A charging pile control system based on digital twin

Abstract

This invention belongs to the field of digital twin technology and discloses a charging pile control system based on digital twins. It includes a data acquisition module for collecting charging pile status data, vehicle information data, environmental data, and safety monitoring data; a data processing module for preprocessing and fusing the collected charging pile status data, vehicle information data, environmental data, and safety monitoring data to obtain a comprehensive feature dataset; a digital twin module for constructing a digital twin model, inputting the comprehensive feature dataset into the virtual model to simulate the charging pile's operation and predict its operating data; and a status monitoring module for constructing a status monitoring model, importing the predicted charging pile operating data into the trained status monitoring model to obtain the charging pile's safety status. This system can effectively predict the charging pile's operating condition, thereby identifying potential problems in advance and avoiding system failures.

Description

Technical Field

[0001] This invention relates to the field of digital twin technology, and more specifically, to a charging pile control system based on digital twins. Background Technology

[0002] With increasing global awareness of environmental protection, reducing carbon emissions and promoting green travel have become a global consensus. Electric vehicles, as a representative of low-carbon travel, are seeing their widespread adoption become an inevitable trend. To address environmental challenges and energy transition, governments worldwide have introduced policies to support the development of the electric vehicle industry. The rapid growth in the number of electric vehicles has placed higher demands on charging infrastructure, driving the research and application of charging pile control systems.

[0003] Patent publication number CN118082585A discloses a charging pile safety charging control system and method. This system and method can promptly eliminate safety hazards, ensure vehicle charging safety, provide more reasonable safety measures, offer more humanized operation services, and reduce the probability of accidents at charging stations. The charging pile safety charging control system includes an IoT communication gateway, a data preprocessing unit, a charging safety feature model, a risk assessment engine, a strategy distributor, and a scheduling control unit. This invention, through the charging safety control system, comprehensively collects and analyzes data from multiple aspects such as the vehicle, charging pile, and battery in real time. Based on the assessment, it generates control measures to achieve different levels and strategy modes, implementing differentiated adjustment and control for different risk levels, and outputting comprehensive charging strategy control and health reports. This invention is applied to the technical field of safe charging of charging piles.

[0004] The existing charging pile control system mainly has the following problems:

[0005] Without considering dynamic adjustment of feature weights, the model's predictions may be inaccurate because they fail to reflect the constantly changing environment and conditions during charging pile operation. Low-precision predictions may lead managers to rely on incorrect information when making decisions, thus affecting operational efficiency and safety. Failure to dynamically monitor and adjust feature weights may result in excessive maintenance or unnecessary downtime, increasing operating costs. This will affect overall economic benefits and challenge the sustainability of charging pile operation. Existing technologies, without dynamic monitoring and intelligent weight adjustment, may not be able to keep up with the trend of intelligent development, causing the technology to lag behind market demands and hindering industry innovation and progress.

[0006] Failure to effectively analyze and manage the thermal effects of charging piles may lead to prolonged operation of equipment in high-temperature environments; poor heat dissipation of charging piles, increasing the risk of failure, and in severe cases, potentially causing fires or explosions; failure to use thermal conductivity limiting formulas to control heat conduction may result in overheating of charging piles; overheating not only shortens the service life of the equipment but may also cause permanent damage, resulting in costly repairs or replacements; failure to consider the impact of thermal conductivity on temperature means that the safe operation of charging piles cannot be guaranteed; and it increases safety hazards, especially under high loads or extreme environments, potentially threatening users and the surrounding environment.

[0007] Failure to monitor and assess the operating status of charging piles in real time may lead to the neglect of potential risks; potential faults or anomalies may continue to develop without being detected in time, eventually leading to accidents; the lack of a dynamically adjusted risk assessment model makes it impossible to adapt to environmental changes; the operational safety of charging piles under different environmental conditions may not be guaranteed, resulting in excessively high risks in some cases; and the lack of effective risk monitoring and management may lead to the failure to detect potential risks in time, potentially resulting in accidents and increased maintenance costs.

[0008] In view of this, the present invention proposes a charging pile control system based on digital twins to solve the above problems. Summary of the Invention

[0009] To overcome the aforementioned shortcomings of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a charging pile control system based on digital twins, comprising:

[0010] The data acquisition module is used to collect charging pile status data, vehicle information data, environmental data, and safety monitoring data.

[0011] The data processing module is used to preprocess and correlate the collected charging pile status data, vehicle information data, environmental data, and safety monitoring data to obtain a comprehensive feature dataset.

[0012] The digital twin module is used to build a digital twin model. By inputting a comprehensive feature dataset into the virtual model, the operation process of the charging pile is simulated, and the operation data of the charging pile is predicted.

[0013] The status monitoring module is used to build a status monitoring model. It imports the predicted charging pile operation data into the trained status monitoring model to obtain the safety status of the charging pile.

[0014] The status assessment module is used to compare the predicted safety status of the charging pile with the preset safety status threshold of the charging pile to assess whether the safety status of the charging pile meets the standard.

[0015] The intelligent scheduling module is used to optimize the scheduling and usage strategies of charging piles; the modules are connected to each other via wired and / or wireless means.

[0016] Furthermore, the charging pile status data includes operational data and basic information data; the operational data includes the power, current, voltage, timestamp, charging duration, utilization frequency, and usage frequency of the charging pile during operation; the basic information data includes the physical characteristics, satellite positioning, design drawings, service life, and charging type of the charging pile.

[0017] Vehicle information data includes vehicle type data and battery data; battery data includes battery type, rated capacity, health status, charge status, and battery temperature;

[0018] Environmental data includes ambient temperature, ambient humidity, ambient brightness, and traffic flow.

[0019] Safety monitoring data includes smoke monitoring data, fire monitoring data, temperature monitoring data, electrical monitoring data, insulation resistance testing data, short circuit monitoring data, and overload protection data.

[0020] Furthermore, the method for preprocessing the collected charging pile status data, vehicle information data, environmental data, and safety monitoring data includes:

[0021] Clustering algorithms are used to detect outliers in charging pile status data, vehicle information data, environmental data, and safety monitoring data. The identified outliers are then removed to obtain charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset.

[0022] The acquired charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset are standardized and normalized, and converted into a unified standard normal distribution form according to timestamps to obtain the final charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset.

[0023] Furthermore, the method for obtaining the comprehensive feature dataset includes:

[0024] Calculate the Jaccard similarity between different features in the charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset and the charging pile operation data, and assign weights to features in each dataset based on the Jaccard similarity.

[0025] After standardizing and weighting each feature, a quantified charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset are formed. These datasets are then fused together using a weighted model to form a comprehensive feature dataset. The charging pile status dataset is denoted as D1, the vehicle information dataset as D2, the environmental dataset as D3, and the safety monitoring dataset as D4.

[0026] The weighted model is: YR=D1·ω1+D2·ω2+D3·ω3+D4·ω4; where YR is the comprehensive feature dataset; ω1 is the weight coefficient of the charging pile status dataset; ω2 is the weight coefficient of the vehicle information dataset; ω3 is the weight coefficient of the environmental dataset; and ω4 is the weight coefficient of the safety monitoring dataset.

[0027] Furthermore, the method of assigning weights to features in each dataset using Jaccard similarity includes:

[0028] Define a time window W. At the end of each time window, collect the latest data from the charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset within that window. For each feature in the data, calculate the Jaccard similarity using the latest dataset at the end of time window W. Let the set of values ​​for the charging pile operation data be C and the feature data be F. g The Jaccard similarity calculation formula is: Among them, f g Let J be the set of feature values; J is the similarity score; |f g ∩C| represents the set of feature values ​​f. g The size of the intersection of the set of values ​​C with the charging pile operation data; |f g ∪C| is the set of feature values ​​f g The size of the union of the set of values ​​C of the charging pile operation data;

[0029] Based on the calculated Jaccar similarity, a weight factor σ is assigned to each feature in the charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset. h The weighting factor of each feature is adjusted by adjusting the weighting factor of the model.

[0030] The weight factor adjustment model is as follows: Where, σ′ h The adjusted weighting factor; r h To adjust the adjustment coefficient of the h-th weighting factor; This represents the accumulation of the h-th feature over time t′. The summation of all features over time t′; K is the type of feature; T′ is the number of time windows; t′ is the index of time; h is the index of the weight factor.

[0031] Furthermore, the method for constructing the digital twin model includes:

[0032] S61, Physical Structure Module: A 3D physical model of the charging pile is constructed using CAD software. This model is based on the basic information data of the charging pile and reflects its physical structure. Differential equations are used to simulate the electrical behavior of the charging pile, mathematically expressed as: Where I(t) is the charging current, V(t) is the charging voltage, and R(t) is the resistance;

[0033] The thermal behavior of charging piles is simulated using the finite element method. Modeled by the heat conduction equation, the mathematical expression is as follows: Where T is the temperature distribution, ρ is the density, and c is the density. p is the specific heat capacity, t is time, k is thermal conductivity, Q is the heat source; ▽ is the vector differential operator; The differential symbol;

[0034] The thermal conductivity k in the heat conduction equation is constrained by a thermal conductivity constraint formula, which is: Where k(t) is the thermal conductivity after constraint; T(t) is the temperature distribution function as a function of time; α is the coefficient of influence of temperature on thermal conductivity; and β is the coefficient of influence of charging time on thermal conductivity.

[0035] S62, Operation Mechanism Module; This module uses the law of electromagnetic induction to construct the power equations, thereby describing the power transmission process; the mathematical formulas of the power equations are as follows: Where V is voltage, L is inductance, I is current, R is resistance, dI is current increment, and dt is time increment;

[0036] S63, Dynamic Response Module: Constructs a dynamic response model to accurately model the input-output relationship and state changes of the charging pile, thereby ensuring that the dynamic behavior of the charging pile under different conditions is consistent with reality; the dynamic response model is constructed using state-space equations, with the specific mathematical formula as follows: Where x(t) is the state vector, u(t) is the input vector, y(t) is the output vector, and A, B, C, and D are the parameters of the state-space model;

[0037] S64, Data-driven module; The data-driven module consists of a status monitoring model, which predicts the operating status of the charging pile.

[0038] Furthermore, the training method for the state monitoring model includes:

[0039] The dataset is divided into training and testing sets to build a status monitoring model. The sample set is a subset of the dataset, and each sample set includes historical charging pile operation data and the corresponding safety status of the charging pile. The input data of the model is the historical charging pile operation data. The output label is the safety status of the charging pile. This status monitoring model is a random forest regressor model.

[0040] Initialize the state monitoring model and set the hyperparameters for the number and depth of trees; train the state monitoring model using the training set, adjust the model's hyperparameters using k-fold cross-validation, and fine-tune the initially set parameters; evaluate the model's performance using the test set data, and use the coefficient of determination to evaluate the difference between the calculated prediction results and the true labels;

[0041] Based on the model performance feedback, the number and depth of trees are adjusted to optimize the model. The model is then retrained using the adjusted hyperparameters. Training stops when the preset model complexity is reached, resulting in the final trained state monitoring model. The trained state monitoring model is then used to predict the current charging pile operation data to determine the safety status of the charging pile.

[0042] Furthermore, the method for optimizing the model by adjusting the number and depth hyperparameters of the trees includes:

[0043] The number and depth of trees are adjusted using a genetic algorithm. During the iteration of the genetic algorithm, the number and depth of the trees corresponding to the chromosome with the highest fitness are recorded as the optimal combination of hyperparameters.

[0044] For the hyperparameters of the number and depth of trees, a minimum threshold for the number of trees and a minimum threshold for the depth are randomly preset; at the same time, a maximum threshold for the number of trees and a maximum threshold for the depth are also randomly preset; the thresholds are gradually increased or decreased until the optimal threshold is found, the performance of the model on the test set is judged, and the optimal hyperparameters of the number and depth of trees are obtained.

[0045] Furthermore, the method for comparing the predicted safety status of the charging pile with a preset safety status threshold for the charging pile to assess whether the safety status of the charging pile meets the standard includes:

[0046] If the predicted safety status of the charging pile matches the preset safety status threshold of the charging pile, then the safety status of the charging pile is deemed to meet the standard.

[0047] If the predicted safety status of the charging pile does not match the preset safety status threshold of the charging pile, the safety status of the charging pile is determined to be substandard.

[0048] Furthermore, the method for optimizing the scheduling and usage strategy of charging piles includes:

[0049] If the charging pile meets the safety standards, the intelligent control terminal of the charging pile will generate a safety command to indicate that the charging pile is in normal condition.

[0050] If the safety status of the charging pile is not up to standard, the intelligent control terminal of the charging pile will generate a danger command, automatically generate a warning message, and collect parameter data of the substandard charging pile.

[0051] The non-compliant parameter data includes the power generated during the operation of the charging pile, the non-compliant current, the non-compliant voltage, the charging pile's operating time, and the charging pile's own temperature;

[0052] A charging pile status risk assessment model is constructed based on the non-compliant parameter data. The model is then used to assess the status risk of the charging piles, yielding a status hazard coefficient. The charging pile risk assessment model is as follows: Where RE is the charging pile's state hazard factor; P is the power generated during the charging pile's operation; τ is the substandard current; V' is the substandard voltage; T sf T represents the charging station's operating time. run δ1 is the charging pile's own temperature; δ2 is the power weighting factor generated during the charging pile's operation; δ3 is the substandard current weighting factor; δ4 is the charging pile's operating time weighting factor; and δ5 is the charging pile's own temperature weighting factor.

[0053] The technical effects and advantages of the charging pile control system based on digital twins of the present invention are as follows:

[0054] This invention enables dynamic monitoring of the charging pile's operating status by collecting the latest charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset at the end of each time window, thus reflecting the actual situation of the system in a timely manner. By calculating the Jaccard similarity of each feature and assigning weight factors to the features based on this similarity, the feature weights can be flexibly adjusted according to the actual operating conditions, enhancing the model's adaptability to changing environments.

[0055] The finite element method is used to model the thermal behavior of charging piles through heat conduction equations, ensuring effective analysis and management of thermal effects during actual operation. The effectiveness of heat conduction is controlled through thermal conductivity limiting formulas to prevent equipment damage due to overheating. A dynamic response model is constructed, using state-space equations to accurately model the input-output relationship and state changes of the charging pile. This ensures that the dynamic behavior of the charging pile under different conditions is consistent with reality, improving system stability and reliability. By limiting thermal conductivity, the temperature of the charging pile during operation can be effectively controlled, preventing overheating and ensuring that the charging pile operates within a safe range.

[0056] By transforming different parameters into a quantified risk coefficient through the charging pile status risk assessment model, the status risk of charging piles becomes easier to understand and manage. Real-time monitoring and evaluation of the operating status of charging piles can promptly identify potential risks and take necessary early warning measures to reduce the possibility of accidents. The model can be dynamically adjusted based on real-time data to reflect risk changes under different environmental conditions, thereby improving the operational safety of charging piles. Attached Figure Description

[0057] Figure 1 This is a schematic diagram of a charging pile control system based on digital twins according to the present invention;

[0058] Figure 2 This is a schematic diagram of a charging pile control method based on digital twins according to the present invention. Detailed Implementation

[0059] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0060] Example 1

[0061] See also Figure 1 As shown in this embodiment, a charging pile control system based on digital twins includes:

[0062] The data acquisition module is used to collect charging pile status data, vehicle information data, environmental data, and safety monitoring data.

[0063] The data processing module is used to preprocess and correlate the collected charging pile status data, vehicle information data, environmental data, and safety monitoring data to obtain a comprehensive feature dataset.

[0064] The digital twin module is used to build a digital twin model. By inputting a comprehensive feature dataset into the virtual model, the operation process of the charging pile is simulated, and the operation data of the charging pile is predicted.

[0065] The status monitoring module is used to build a status monitoring model. It imports the predicted charging pile operation data into the trained status monitoring model to obtain the safety status of the charging pile.

[0066] The status assessment module is used to compare the predicted safety status of the charging pile with the preset safety status threshold of the charging pile to assess whether the safety status of the charging pile meets the standard.

[0067] The intelligent scheduling module is used to optimize the scheduling and usage strategies of charging piles; the modules are connected to each other via wired and / or wireless means.

[0068] The charging pile status data includes operational data and basic information data; the operational data includes the power, current, voltage, timestamp, charging duration, utilization frequency and usage frequency of the charging pile during operation; the basic information data includes the physical characteristics, satellite positioning, design drawings, service life and charging type of the charging pile.

[0069] Vehicle information data includes vehicle type data and battery data; battery data includes battery type, rated capacity, health status, charge status, and battery temperature;

[0070] Environmental data includes ambient temperature, ambient humidity, ambient brightness, and traffic flow.

[0071] Safety monitoring data includes smoke monitoring data, fire monitoring data, temperature monitoring data, electrical monitoring data, insulation resistance testing data, short circuit monitoring data, and overload protection data.

[0072] Methods for preprocessing the collected charging pile status data, vehicle information data, environmental data, and safety monitoring data include:

[0073] Clustering algorithms are used to detect outliers in charging pile status data, vehicle information data, environmental data, and safety monitoring data. The identified outliers are then removed to obtain charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset.

[0074] The acquired charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset are standardized and normalized, and converted into a unified standard normal distribution form according to timestamps to ensure that the data can be integrated and analyzed in chronological order; thus obtaining the final charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset.

[0075] Methods for obtaining comprehensive feature datasets include:

[0076] Calculate the Jaccard similarity between different features in the charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset and the charging pile operation data, and assign weights to features in each dataset based on the Jaccard similarity.

[0077] After standardizing and weighting each feature, a quantified charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset are formed. These datasets are then fused together using a weighted model to form a comprehensive feature dataset. The charging pile status dataset is denoted as D1, the vehicle information dataset as D2, the environmental dataset as D3, and the safety monitoring dataset as D4.

[0078] The weighted model is: YR=D1·ω1+D2·ω2+D3·ω3+D4·ω4; where YR is the comprehensive feature dataset; ω1 is the weight coefficient of the charging pile status dataset; ω2 is the weight coefficient of the vehicle information dataset; ω3 is the weight coefficient of the environmental dataset; and ω4 is the weight coefficient of the safety monitoring dataset.

[0079] Methods for assigning weights to features in each dataset using Jaccard similarity include:

[0080] Define a time window W. At the end of each time window, collect the latest data from the charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset within that window. For each feature in the data, calculate the Jaccard similarity using the latest dataset at the end of time window W. Let the set of values ​​for the charging pile operation data be C and the feature data be F. g The Jaccard similarity calculation formula is: Among them, f g Let J be the set of feature values; J is the similarity score; |f g ∩C| represents the set of feature values ​​f. g The size of the intersection of the set of values ​​C with the charging pile operation data; |f g ∪C| is the set of feature values ​​f g The size of the union of the set of values ​​C of the charging pile operation data;

[0081] Based on the calculated Jaccar similarity, a weight factor σ is assigned to each feature in the charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset. h The weighting factor of each feature is adjusted by adjusting the weighting factor of the model.

[0082] The weight factor adjustment model is as follows: Where, σ′ h The adjusted weighting factor; r h To adjust the adjustment coefficient of the h-th weighting factor; This represents the accumulation of the h-th feature over time t′. The summation of all features over time t′; K is the type of feature; T′ is the number of time windows; t′ is the index of time; h is the index of the weight factor.

[0083] Methods for constructing digital twin models include:

[0084] S61, Physical Structure Module: A 3D physical model of the charging pile is constructed using CAD software. This model is based on the basic information data of the charging pile and reflects its physical structure. Differential equations are used to simulate the electrical behavior of the charging pile, mathematically expressed as: Where I(t) is the charging current, V(t) is the charging voltage, and R(t) is the resistance, which depends on the changes in time and temperature. The thermal behavior of the charging pile is simulated using the finite element method, modeled through the heat conduction equation, and mathematically expressed as: Where T is the temperature distribution, ρ is the density, cp is the specific heat capacity, t is the time, k is the thermal conductivity, Q is the heat source; ▽ is the vector differential operator; The differential symbol;

[0085] The thermal conductivity k in the heat conduction equation is constrained by a thermal conductivity constraint formula, which is: Where k(t) is the thermal conductivity after constraint; T(t) is the temperature distribution function as a function of time; α is the coefficient of influence of temperature on thermal conductivity; and β is the coefficient of influence of charging time on thermal conductivity.

[0086] For example, the temperature T changes with time during the charging process and is related to the charging time; the function can be described as: T(t) = 2 + 0.5t, in degrees Celsius, with an initial temperature of 20°C and a temperature increase of 5°C per hour.

[0087] The influence coefficient α of temperature on thermal conductivity is 0.2, and the influence coefficient β of charging time on thermal conductivity is 0.1.

[0088] When t is 0 That is, the thermal conductivity is 40% at the beginning of charging;

[0089] When t is 2 hours That is, after charging for 2 hours, the thermal conductivity is 50%;

[0090] When t is 5 hours That is, after charging for 5 hours, the thermal conductivity reaches 60%;

[0091] When t is 10 That is, after charging for 10 hours, the thermal conductivity reaches 70%;

[0092] When the temperature T increases, the thermal conductivity k increases, indicating that the device needs better thermal conductivity when the temperature rises; when the charging time t increases, the denominator increases, and the thermal conductivity k decreases accordingly, indicating that long-term charging may reduce the thermal conductivity effect.

[0093] S62, Operation Mechanism Module; This module uses the law of electromagnetic induction to construct the power equations, thereby describing the power transmission process; the mathematical formulas of the power equations are as follows: Where V is voltage, L is inductance, I is current, R is resistance, dI is current increment, and dt is time increment;

[0094] S63, Dynamic Response Module: Constructs a dynamic response model to accurately model the input-output relationship and state changes of the charging pile, and tracks the behavior of the charging pile under different loads and external conditions in real time, thereby ensuring that the dynamic behavior of the charging pile under different conditions is consistent with reality; the dynamic response model is constructed using state-space equations, and the specific mathematical formula is as follows: Where x(t) is the state vector (such as physical quantities like current and voltage), u(t) is the input vector (external power or control signal); y(t) is the output vector (monitored charging pile status, such as temperature and current); A, B, C and D are the parameters of the state-space model;

[0095] S64, Data-driven module; The data-driven module consists of a status monitoring model, which predicts the operating status of the charging pile.

[0096] Training methods for state monitoring models include:

[0097] The dataset is divided into training and testing sets to build a status monitoring model. The sample set is a subset of the dataset, and each sample set includes historical charging pile operation data and the corresponding safety status of the charging pile. The input data of the model is the historical charging pile operation data. The output label is the safety status of the charging pile. This status monitoring model is a random forest regressor model.

[0098] Initialize the state monitoring model and set the hyperparameters for the number and depth of trees; train the state monitoring model using the training set, adjust the model's hyperparameters using k-fold cross-validation, and fine-tune the initially set parameters; evaluate the model's performance using the test set data, and use the coefficient of determination to evaluate the difference between the calculated prediction results and the true labels;

[0099] Based on the model performance feedback, the number and depth of trees are adjusted to optimize the model. The model is then retrained using the adjusted hyperparameters. Training stops when the preset model complexity is reached, resulting in the final trained state monitoring model. The trained state monitoring model is then used to predict the current charging pile operation data to determine the safety status of the charging pile.

[0100] Methods for optimizing the model by adjusting the hyperparameters of the number and depth of trees include:

[0101] The number and depth of trees are adjusted using a genetic algorithm. During the iteration of the genetic algorithm, the number and depth of the trees corresponding to the chromosome with the highest fitness are recorded as the optimal combination of hyperparameters.

[0102] For the hyperparameters of the number and depth of trees, a minimum threshold for the number of trees and a minimum threshold for the depth are randomly preset; at the same time, a maximum threshold for the number of trees and a maximum threshold for the depth are also randomly preset; the thresholds are gradually increased or decreased until the optimal threshold is found, the performance of the model on the test set is judged, and the optimal hyperparameters of the number and depth of trees are obtained.

[0103] The method for assessing whether the safety status of a charging pile meets the standards by comparing the predicted safety status with the preset safety status thresholds for the charging pile includes:

[0104] If the predicted safety status of the charging pile matches the preset safety status threshold of the charging pile, then the safety status of the charging pile is deemed to meet the standard.

[0105] If the predicted safety status of the charging pile does not match the preset safety status threshold of the charging pile, the safety status of the charging pile is determined to be substandard.

[0106] Methods for optimizing the scheduling and usage strategies of charging stations include:

[0107] If the charging pile meets the safety standards, the intelligent control terminal of the charging pile will generate a safety command to indicate that the charging pile is in normal condition.

[0108] If the safety status of the charging pile is not up to standard, the intelligent control terminal of the charging pile will generate a danger command, automatically generate a warning message, and collect parameter data of the substandard charging pile.

[0109] The non-compliant parameter data includes the power generated during the operation of the charging pile, the non-compliant current, the non-compliant voltage, the charging pile's operating time, and the charging pile's own temperature;

[0110] A charging pile status risk assessment model is constructed based on the non-compliant parameter data. The model is then used to assess the status risk of the charging piles, yielding a status hazard coefficient. The charging pile risk assessment model is as follows: Where RE is the charging pile's state hazard factor; P is the power generated during the charging pile's operation; τ is the substandard current; V' is the substandard voltage; T sf T represents the charging station's operating time. run δ1 is the charging pile's own temperature; δ2 is the power weighting factor generated during the charging pile's operation; δ3 is the substandard current weighting factor; δ4 is the charging pile's operating time weighting factor; δ5 is the charging pile's own temperature weighting factor.

[0111] For example, if the power generated by the charging pile during operation is 2 kilowatts, the substandard current is 5A, the substandard voltage is 300V, the charging pile operation time is 10 hours, the charging pile temperature is 75℃, and δ1 is 0.3, δ2 is 0.3, δ3 is 0.2, δ4 is 0.1, and δ5 is 0.1, then the charging pile's state danger factor is 1.2.

[0112] The preset safety status threshold for charging piles is set by staff. The safety status of different charging piles is collected through the intelligent control terminal of the charging piles, and the average value of the safety status of multiple charging piles is taken as the preset safety status threshold for the charging piles.

[0113] In this embodiment, by collecting the latest charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset at the end of each time window, dynamic monitoring of the charging pile's operating status can be achieved, reflecting the actual situation of the system in a timely manner. By calculating the Jaccard similarity of each feature and assigning weight factors to the features based on the similarity, the feature weights can be flexibly adjusted according to the actual operating conditions, enhancing the model's adaptability to changing environments.

[0114] The finite element method is used to model the thermal behavior of charging piles through heat conduction equations, ensuring effective analysis and management of thermal effects during actual operation. The effectiveness of heat conduction is controlled through thermal conductivity limiting formulas to prevent equipment damage due to overheating. A dynamic response model is constructed, using state-space equations to accurately model the input-output relationship and state changes of the charging pile. This ensures that the dynamic behavior of the charging pile under different conditions is consistent with reality, improving system stability and reliability. By limiting thermal conductivity, the temperature of the charging pile during operation can be effectively controlled, preventing overheating and ensuring that the charging pile operates within a safe range.

[0115] By transforming different parameters into a quantified risk coefficient through the charging pile status risk assessment model, the status risk of charging piles becomes easier to understand and manage. Real-time monitoring and evaluation of the operating status of charging piles can promptly identify potential risks and take necessary early warning measures to reduce the possibility of accidents. The model can be dynamically adjusted based on real-time data to reflect risk changes under different environmental conditions, thereby improving the operational safety of charging piles.

[0116] Example 2

[0117] See also Figure 2 As shown, parts not described in detail in this embodiment are described in Embodiment 1. A charging pile control method based on digital twins is provided, including:

[0118] S1. Collect charging pile status data, vehicle information data, environmental data, and safety monitoring data;

[0119] S2. Preprocess and correlate the collected charging pile status data, vehicle information data, environmental data, and safety monitoring data to obtain a comprehensive feature dataset;

[0120] S3. Construct a digital twin model, input the comprehensive feature dataset into the virtual model, simulate the operation process of the charging pile, and predict the charging pile operation data;

[0121] S4. Construct a status monitoring model, import the predicted charging pile operation data into the trained status monitoring model, and obtain the safety status of the charging pile.

[0122] S5. Compare the predicted safety status of the charging pile with the preset safety status threshold of the charging pile to evaluate whether the safety status of the charging pile meets the standard.

[0123] S6. Optimize the scheduling and usage strategies of charging stations.

[0124] Since the electronic device described in this embodiment is the electronic device used in implementing the charging pile control system based on digital twins in this application embodiment, those skilled in the art can understand the specific implementation methods and various variations of the electronic device in this embodiment based on the charging pile control system based on digital twins described in this application embodiment. Therefore, how the electronic device implements the method in this application embodiment will not be described in detail here. Any electronic device used by those skilled in the art in implementing the charging pile control system based on digital twins in this application embodiment falls within the scope of protection of this application.

[0125] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0126] The above description is merely a preferred embodiment of the present invention, and the scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for users of ordinary technical skills, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A charging pile control system based on digital twins, characterized in that, include: The data acquisition module is used to collect charging pile status data, vehicle information data, environmental data, and safety monitoring data. The charging pile status data includes operational data and basic information data; the operational data includes the power, current, voltage, timestamp, charging duration, utilization frequency, and usage frequency of the charging pile during operation; the basic information data includes the physical characteristics, satellite positioning, design drawings, service life, and charging type of the charging pile. Vehicle information data includes vehicle type data and battery data; battery data includes battery type, rated capacity, health status, charge status, and battery temperature; Environmental data includes ambient temperature, ambient humidity, ambient brightness, and traffic flow. Safety monitoring data includes smoke monitoring data, fire monitoring data, temperature monitoring data, electrical monitoring data, insulation resistance testing data, short circuit monitoring data, and overload protection data; The data processing module is used to preprocess and correlate the collected charging pile status data, vehicle information data, environmental data, and safety monitoring data to obtain a comprehensive feature dataset. The digital twin module is used to build a digital twin model. By inputting a comprehensive feature dataset into the virtual model, the operation process of the charging pile is simulated, and the operation data of the charging pile is predicted. The status monitoring module is used to build a status monitoring model. It imports the predicted charging pile operation data into the trained status monitoring model to obtain the safety status of the charging pile. The status assessment module is used to compare the predicted safety status of the charging pile with the preset safety status threshold of the charging pile to assess whether the safety status of the charging pile meets the standard. The intelligent scheduling module is used to optimize the scheduling and usage strategies of charging piles; the modules are connected to each other via wired and / or wireless means.

2. The charging pile control system based on digital twin according to claim 1, characterized in that, The method for preprocessing the collected charging pile status data, vehicle information data, environmental data, and safety monitoring data includes: Density clustering algorithm is used to detect outliers in charging pile status data, vehicle information data, environmental data and safety monitoring data. The identified outliers are then removed to obtain charging pile status dataset, vehicle information dataset, environmental dataset and safety monitoring dataset. The acquired charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset are standardized and normalized, and converted into a unified standard normal distribution form according to timestamps to obtain the final charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset.

3. A charging pile control system based on digital twins according to claim 2, characterized in that, The method for obtaining the comprehensive feature dataset includes: Calculate the Jaccard similarity between different features in the charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset and the charging pile operation data, and assign weights to features in each dataset based on the Jaccard similarity. After standardizing and weighting each feature, quantified charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset are generated. These datasets are then fused using a weighted model to form a comprehensive feature dataset. The charging pile status dataset is denoted as... The vehicle information dataset is denoted as The environmental dataset is denoted as The safety monitoring dataset is denoted as ; The weighted model is: ;in, For comprehensive feature datasets; These are the weighting coefficients for the charging pile status dataset; These are the weighting coefficients for the vehicle information dataset; These are the weighting coefficients for the environmental dataset; These are the weighting coefficients for the safety monitoring dataset.

4. A charging pile control system based on digital twins according to claim 3, characterized in that, The method of assigning weights to features in each dataset using Jaccard similarity includes: Define a time window At the end of each time window, the latest charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset within that window are collected; for each feature in the data, within the time window... At the end, the Jaccard similarity is calculated using the latest dataset; let the set of values ​​for the charging pile operation data be... and feature data are The Jaccard similarity calculation formula is: ;in, The set of values ​​for the feature; The similarity score; The set of values ​​for the features The set of values ​​related to the charging pile operation data The size of the intersection; The set of values ​​for the features The set of values ​​related to the charging pile operation data The size of the union; Based on the calculated Jaccar similarity, weight factors are assigned to each feature in the charging pile status dataset, vehicle information dataset, environmental dataset, and safety monitoring dataset. The weighting factor of each feature is adjusted by adjusting the weighting factor of the model. The weight factor adjustment model is as follows: ;in, The adjusted weighting factor; To adjust the first Adjustment coefficients for each weighting factor; For the first One characteristic in time Accumulation on; For all features in time Accumulation on; Types characterized by features; The number of time windows; For time indexing; This is the index of the weighting factor.

5. A charging pile control system based on digital twins according to claim 4, characterized in that, The method for constructing the digital twin model includes: S61, Physical Structure Module: A 3D physical model of the charging pile is constructed using CAD software. This model is based on the basic information data of the charging pile and reflects its physical structure. Differential equations are used to simulate the electrical behavior of the charging pile, mathematically expressed as: ;in, For charging current, This is the charging voltage. For resistance; The thermal behavior of charging piles is simulated using the finite element method. Modeled by the heat conduction equation, the mathematical expression is as follows: ;in, For temperature distribution, For density, For specific heat capacity, For time, Thermal conductivity, As a heat source; It is a vector differential operator; The differential symbol; Thermal conductivity in the heat conduction equation is determined by the thermal conductivity constraint formula. To impose a limit, the thermal conductivity limit formula is: ;in, Thermal conductivity after limitation; This is a time-varying temperature distribution function; This is the coefficient representing the effect of temperature on thermal conductivity. The coefficient representing the effect of charging time on thermal conductivity; S62, Operation Mechanism Module; This module uses the law of electromagnetic induction to construct the power equations, thereby describing the power transmission process; the mathematical formulas of the power equations are as follows: ;in, For voltage, For inductance, For current, For resistance, For current increment, For time increments; S63, Dynamic Response Module: Constructs a dynamic response model to accurately model the input-output relationship and state changes of the charging pile, thereby ensuring that the dynamic behavior of the charging pile under different conditions is consistent with reality; the dynamic response model is constructed using state-space equations, with the specific mathematical formula as follows: ;in, For state vectors, The input vector; This is the output vector; , , and These are the parameters of the state-space model; S64, Data-driven module; The data-driven module consists of a status monitoring model, which predicts the operating status of the charging pile.

6. A charging pile control system based on digital twins according to claim 5, characterized in that, The training method for the state monitoring model includes: The dataset is divided into training and testing sets to build a status monitoring model. The sample set is a subset of the dataset, and each sample set includes historical charging pile operation data and the corresponding safety status of the charging pile. The input data of the model is the historical charging pile operation data. The output label is the safety status of the charging pile. This status monitoring model is a random forest regressor model. Initialize the state monitoring model and set the hyperparameters for the number and depth of trees; train the state monitoring model using the training set, adjust the model's hyperparameters using k-fold cross-validation, and fine-tune the initially set parameters; evaluate the model's performance using the test set data, and use the coefficient of determination to evaluate the difference between the calculated prediction results and the true labels; Based on the model performance feedback, the number and depth of trees are adjusted to optimize the model. The model is then retrained using the adjusted hyperparameters. Training stops when the preset model complexity is reached, resulting in the final trained state monitoring model. The trained state monitoring model is then used to predict the current charging pile operation data to determine the safety status of the charging pile.

7. A charging pile control system based on digital twins according to claim 6, characterized in that, The methods for optimizing the model by adjusting the number and depth hyperparameters of the trees include: The number and depth of trees are adjusted using a genetic algorithm. During the iteration of the genetic algorithm, the number and depth of the trees corresponding to the chromosome with the highest fitness are recorded as the optimal combination of hyperparameters. For the hyperparameters of the number and depth of trees, a minimum threshold for the number of trees and a minimum threshold for the depth are randomly preset; at the same time, a maximum threshold for the number of trees and a maximum threshold for the depth are also randomly preset; the thresholds are gradually increased or decreased until the optimal threshold is found, the performance of the model on the test set is judged, and the optimal hyperparameters of the number and depth of trees are obtained.

8. A charging pile control system based on digital twins according to claim 7, characterized in that, The method for comparing the predicted safety status of the charging pile with a preset safety status threshold to assess whether the safety status of the charging pile meets the standard includes: If the predicted safety status of the charging pile matches the preset safety status threshold of the charging pile, then the safety status of the charging pile is deemed to meet the standard. If the predicted safety status of the charging pile does not match the preset safety status threshold of the charging pile, the safety status of the charging pile is determined to be substandard.

9. A charging pile control system based on digital twins according to claim 8, characterized in that, The method for optimizing the scheduling and usage strategy of charging piles includes: If the charging pile meets the safety standards, the intelligent control terminal of the charging pile will generate a safety command to indicate that the charging pile is in normal condition. If the safety status of the charging pile is not up to standard, the intelligent control terminal of the charging pile will generate a danger command, automatically generate a warning message, and collect parameter data of the substandard charging pile. The non-compliant parameter data includes the power generated during the operation of the charging pile, the non-compliant current, the non-compliant voltage, the charging pile's operating time, and the charging pile's own temperature; A charging pile status risk assessment model is constructed based on the non-compliant parameter data. The model is then used to assess the status risk of the charging piles, yielding a status hazard coefficient. The charging pile risk assessment model is as follows: ;in, The condition risk factor of the charging pile; This refers to the power generated during the operation of the charging station; The current is below standard. The voltage is substandard. This refers to the charging station's operating time. The temperature of the charging station itself; This refers to the power weighting factor generated during the operation of the charging pile. Weighting factor for substandard current; Weighting factor for substandard voltage; The weighting factor for charging pile operating time; This is the temperature weighting factor for the charging pile itself.

Images