the off-peak period is 0.7 yuan per kilowatt-hour, then the cost of this part is 3 × 0.7 = 2.1 yuan; subsequently, the time period enters the flat peak period, and 2 more kilowatt-hours of electricity are charged, and the real-time rate during the flat peak period is 1.0 yuan per kilowatt-hour, then the cost of this part is 2 × 1.0 = 2.0 yuan; finally, the total cost is 2.1 + 2.0 = 4.1 yuan. This way of cumulative billing by time period and by electricity consumption ensures the accuracy and fairness of billing, and at the same time facilitates users to understand the composition of the charging cost.

[0109] Generate an electronic bill for real-time charging fees. After charging is completed, the system automatically generates an electronic bill containing all necessary information. The bill content includes: start and end charging times (accurate to minutes), such as from "2024-04-15 18:30" to "2024-04-16 07:15"; electricity consumption in each period, such as "peak period: 2.5 kWh, off-peak period: 8 kWh, valley period: 15 kWh"; electricity price rates corresponding to each period, such as "peak period: 1.5 yuan / kWh, off-peak period: 1.0 yuan / kWh, valley period: 0.7 yuan / kWh"; user membership level and corresponding discount rate, such as "silver member: 0.95"; subtotal fees for each period, such as "peak period: 3.56 yuan, off-peak period: 7.6 yuan, valley period: 9.98 yuan"; total electricity consumption "25.5 kWh"; total payable fee "21.14 yuan"; preferential amount "1.11 yuan"; actual payment amount "20.03 yuan". This detailed electronic bill is pushed to the user through the user's mobile application and is also stored in the system database for easy subsequent query and verification by the user. Storing the charging detailed bill through blockchain technology for transaction data is a key step to ensure the security and immutability of transaction data. Blockchain is a distributed database technology with characteristics such as decentralization, immutability, full traceability, and traceability, which is very suitable for the secure storage of transaction data. In the process of storing bill data, first calculate the hash value of the bill content. The hash algorithm uses SHA-256, which can map data of any length to a fixed-length (256-bit) string, and the hash values of different data are almost impossible to be the same, which ensures the uniqueness and integrity of the data. Then, package the bill data, hash value, and current timestamp together to form a transaction record. The timestamp is accurate to the millisecond level to ensure the timing of the transaction. Next, broadcast this transaction record to all nodes in the charging pile network, and the nodes verify the validity of the transaction through the consensus mechanism. After verification, add the transaction record to the block. When the block reaches the predetermined size or time interval, the node links the current block with the hash value of the previous block to form a chain structure and broadcasts the new block to the entire network to achieve the immutability of the data. Finally, the user can query the bill details in the blockchain browser through the transaction ID or bill ID to ensure the transparency and traceability of the transaction.

[0110] For example, Mr. Zhang, a user, uses charging pile CP20240417 to charge his electric vehicle. Mr. Zhang completes identity verification with the charging pile through the mobile phone NFC function, and the system identifies him as a gold member (discount rate 0.93). The benchmark electricity price on that day is 1.0 yuan per kilowatt-hour. The charging starts at 20:00 in the evening (flat peak period, coefficient 1.0) and continues until 2:00 in the early morning of the next day (valley period, coefficient 0.7). During the flat peak period (20:00 - 22:00), Mr. Zhang charges 8 kilowatt-hours. The system calculates the real-time rate as 1.0×1.0×0.93 = 0.93 yuan per kilowatt-hour, and the cumulative cost is 8×0.93 = 7.44 yuan. During the valley period (22:00 - 02:00), 20 kilowatt-hours are charged. The system calculates the real-time rate as 1.0×0.7×0.93 = 0.651 yuan per kilowatt-hour, and the cumulative cost is 20×0.651 = 13.02 yuan. After the charging ends, the system generates an electronic bill, which details the charging start and end times, the electricity consumption in each period, the rate in each period, the total electricity consumption of 28 kilowatt-hours, the total cost of 20.46 yuan, and the preferential amount of 1.54 yuan (compared with ordinary members). The system then calculates the SHA-256 hash value of the bill content, generating "3f7d8e2a1c9b6f5d4e2a1c9b8f7d6e5a4c3b2a1", and adds a timestamp "2024-04-18 02:00:15.382". These information are packaged into a transaction record and written into the blockchain. Mr. Zhang can view the bill details through the mobile application, and at the same time, he can also verify the authenticity and immutability of the bill through the blockchain browser.

[0111] The above describes the Internet of Things-based charging pile management method in the embodiments of the present application. Next, the Internet of Things-based charging pile management system in the embodiments of the present application will be described. Please refer to Figure 2 , an embodiment of the Internet of Things-based charging pile management system in the embodiments of the present application includes:

[0112] The acquisition module 201 is used to collect real-time data of the charging pile, and collect voltage parameters, current parameters, temperature parameters and power parameters through Internet of Things sensors to obtain the original data of the charging pile;

[0113] The authorization module 202 is used to perform identity recognition and authorization management on users according to the original data of the charging pile, verify user information through a multi-factor authentication mechanism and assign an authorization level to obtain user identity authorization data;

[0114] The detection module 203 is used to perform temperature anomaly detection on the internal cables of the charging pile based on the user identity authorization data and the original data of the charging pile, compare the set temperature threshold with the measured temperature and calculate the temperature rise rate to obtain a charging safety warning signal;

[0115] An analysis module 204, configured to perform feature extraction and clustering analysis on user charging behaviors based on the charging safety warning signal and the original data of the charging pile, calculate user behavior feature vectors through an improved K-Means clustering algorithm, and obtain user behavior clustering results;

[0116] A scheduling module 205, configured to predict and schedule the charging load based on the user behavior clustering results and the original data of the charging pile, calculate a load prediction curve through an LSTM neural network model and optimize the charging scheduling strategy, and obtain a charging scheduling plan;

[0117] A calculation module 206, configured to perform multi-period dynamic calculation of electricity prices based on the charging scheduling plan and the user identity authorization data, calculate real-time charging rates by multiplying the benchmark electricity price by the period coefficient and perform cost accumulation processing, and obtain a user charging bill.

[0118] Through the collaborative cooperation of the above-mentioned various components, real-time data collection of the charging pile is carried out through Internet of Things sensors, realizing accurate monitoring of key parameters such as voltage, current, temperature, and power, laying a data foundation for safe and efficient charging management. The multiple identity authentication mechanism combined with the authorization level allocation not only improves the system security, but also realizes a differentiated service strategy, enhancing the user experience while optimizing the resource allocation efficiency. Based on the user identity authorization data, personalized temperature threshold setting and real-time temperature rise rate calculation are carried out to construct an accurate charging safety warning mechanism, greatly reducing the safety risks during the charging process and ensuring the safety of charging equipment and user property. The improved K-Means clustering algorithm extracts and classifies the features of user charging behaviors, revealing the charging patterns and preferences of different user groups. These refined user portraits provide a decision-making basis for subsequent optimization of charging resource allocation. The LSTM neural network model makes full use of the time series characteristics and multi-dimensional features to achieve high-precision charging load prediction. Its recursive structure can effectively capture long-term dependence relationships. Compared with traditional time series prediction methods, the prediction accuracy is improved by more than 15%, providing a reliable prediction basis for intelligent scheduling strategies. The multi-period dynamic electricity price calculation mechanism combines the user level with the period coefficient to form a flexible price incentive system, effectively guiding users to charge during off-peak hours, reducing the peak load of the power grid, and improving the overall operating efficiency of the charging network. Overall, this solution deeply integrates Internet of Things technology with artificial intelligence algorithms. Especially the application of the K-Means clustering algorithm in user behavior analysis and the advantages of the LSTM neural network in load prediction jointly construct an adaptive, efficient and safe charging pile management system. This integration not only improves the intelligent level of electric vehicle charging infrastructure, but also optimizes the utilization of power grid resources by cutting peaks and filling valleys.

[0119] Above Figure 2The embodiments of the present invention are described in detail from the perspective of modular functional entities. Next, the embodiments of the present invention related to an Internet of Things-based charging pile management device are described in detail from the perspective of hardware processing.

[0120] Figure 3 FIG. 4 is a schematic structural diagram of an Internet of Things-based charging pile management device provided by an embodiment of the present invention. The Internet of Things-based charging pile management device 300 may vary greatly due to different configurations or performances, and may include one or more processors (central processing units, CPUs) 310 (for example, one or more processors) and a memory 320, and one or more storage media 330 (for example, one or more mass storage device ends) for storing application programs 333 or data 332. Among them, the memory 320 and the storage media 330 may be transient storage or persistent storage. The program stored in the storage media 330 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations on the Internet of Things-based charging pile management device 300. Further, the processor 310 may be configured to communicate with the storage media 330 and execute a series of instruction operations in the storage media 330 on the Internet of Things-based charging pile management device 300 to implement the steps of the above-mentioned Internet of Things-based charging pile management method.

[0121] The Internet of Things-based charging pile management device 300 may further include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input / output interfaces 360, and / or one or more operating systems 331, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, and so on. Those skilled in the art can understand that Figure 3 the shown structure of the Internet of Things-based charging pile management device does not limit the Internet of Things-based charging pile management device provided by the present invention, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0122] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions run on a computer, the computer is caused to execute the steps of the Internet of Things-based charging pile management method.

[0123] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, systems, and units can refer to the corresponding processes in the foregoing method embodiments and will not be repeated herein.

[0124] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing an Internet of Things-based charging pile management device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0125] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A charging pile management method based on the Internet of Things, characterized in that: The method comprises: Real-time data collection of charging piles is carried out. Voltage parameters, current parameters, temperature parameters and power parameters are collected through IoT sensors to obtain the original data of charging piles. Performing identity identification and authorization management on the user according to the original data of the charging pile, verifying the user information and assigning the authorization level through a multiple identity authentication mechanism, and obtaining the user identity authorization data; Based on the user identity authorization data and the charging pile original data, the temperature abnormality detection is performed on the cables inside the charging pile, and the set temperature threshold is compared with the measured temperature and the temperature rise rate is calculated to obtain a charging safety warning signal; Performing feature extraction and cluster analysis on the user's charging behavior based on the charging safety warning signal and the original data of the charging pile, calculating the user behavior feature vector by using an improved K-Means clustering algorithm, and obtaining a user behavior clustering result; Based on the user behavior clustering results and the original data of the charging piles, the charging load is predicted and scheduled, the load prediction curve is calculated through the LSTM neural network model, and the charging scheduling strategy is optimized to obtain a charging scheduling plan; The electricity price is dynamically calculated in multiple time periods according to the charging scheduling plan and the user identity authorization data, the real-time charging rate is calculated by multiplying the benchmark electricity price by the time period coefficient, and the fee accumulation processing is performed to obtain the user's charging bill.

2. The charging pile management method based on the Internet of Things according to claim 1 is characterized in that: The real-time data collection of the charging pile is performed by collecting voltage parameters, current parameters, temperature parameters and power parameters through the Internet of Things sensor to obtain the original data of the charging pile, including: The input voltage and output voltage of the charging pile are collected through the voltage sensor, and the collected voltage data is sampled and stored every 5 seconds to obtain the voltage parameters; The input current and output current of the charging pile are collected by the current sensor, and the collected current data is sampled and stored every 5 seconds to obtain the current parameters; The temperature sensor is used to collect the internal temperature of the charging pile and the temperature of the charging interface. The collected temperature data is sampled and stored every 30 seconds to obtain temperature parameters. The real-time charging power of the charging pile is collected through the power sensor, and the collected power data is sampled and stored every 5 seconds to obtain the power parameters; Perform AES-256 encryption processing on the voltage parameter, the current parameter, the temperature parameter, and the power parameter, and add a unique identification code and a timestamp of the charging pile to obtain an encrypted data packet; The encrypted data packet is transmitted to the central server through the Internet of Things communication module, and the encrypted data packet is decrypted, verified and indexed and stored to obtain the original data of the charging pile.

3. The charging pile management method based on the Internet of Things according to claim 1 is characterized in that: The user is identified and authorized according to the original data of the charging pile, and the user information is verified and the authorization level is assigned through a multiple identity authentication mechanism to obtain the user identity authorization data, including: Acquire user information based on the original data of the charging pile, receive user real-name registration information through a mobile application and generate a unique user identification code to obtain basic user information; Perform multiple identity authentication on the user's basic information, verify the user's identity through at least one of NFC near-field communication technology, QR code scanning or RFID radio frequency identification, and obtain an identity authentication result; The identity authentication result is processed by RSA asymmetric encryption, and the user identity information is protected and transmitted through an encryption algorithm to obtain encrypted identity data; Calculate the user's credit rating based on the encrypted identity data and historical charging records, and obtain the user's credit score by weighted summing the three indicators of cumulative charging times, average charging amount and payment timeliness; The user is divided into authorization levels according to the user credit score, and the user authorization level is obtained by mapping the credit score to four level intervals: diamond member, gold member, silver member and ordinary member; Based on the user authorization level, the charging power upper limit and charging time are allocated to the user, and the personalized charging parameters are calculated through a differentiated resource allocation algorithm for users of different levels to obtain the user identity authorization data.

4. The charging pile management method based on the Internet of Things according to claim 1 is characterized in that: The method of detecting abnormal temperature of the cables inside the charging pile based on the user identity authorization data and the original data of the charging pile, comparing the set temperature threshold with the measured temperature and calculating the temperature rise rate to obtain a charging safety warning signal includes: Extract temperature parameters from the raw data of the charging pile, construct a temperature time series by selecting the temperature data of the cables inside the charging pile within the last 30 minutes, and obtain the cable temperature history data; The temperature threshold is personalized according to the user identity authorization data, and the personalized temperature threshold is obtained by multiplying the basic temperature threshold by the safety factor corresponding to the user authorization level; The cable temperature history data is divided into time periods, and the temperature change data points are obtained by calculating the average temperature value and recording the data points every 5 minutes; Calculating the temperature rise rate according to the temperature change data points, and obtaining the temperature rise rate value for each time period by dividing the temperature difference between adjacent time points by the time interval; The temperature rise rate value is compared with a preset rate standard, and the safety level is divided by setting three intervals of normal rate, warning rate and dangerous rate to obtain the temperature change level; The safety risk value is calculated based on the temperature change level and the difference between the current measured temperature and the personalized temperature threshold, and the cable safety status is comprehensively evaluated through weighted summation to obtain the charging safety warning signal.

5. The charging pile management method based on the Internet of Things according to claim 1 is characterized in that: The method of extracting features and performing cluster analysis on the user's charging behavior based on the charging safety warning signal and the original data of the charging pile, calculating the user behavior feature vector by improving the K-Means clustering algorithm, and obtaining the user behavior clustering result includes: Extracting the user charging history records from the charging pile raw data, and obtaining the user charging raw characteristics by screening the charging time, charging duration, charging power and charging frequency in the past 90 days; The original charging characteristics of the user are processed in the time dimension, and the user's time preference characteristics are obtained by counting the charging distribution frequency of each user within seven days from Monday to Sunday and within 24 hours every day; Performing a safety rating on the user's charging behavior according to the charging safety warning signal, and obtaining the user's charging safety index by calculating the ratio of the frequency of triggering the safety warning during the user's historical charging process to the total number of charging times; Normalizing the user time preference feature and the user charging safety index, and obtaining a standardized feature vector by mapping each dimension data to a range of 0-1 to eliminate dimensional differences; Based on the standardized feature vector, the K value is adaptively determined, and the optimal number of clusters is obtained by calculating the silhouette coefficients under different K values ​​and selecting the K value corresponding to the maximum silhouette coefficient; The standardized feature vectors are grouped and calculated according to the optimal clustering number, and the user behavior clustering result is obtained by initially selecting samples far away from the center point as cluster centers and iteratively updating until the cluster centers are stable.

6. The charging pile management method based on the Internet of Things according to claim 1 is characterized in that: The charging load is predicted and scheduled based on the user behavior clustering result and the original data of the charging pile, and the load prediction curve is calculated and the charging scheduling strategy is optimized through the LSTM neural network model to obtain a charging scheduling plan, including: A time series data set is constructed based on the user behavior clustering results and the charging pile raw data, and the charging load time series data is obtained by arranging the historical load data of nearly 30 days at one data point per hour and adding time tags; Performing feature expansion processing on the charging load time series data, by adding time features, weather features and the user behavior clustering ratio features as input variables, to obtain a multi-dimensional input feature matrix; The multi-dimensional input feature matrix is ​​segmented into sliding windows, and data is organized by setting 24 hours as the input window length and the next 24 hours as the prediction target length to obtain input and output sample pairs; An LSTM network structure is constructed for the input and output sample pairs, and a load prediction model is obtained by processing time series data by setting a network structure of an input layer, two LSTM hidden layers, a Dropout layer, and a fully connected output layer; The charging load for the next 24 hours is calculated based on the load prediction model, and the load prediction curve is obtained by inputting the load data of the current 24 hours into the model and gradually rolling predicting the load values ​​of subsequent time periods; The charging piles are resource scheduled according to the load prediction curve, and the charging scheduling plan is obtained by dividing the prediction period into peak period, off-peak period and valley period and calculating the scheduling priority weight of each charging pile.

7. The charging pile management method based on the Internet of Things according to claim 1 is characterized in that: The method of dynamically calculating the electricity price in multiple time periods according to the charging scheduling scheme and the user identity authorization data, calculating the real-time charging rate by multiplying the benchmark electricity price by the time period coefficient and performing fee accumulation processing to obtain the user's charging bill includes: The charging time period is divided according to the charging scheduling scheme, and the time period electricity price coefficient is obtained by dividing 24 hours into three time periods: peak period, off-peak period and off-peak period and assigning time period coefficients of 1.5, 1.0 and 0.7 respectively; Performing level discount calculation on the user identity authorization data, obtaining a user discount coefficient by setting discount rates of 0.9, 0.93, 0.95 and 1.0 for diamond members, gold members, silver members and ordinary members respectively; The real-time charging rate is calculated based on the electricity price coefficient of the time period and the current market benchmark electricity price, and the real-time user rate is obtained by multiplying the product of the benchmark electricity price and the corresponding time period coefficient by the user discount coefficient; The single charging fee is accumulated according to the user's real-time rate and the amount of electricity charged each time, and the accumulated fee is calculated and recorded once for each completed charging amount of 1 kWh to obtain the real-time charging fee; Generate an electronic bill for the real-time charging fee, and obtain a detailed charging bill by recording the charging start and end time, power consumption in each time period, time period rate, total power consumption, total fee and discount amount; The detailed charging bill is stored in the transaction data through blockchain technology, and an unalterable transaction record is formed by calculating the bill data hash value and adding a timestamp to obtain the user's charging bill.

8. A charging pile management system based on the Internet of Things, characterized in that: Used to implement the charging pile management method based on the Internet of Things as described in any one of claims 1 to 7, the charging pile management system based on the Internet of Things includes: The acquisition module is used to collect real-time data of the charging pile, collect voltage parameters, current parameters, temperature parameters and power parameters through the IoT sensor, and obtain the original data of the charging pile; An authorization module is used to identify and authorize the user according to the original data of the charging pile, verify the user information and assign the authorization level through a multiple identity authentication mechanism, and obtain the user identity authorization data; A detection module, used to detect temperature anomalies of cables inside the charging pile based on the user identity authorization data and the original data of the charging pile, compare the set temperature threshold with the measured temperature and calculate the temperature rise rate to obtain a charging safety warning signal; An analysis module is used to extract features and perform cluster analysis on the user's charging behavior based on the charging safety warning signal and the charging pile raw data, calculate the user behavior feature vector by improving the K-Means clustering algorithm, and obtain the user behavior clustering result; A scheduling module, used to predict and schedule the charging load based on the user behavior clustering results and the original data of the charging piles, calculate the load prediction curve through the LSTM neural network model and optimize the charging scheduling strategy to obtain a charging scheduling plan; The calculation module is used to dynamically calculate the electricity price in multiple time periods according to the charging scheduling plan and the user identity authorization data, calculate the real-time charging rate by multiplying the benchmark electricity price by the time period coefficient, and perform fee accumulation processing to obtain the user's charging bill.

9. A charging pile management device based on the Internet of Things, characterized in that: It includes a memory and a processor, the memory stores a computer program that can be run on the processor, and the processor implements the charging pile management method based on the Internet of Things as described in any one of claims 1 to 7 when executing the computer program.

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 processor executes the charging pile management method based on the Internet of Things as described in any one of claims 1 to 7.

Images