Intelligent charging station monitoring method and system based on digital twinning

Abstract

The invention relates to the technical field of charging stations, and discloses an intelligent charging station monitoring method based on digital twinning, which comprises the steps of 1, data acquisition and preprocessing, 2, digital twinning model construction and calibration, 3, real-time monitoring and intelligent decision making, and 4, full life cycle management. The invention also provides an intelligent charging station monitoring system based on digital twinning, which comprises a physical module, a virtual module, a data module, an application module, a connection module and a visualization module. The system has the beneficial effects that multi-source data are collected in real time through the Internet of Things, a digital twin model is constructed, a perfect 3D visual interface is provided, and information such as the equipment running state, the charging process and the energy flow of the charging station can be accurately displayed in real time. And meanwhile, remote monitoring and intelligent control of charging station equipment, such as functions of remotely adjusting the power of the charging pile, switching equipment and the like, are realized, and convenient and efficient operation experience is provided for a user.

Description

Technical Field

[0001] The present invention relates to the technical field of charging stations, and in particular to an intelligent charging station monitoring method based on digital twin, and also to an intelligent charging station monitoring system based on digital twin. Background Art

[0002] With the rapid popularization of new energy vehicles, the scale of charging stations has shown an explosive growth. Data shows that from 2020 to 2024, the sales volume of new energy vehicles increased from 1.367 million to 12.866 million, and the cumulative number of charging piles increased from about 1.3 million to over 10 million (including public piles and private piles) during the same period. The traditional charging station management model has exposed four core problems in large-scale operation: Low monitoring efficiency: Relying on manual inspections, it is impossible to grasp the status of charging piles in real time (such as voltage, current fluctuations, fault warnings), resulting in a lag in the response to equipment abnormalities.

[0003] Unbalanced energy scheduling: The charging load fluctuates greatly, which does not match the grid load, easily causes overload or energy waste, and it is difficult to integrate distributed energy such as photovoltaic and energy storage.

[0004] Insufficient user experience: Users face problems such as "difficulty in finding charging piles" and "long queuing times", and the traditional two-dimensional interface cannot intuitively display the status of charging pile positions and the charging process.

[0005] High operation and maintenance costs: The cost of manual inspections accounts for more than 40%, fault troubleshooting relies on experience, the equipment downtime is long, and the life cycle management lacks data support. Summary of the Invention

[0006] In view of the above problems existing in the existing intelligent charging station monitoring method based on digital twin, we have proposed the present invention.

[0007] Therefore, the purpose of the present invention is to achieve real-time bidirectional mapping between the physical charging station and the virtual model through digital twin, 3D modeling, and data interaction visualization. The system covers core functions such as 3D modeling, data interaction visualization, and energy optimization, accurately monitors the status of the charging station, intelligently predicts equipment failures, optimizes energy scheduling, improves management efficiency and user charging experience, and helps the intelligent upgrade and sustainable development of the charging station.

[0008] For this purpose, an intelligent charging station monitoring method based on digital twin is provided, including the following steps: Step 1: Data collection and preprocessing. Multisource heterogeneous data, including device operation parameters, environmental parameters, and user behavior data, are collected in real time through the Internet of Things sensor network deployed in the charging station. The sliding window filtering algorithm is used for data denoising. The multi-dimensional data is mapped by combining the Z-score normalization method. The synchronous processing of data with different sampling frequencies is achieved through the timestamp alignment technology, and a distributed data storage architecture is constructed. Step 2: Digital twin model construction and calibration. Based on the theory of cyber-physical fusion, the three-dimensional geometric model of the charging station is constructed using BIM (Building Information Modeling) technology. The physical model of the device is established by combining multi-physics field simulation technology. The extended Kalman filter (EKF) algorithm is used to achieve the dynamic calibration of the digital twin model. The steps are as follows. Prediction: According to the previous state xk−1 and control input uk, calculate the prior estimate ^kk|k−1 and covariance Pk|k-1. Update: Use the current observation value zk (such as the measured temperature) to calculate the Kalman gain Kk, and update the posterior estimate ^xk|k and covariance Pk|k-1. The calibration period is 1 second to ensure that the state error between the virtual model and the physical entity is ≤ 3%. A twin data consistency evaluation mechanism is introduced. By calculating the state residuals between the physical entity and the virtual model, when the residuals exceed the preset threshold (such as 3σ), the model recalibration process is triggered. Step 3: Real-time monitoring and intelligent decision-making. A three-dimensional visual monitoring platform is developed based on WebGL technology. The hierarchical data display strategy (macro-region level → meso-station level → micro-device level) is adopted to achieve multi-scale monitoring. A multi-dimensional early warning index system is constructed, including device status early warning, energy fluctuation early warning, and safety risk early warning, etc. An intelligent decision-making engine based on multi-objective optimization is developed, and different optimization strategies are adopted for different application scenarios. Step 4: Full life cycle management, including device life prediction. Based on the Weibull distribution model, the calculation formula is: F(t) = 1 - e^(-(t / η)^β), where η = 50000 hours (characteristic life), β = 1.2 (shape parameter, representing the failure rate trend), and the error in predicting the mean time between failures (MTBF) of the charging pile is ≤ 5%. Maintenance plan: When the remaining life of the device < 30 days, a maintenance work order is automatically generated, including: list of replacement parts, estimated maintenance duration, and query results of spare parts inventory. User behavior analysis: Through association rule mining (such as the Apriori algorithm), analyze the association relationship between the user's charging time period, vehicle type, and battery level. For example: "When the battery level of SUV models < 20%, 80% choose fast charging piles", and optimize the charging pile layout accordingly. Also provided is an intelligent charging station monitoring system based on digital twins, including a physical module, a virtual module, a data module, and an application module, and also including a connection module and a visualization module.

[0009] Specifically, the physical module includes a voltage / current sensor (accuracy ±0.5%), a temperature / humidity sensor (accuracy ±2%RH / ±0.5°C), an RFID reader (identification distance 0.1 - 10m), covering physical entities such as charging piles (AC / DC piles), energy storage devices, and environmental monitoring points. It integrates a SCADA system and a 5G communication module to achieve a data acquisition frequency ≥10Hz and a wireless transmission delay ≤100ms. The sensor network of the physical module adopts a hierarchical deployment architecture. The device layer deploys current / voltage sensors inside the charging pile to collect core operation parameters; the environmental layer deploys meteorological sensors (including wind speed and precipitation detection) on the top of the station building to monitor the impact of extreme weather on charging safety; the user layer deploys RFID readers at the charging positions to automatically identify user identities and vehicle charging requirements, with an identification accuracy ≥99%.

[0010] Specifically, the virtual module includes a geometric model built based on 3D modeling software, a thermal - electrical coupling physical model of the charging device established using simulation software, a behavior model trained based on data such as user charging duration and arrival time, and built - in charging power adjustment rules (such as automatically adjusting power during peak - valley electricity price periods) and fault response rules (such as automatically powering off when the temperature exceeds the threshold), supporting custom policy configuration. The geometric model includes 1:1 high - precision 3D models of charging piles (including AC piles and DC fast - charging piles), station buildings, and roads, and the model detail accuracy reaches 0.1mm. The physical model includes simulating the dynamic distribution of the current - temperature field, with an error ≤5%.

[0011] Specifically, the data module includes a collection layer, a processing layer, and a storage layer. The collection layer captures data such as charging power, voltage fluctuations, device status (charging / fault / standby), user ID, and charging amount in real time. The processing layer uses moving average filtering for denoising and combines Min - Max normalization to process multi - source data. The time - series database (InfluxDB) in the storage layer is used to store real - time data (retaining for 30 days), MySQL stores structured data such as user profiles and device ledgers, and MongoDB stores unstructured data such as video monitoring.

[0012] Specifically, the application module includes a fault prediction algorithm, an energy scheduling optimization algorithm, and a user queuing prediction algorithm. The fault prediction algorithm uses an LSTM neural network model. The input layer contains 12-dimensional features (such as voltage, current, temperature, charging duration, etc.), there are 2 hidden layers (each with 128 neurons), and the output layer is the device health status value (0 - 1, the closer to 0, the higher the fault risk). The energy scheduling optimization algorithm is based on a mixed integer programming (MIP) model, and the objective function is min α.Ccost + β.Cgrid, where Ccost is the charging cost (yuan), Cgrid is the grid load fluctuation penalty term, and α + β = 1 (the weights can be dynamically adjusted). The constraint conditions include the upper and lower limits of the charging pile power (such as 30 - 120kW for DC piles), the grid capacity limit (such as the maximum power supply of the substation area is 500kW), and the optimization period is 15 minutes. The user queuing prediction algorithm is based on the M / M / c model, and the calculation formula is: Pwait = . P0 (ρ = λ / (cμ)) where λ is the user arrival rate (vehicles / hour), μ is the charging pile service rate (vehicles / hour, μ = 4 for DC piles, μ = 1 for AC piles), c is the number of charging piles, and Pwait is the queuing probability. When the predicted queuing time > 30 minutes, the system automatically triggers the "fast charging pile priority scheduling" strategy.

[0013] Specifically, the connection module uses the MQTT protocol to achieve bidirectional data transmission, supports TLS encryption (AES - 256 algorithm), the communication delay ≤ 50ms, and the packet loss rate ≤ 0.1%. An edge computing node (EdgeGateway) is built to achieve local data pre - processing (such as outlier filtering).

[0014] Specifically, the visualization module includes a three - dimensional monitoring interface developed based on Unity3D, supports multi - terminal access on PC / mobile / VR, and finally realizes the visualization of device status: the charging piles are rendered in real - time in three colors: red (fault) / yellow (standby) / green (charging), and clicking on them can view real - time parameters such as power and temperature; the visualization of energy flow: the current direction is simulated through particle effects, and the energy distribution path of photovoltaic - energy storage - charging pile is dynamically displayed; and a data dashboard: integrating components such as charging volume statistics, fault trend analysis, and user behavior heat maps, and supports custom report export.

[0015] Specifically, the sensor network of the physical module adopts a hierarchical deployment architecture: Device layer: Current / voltage sensors deployed inside the charging pile, used to collect core operation parameters; Environment layer: Meteorological sensors (including wind speed and precipitation detection) deployed on the top of the station building, used to monitor the impact of extreme weather on charging safety; User layer: RFID card readers deployed at the charging spaces, used to automatically identify user identities and vehicle charging requirements, with an identification accuracy rate ≥ 99%.

[0016] Specifically, the virtual module adopts multi-precision modeling technology, which includes a macroscopic layer: representing the overall layout of the charging station with simplified geometric bodies (such as cubes), used for energy flow simulation, with a rendering frame rate ≥ 60 FPS; Microscopic layer: Conducting refined modeling on the internal circuits and heat dissipation structures of the charging pile (mesh size ≤ 1 mm), used for thermal failure analysis, with a simulation error ≤ 2%; Supporting dynamic switching of model precision. When the user zooms the interface to the device level, the microscopic layer model is automatically loaded.

[0017] Specifically, the connection module integrates the blockchain evidence storage function: Conducting blockchain evidence storage on key data (such as charging transaction records and device fault logs), using the SHA-256 hash algorithm to ensure data immutability; Designing an attribute-based encryption (ABE) scheme, where user privacy data (such as charging amount and location information) is only open to authorized personnel, with access control granularity reaching the field level; Deploying an intrusion detection system (IDS), identifying network attacks through the LSTM anomaly detection model, with a false alarm rate < 0.5% and a response time < 5 seconds.

[0018] The beneficial effects of the present invention: It has a complete 3D visualization interface, which can display information such as the operation status of charging station equipment, charging process, and energy flow in real time and accurately. At the same time, it realizes remote monitoring and intelligent control of charging station equipment, such as functions like remotely adjusting the power of the charging pile and switching devices, providing users with a convenient and efficient operation experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts. Among them: Figure 1 It is a schematic diagram of the overall system framework structure of the present invention.

[0020] Figure 2 It is a schematic diagram of the system operation process of the present invention.

[0021] Figure 3 It is a schematic diagram of the system data transmission design of the present invention.

[0022] Figure 4 This is a schematic diagram of a simple flowchart for creating a 3D model of the present invention. Detailed implementation manners

[0023] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe the detailed implementation manners of the present invention in conjunction with the drawings of the specification.

[0024] In the following description, many specific details are set forth to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0025] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that can be included in at least one implementation manner of the present invention. The "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that mutually excludes other embodiments.

[0026] Thirdly, the present invention is described in detail in conjunction with schematic diagrams. When describing the embodiments of the present invention in detail, for the convenience of explanation, the cross-sectional views showing the device structure will be enlarged locally not in accordance with the general scale, and the schematic diagrams are only examples and should not limit the protection scope of the present invention herein. In addition, in actual production, three-dimensional spatial dimensions including length, width, and depth should be included.

[0027] Embodiment Referring to Figures 1-4 , an embodiment of the present invention provides an intelligent charging station monitoring method based on digital twin, including the following steps: Step 1: Data collection and preprocessing. Multisource heterogeneous data, including device operation parameters, environmental parameters, and user behavior data, are collected in real time through the Internet of Things sensor network deployed in the charging station. The sliding window filtering algorithm is used for data denoising. The multi-dimensional data is mapped by combining the Z-score normalization method, and the synchronization processing of data with different sampling frequencies is realized through the timestamp alignment technology, and a distributed data storage architecture is constructed; Step 2: Digital twin model construction and calibration. Based on the theory of physical information fusion, a three-dimensional geometric model of the charging station is constructed using BIM (Building Information Modeling) technology, a device physical model is established by combining multi-physics field simulation technology, and the extended Kalman filter (EKF) algorithm is used to realize the dynamic calibration of the digital twin model. The steps are as follows. Prediction: According to the previous moment state xk−1 and the control input uk, calculate the prior estimate ^kk|k−1 and the covariance Pk|k-1. Update: Using the current observation value zk (such as the measured temperature), calculate the Kalman gain Kk, and update the posterior estimate ^xk|k and covariance Pk|k-1. The calibration period is 1 second to ensure that the state error between the virtual model and the physical entity is ≤ 3%. Introduce a twin data consistency evaluation mechanism. By calculating the state residuals between the physical entity and the virtual model, when the residuals exceed the preset threshold (such as 3σ), trigger the model recalibration process. Step 3: Real-time monitoring and intelligent decision-making. Develop a 3D visualization monitoring platform based on WebGL technology. Adopt a hierarchical data display strategy (macro-region level → meso-station level → micro-device level) to achieve multi-scale monitoring. Build a multi-dimensional early warning index system, including equipment status early warning, energy fluctuation early warning, and safety risk early warning, etc. Develop an intelligent decision-making engine based on multi-objective optimization, and adopt different optimization strategies for different application scenarios. Step 4: Full life cycle management, including equipment life prediction. Based on the Weibull distribution model, the calculation formula is: F(t) = 1 - e^(-(t / η)^β), where η = 50000 hours (characteristic life), β = 1.2 (shape parameter, representing the failure rate trend), and the error in predicting the mean time between failures (MTBF) of the charging pile is ≤ 5%. Maintenance plan: When the remaining life of the equipment < 30 days, automatically generate a maintenance work order, the content of which includes: a list of replacement parts, the estimated maintenance duration, and the query results of spare parts inventory. User behavior analysis: Through association rule mining (such as the Apriori algorithm), analyze the association relationship between the user's charging time period, vehicle type, and battery level. For example: "When the battery level of an SUV model < 20%, 80% choose fast charging piles", and optimize the charging pile layout accordingly. At the same time, it also provides an intelligent charging station monitoring system based on digital twins, including a physical module, a virtual module, a data module, and an application module, and also includes a connection module and a visualization module.

[0028] Specifically, the physical module includes voltage / current sensors (accuracy ±0.5%), temperature / humidity sensors (accuracy ±2%RH / ±0.5°C), RFID readers (identification distance 0.1 - 10m), covering physical entities such as charging piles (AC / DC piles), energy storage devices, and environmental monitoring points. Integrate the SCADA system and the 5G communication module to achieve a data acquisition frequency ≥ 10Hz and a wireless transmission delay ≤ 100ms. The sensor network of the physical module adopts a hierarchical deployment architecture. The device layer deploys current / voltage sensors inside the charging pile to collect core operation parameters; the environment layer deploys meteorological sensors (including wind speed and precipitation detection) on the top of the station building to monitor the impact of extreme weather on charging safety; the user layer deploys RFID readers at the charging parking spaces to automatically identify the user's identity and vehicle charging requirements, and the identification accuracy ≥ 99%.

[0029] Specifically, the virtual module includes a geometric model built based on 3D model software, a thermal-electrical coupling physical model of the charging device established using simulation software, a behavior model trained based on data such as user charging duration and arrival time, and built-in charging power adjustment rules (such as automatically adjusting power during peak-valley electricity price periods) and fault response rules (such as automatically powering off when the temperature exceeds the threshold), supporting custom policy configuration. The geometric model includes 1:1 high-precision 3D models of charging piles (including AC piles and DC fast-charging piles), station houses, and roads, and the model detail accuracy reaches 0.1 mm. The physical model includes simulating the dynamic distribution of current-temperature fields, with an error ≤ 5%.

[0030] Specifically, the data module includes an acquisition layer, a processing layer, and a storage layer. Among them, the acquisition layer captures data such as charging power, voltage fluctuations, device status (charging / fault / standby), user ID, and charging amount in real time. The processing layer uses moving average filtering for denoising and combines Min-Max normalization to process multi-source data. The time-series database (InfluxDB) in the storage layer is used to store real-time data (retaining for 30 days), MySQL stores structured data such as user profiles and device ledgers, and MongoDB stores unstructured data such as video monitoring.

[0031] Specifically, the application module includes a fault prediction algorithm, an energy scheduling optimization algorithm, and a user queuing prediction algorithm. The fault prediction algorithm uses an LSTM neural network model. The input layer includes 12-dimensional features (voltage, current, temperature, charging duration, etc.), there are 2 hidden layers (each with 128 neurons), and the output layer is the device health status value (0 - 1, the closer to 0 indicates a higher fault risk). The energy scheduling optimization algorithm is based on a mixed-integer programming (MIP) model, and the objective function is min α.Ccost + β.Cgrid, where Ccost is the charging cost (yuan), Cgrid is the grid load fluctuation penalty term, and α + β = 1 (the weights can be dynamically adjusted). The constraint conditions include the upper and lower limits of the charging pile power (such as 30 - 120 kW for DC piles), the grid capacity limit (such as the maximum power supply of the substation area is 500 kW), and the optimization period is 15 minutes. The user queuing prediction algorithm is based on the M / M / c model, and the calculation formula is: Pwait= . P0 (ρ = λ / (cμ)) Where λ is the user arrival rate (vehicles / hour), μ is the charging pile service rate (vehicles / hour, μ = 4 for DC piles, μ = 1 for AC piles), c is the number of charging piles, Pwait is the queuing probability. When the predicted queuing time > 30 minutes, the system automatically triggers the "priority scheduling of fast-charging piles" strategy.

[0032] Specifically, the connection module uses the MQTT protocol to achieve bidirectional data transmission, supports TLS encryption (AES-256 algorithm), has a communication latency ≤ 50 ms and a packet loss rate ≤ 0.1%, and constructs an edge computing node (EdgeGateway) to implement local data preprocessing (such as outlier filtering).

[0033] Specifically, the visualization module includes a 3D monitoring interface developed based on Unity3D, supports multi-terminal access on PC / mobile / VR, and finally realizes the visualization of device status: the charging pile is rendered in real time in three colors: red (fault) / yellow (standby) / green (charging), and clicking on it can view real-time parameters such as power and temperature; the visualization of energy flow: simulates the current direction through particle effects and dynamically displays the energy distribution path of photovoltaic - energy storage - charging pile; and a data dashboard: integrates components such as charging volume statistics, fault trend analysis, and user behavior heat maps, and supports the export of custom reports.

[0034] Specifically, the sensor network of the physical module adopts a hierarchical deployment architecture: device layer: current / voltage sensors deployed inside the charging pile, used to collect core operation parameters; environment layer: meteorological sensors (including wind speed and precipitation detection) deployed on the top of the station building, used to monitor the impact of extreme weather on charging safety; user layer: RFID card readers deployed at the charging positions, used to automatically identify user identities and vehicle charging requirements, with an identification accuracy ≥ 99%.

[0035] Specifically, the virtual module uses multi-precision modeling technology, which includes a macroscopic layer: represents the overall layout of the charging station with simplified geometric bodies (such as cubes), used for energy flow simulation, with a rendering frame rate ≥ 60 FPS; microscopic layer: conducts refined modeling on the internal circuit and heat dissipation structure of the charging pile (mesh size ≤ 1 mm), used for thermal failure analysis, with a simulation error ≤ 2%; supports dynamic switching of model accuracy, and when the user zooms the interface to the device level, the microscopic layer model is automatically loaded.

[0036] Specifically, the connection module integrates the blockchain evidence storage function: conducts blockchain evidence storage on key data (such as charging transaction records, device fault logs), uses the SHA-256 hash algorithm to ensure the immutability of data; designs an attribute-based encryption (ABE) scheme, and user privacy data (such as charging volume, location information) is only open to authorized personnel, with access control granularity reaching the field level; deploys an intrusion detection system (IDS), uses the LSTM anomaly detection model to identify network attacks, with a false alarm rate < 0.5% and a response time < 5 seconds.

[0037] In summary, multi-source data (sampling frequency ≥ 10Hz) is collected in real time through Internet of Things devices (voltage / current sensors, RFID readers, etc.), and after preprocessing such as moving average filtering and Min-Max normalization, a 1:1 digital twin model of the physical charging station is constructed. The virtual model is dynamically calibrated through the extended Kalman filter algorithm (calibration period 1 second, state error ≤ 3%), achieving real-time synchronization of charging pile status, environmental parameters and user behavior, with an 80% improvement in monitoring efficiency and a 60% reduction in manual inspection costs.

[0038] The visualization interface developed based on Unity3D or other available development engines supports multi-terminal access on PC / mobile / VR, and real-time renders the device status through red / yellow / green three-color coding, integrating energy flow particle effects and multi-dimensional data dashboards (such as charging volume statistics, fault trend analysis). Managers can intuitively grasp the station situation through operations such as zooming and rotating, with a 70% improvement in the efficiency of complex data interpretation.

[0039] The mixed integer programming (MIP) algorithm dynamically balances the charging load and the grid capacity, reducing the charging cost by 30% in the peak-valley electricity price scenario and improving the grid load balance by 40%; after accessing photovoltaic and energy storage data, the energy utilization rate is increased by 15%-20%, promoting the transformation of the charging station to low-carbon operation.

[0040] The LSTM neural network model is trained based on more than 100,000 fault data, and the prediction accuracy of the device health status reaches 92%. It can predict faults 72 hours in advance, shorten the operation and maintenance response time by 50%, and extend the device life cycle by 10%; the Weibull distribution model predicts the remaining life of the device (error ≤ 5%). Combined with the preventive maintenance work order system, the operation and maintenance cost is reduced by 35%.

[0041] The M / M / c queuing theory model predicts the user waiting time (error ≤ 8%). When the queue exceeds 30 minutes, the fast charging pile is automatically triggered for priority scheduling, and the average user waiting time is shortened by 30%; the mobile APP displays the pile status and queue prediction results in real time, and the charging satisfaction is increased to over 90%.

[0042] By analyzing user behavior through association rule mining (such as the Apriori algorithm), the charging pile layout and charging strategy are optimized, and the utilization rate of station resources is increased by 25%; blockchain evidence storage and ABE encryption technology ensure data security, and the false alarm rate of the intrusion detection system < 0.5%, ensuring operational compliance.

[0043] The system integrates technologies such as digital twin, Internet of Things, AI algorithms (LSTM / PPO / extended Kalman filter), and edge computing to build a "data collection-modeling-analysis-control" closed loop, realizing collaborative optimization from the device level to the system level.

[0044] Adopt a hierarchical modular architecture (physical layer, virtual layer, application layer, etc.), support flexible expansion of sensor protocols, algorithm models, and visualization interfaces, adapt to charging stations of different scales, provide replicable intelligent upgrade solutions for the industry, and help new models such as "photovoltaic energy storage charging and swapping integration" to be implemented.

[0045] It is important to note that the construction and arrangement of the present application shown in multiple different exemplary embodiments are merely illustrative. Although only a few embodiments are described in detail in this disclosure, those who refer to this disclosure should easily understand that many modifications are possible without substantially departing from the novel teachings and advantages of the subject matter described in this application (for example, the dimensions, scales, structures, shapes, and proportions of various components, as well as parameter values (such as temperature, pressure, etc.), installation arrangements, use of materials, color, orientation changes, etc.). For example, an element shown as integrally formed may be composed of multiple parts or elements, the position of the element may be inverted or otherwise changed, and the nature, number, or position of discrete elements may be changed or altered. Therefore, all such modifications are intended to be included within the scope of the present invention. The order or sequence of any process or method steps may be changed or reordered according to alternative embodiments. In the claims, any "means-plus-function" clause is intended to cover the structures that perform the recited function herein, and not only structural equivalents but also equivalent structures. Other substitutions, modifications, changes, and omissions may be made in the design, operating conditions, and arrangement of the exemplary embodiments without departing from the scope of the present invention. Therefore, the present invention is not limited to a specific embodiment, but extends to various modifications that still fall within the scope of the appended claims.

[0046] In addition, to provide a concise description of the exemplary embodiments, not all features of the actual embodiments may be described (i.e., those features that are not relevant to the currently considered best mode of implementing the present invention or those features that are not relevant to the implementation of the present invention).

[0047] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. An intelligent charging station monitoring method based on digital twin, characterized in that: It includes the following steps: Step 1: Data collection and preprocessing. Multisource heterogeneous data, including device operation parameters, environmental parameters, and user behavior data, are collected in real time through the Internet of Things sensor network deployed in the charging station. The sliding window filtering algorithm is used for data denoising. The multi-dimensional data is mapped by combining the Z-score normalization method. The synchronization processing of data with different sampling frequencies is realized through the timestamp alignment technology, and a distributed data storage architecture is constructed. Step 2: Digital twin model construction and calibration. Based on the theory of cyber-physical fusion, the three-dimensional geometric model of the charging station is constructed using building information modeling technology. The physical model of the device is established by combining multi-physics field simulation technology. The extended Kalman filter algorithm is used to realize the dynamic calibration of the digital twin model. The steps are as follows. Prediction: Based on the state x at the previous moment k−1 and the control input u k , calculate the prior estimate ^k k|k−1 and the covariance P k|k-1 , Update: Calculate the Kalman gain K k using the current observation z k (measured temperature), and update the posterior estimate ^x k|k and covariance P k|k-1 , The calibration period is 1 second to ensure that the state error between the virtual model and the physical entity is ≤ 3%. A twin data consistency evaluation mechanism is introduced. By calculating the state residuals between the physical entity and the virtual model, when the residuals exceed the preset threshold, the model recalibration process is triggered. Step 3: Real-time monitoring and intelligent decision-making. A three-dimensional visualization monitoring platform is developed based on WebGL technology. The hierarchical data display strategy is adopted to achieve multi-scale monitoring. A multi-dimensional early warning index system is constructed, including device status early warning, energy fluctuation early warning, and safety risk early warning, etc. An intelligent decision-making engine based on multi-objective optimization is developed, and different optimization strategies are adopted for different application scenarios. Step 4: Full life cycle management, including equipment life prediction, based on the Weibull distribution model, and the calculation formula is: F(t) = 1 - e -（t / η）β where η = 50,000 hours (characteristic life), β = 1.2 (shape parameter, representing the failure rate trend), and the error in predicting the average time between failures of the charging pile is ≤ 5%; Maintenance plan: When the remaining life of the device < 30 days, a maintenance work order is automatically generated, and the content includes: a list of replacement parts, the estimated maintenance duration, and the query results of spare parts inventory. User behavior analysis: By mining association rules, the association relationship between the user's charging time period, vehicle type, and battery level is analyzed, and the charging pile layout is optimized accordingly.

2. A system for the method of monitoring an intelligent charging station based on digital twin as claimed in claim 1, characterized in that: The system includes a physical module, a virtual module, a data module, and an application module. It also includes a connection module and a visualization module. The physical module includes voltage / current sensors (accuracy ±0.5%), temperature / humidity sensors (accuracy ±2%RH / ±0.5°C), and RFID readers (identification distance 0.1 - 10m), covering physical entities such as charging piles (AC / DC piles), energy storage devices, and environmental monitoring points. The SCADA system and the 5G communication module are integrated to achieve a data collection frequency ≥ 10Hz and a wireless transmission delay ≤ 100ms. The sensor network of the physical module adopts a hierarchical deployment architecture. The device layer deploys current / voltage sensors inside the charging pile to collect core operation parameters; the environmental layer deploys meteorological sensors (including wind speed and precipitation detection) on the top of the station building to monitor the impact of extreme weather on charging safety. The user layer deploys RFID readers at the charging parking spaces to automatically identify the user's identity and the vehicle's charging requirements, and the identification accuracy is ≥ 99%.

3. The intelligent charging station monitoring system based on digital twin according to claim 2, characterized in that: The virtual module includes a geometric model built based on 3D modeling software, a thermal-electrical coupling physical model of the charging device established using simulation software, a behavior model trained based on data such as user charging duration and arrival time, and built-in charging power adjustment rules and fault response rules that support custom policy configuration. The geometric model includes 1:1 high-precision 3D models of charging piles (including AC piles and DC fast-charging piles), station houses, and roads, and the model detail accuracy reaches 0.1 mm. The physical model includes simulating the dynamic distribution of current-temperature fields with an error ≤ 5%.

4. The intelligent charging station monitoring system based on digital twin according to claim 3, characterized in that: The data module includes an acquisition layer, a processing layer, and a storage layer. The acquisition layer captures data such as charging power, voltage fluctuations, device status (charging / fault / standby), user ID, and charging amount in real time. The processing layer uses moving average filtering for denoising and combines Min-Max normalization to process multi-source data. The time-series database (InfluxDB) in the storage layer is used to store real-time data (retaining for 30 days), MySQL stores structured data such as user profiles and device ledgers, and MongoDB stores unstructured data such as video monitoring.

5. The intelligent charging station monitoring system based on digital twin according to claim 4, characterized in that: The application module includes a fault prediction algorithm, an energy scheduling optimization algorithm, and a user queuing prediction algorithm. The fault prediction algorithm uses an LSTM neural network model. The input layer contains 12-dimensional features (such as voltage, current, temperature, charging duration, etc.), there are 2 hidden layers (each with 128 neurons), and the output layer is the device health status value (0 - 1, the closer to 0 indicates a higher fault risk). The energy scheduling optimization algorithm is based on a mixed integer programming (MIP) model, and the objective function is min α . C cost +β . C grid , where C cost is the charging cost (yuan), C grid is the penalty term for grid load fluctuations, α + β = 1 (the weights can be dynamically adjusted). The constraint conditions include the upper and lower limits of the charging pile power and the grid capacity limit. The optimization period is 15 minutes. The user queuing prediction algorithm is based on the M / M / c model, and the calculation formula is: P wait = . P0(ρ = λ / (cμ)) Among them, λ is the user arrival rate (vehicles per hour), μ is the charging pile service rate (vehicles per hour, μ = 4 for DC piles and μ = 1 for AC piles), c is the number of charging piles, and Pwait is the queuing probability. When the predicted queuing time > 30 minutes, the system automatically triggers the "fast-charging pile priority scheduling" strategy.

6. The intelligent charging station monitoring system based on digital twin according to claim 5, characterized in that: The connection module uses the MQTT protocol to achieve bidirectional data transmission, supports TLS encryption (AES-256 algorithm), with a communication delay ≤ 50 ms and a packet loss rate ≤ 0.1%. It constructs an edge computing node (EdgeGateway) to achieve local data preprocessing.

7. The intelligent charging station monitoring system based on digital twin according to claim 6, characterized in that: The visualization module includes a 3D monitoring interface developed based on Unity3D, supporting multi-terminal access on PC / mobile / VR. Finally, it realizes device status visualization: charging piles are rendered in real time in three colors: red (fault) / yellow (standby) / green (charging), and clicking can view parameters such as real-time power and temperature; energy flow visualization: simulating the current direction through particle effects and dynamically displaying the energy distribution path of photovoltaic-storage-charging piles; and a data dashboard: integrating components such as charging amount statistics, fault trend analysis, and user behavior heat maps, and supporting custom report export.

8. The intelligent charging station monitoring system based on digital twin according to claim 7, wherein: The sensor network of the physical module adopts a hierarchical deployment architecture: Device layer: Current / voltage sensors deployed inside the charging pile are used to collect core operation parameters; Environment layer: Meteorological sensors (including wind speed and precipitation detection) deployed on the top of the station house are used to monitor the impact of extreme weather on charging safety; User layer: RFID card readers deployed at the charging parking spaces are used to automatically identify user identities and vehicle charging requirements, with an identification accuracy ≥ 99%.

9. The intelligent charging station monitoring system based on digital twin according to claim 8, characterized in that: The virtual module adopts multi-precision modeling technology, which includes a macroscopic layer: representing the overall layout of the charging station with a simplified geometric body for energy flow simulation, and the rendering frame rate is ≥60FPS; a microscopic layer: performing refined modeling on the internal circuit and heat dissipation structure of the charging pile (the grid size is ≤1mm) for thermal failure analysis, and the simulation error is ≤2%; it supports dynamic switching of model accuracy. When the user zooms the interface to the device level, the microscopic layer model is automatically loaded.

10. The intelligent charging station monitoring system based on digital twin according to claim 8, characterized in that: The connection module integrates the blockchain evidence storage function: performing blockchain evidence storage on key data, using the SHA-256 hash algorithm to ensure the immutability of data; designing an attribute-based encryption (ABE) scheme, where user privacy data is only open to authorized personnel, and the access control granularity reaches the field level; deploying an intrusion detection system (IDS), identifying network attacks through the LSTM anomaly detection model, with a false alarm rate <0.5% and a response time <5 seconds.

Images