Overcharge station direct current micro-grid system control method

Abstract

The invention discloses an overcharge station direct current micro-grid system control method, which comprises the following steps: modeling an overcharge station direct current micro-grid system to obtain system parameters; constructing an initial deep learning network model, and training the initial deep learning network model based on the system parameters to obtain a target deep learning network model; based on the target deep learning network model, according to the real-time energy supply condition and the charging demand, the input proportion of different energy sources is dynamically adjusted, and the energy storage device is introduced to balance the supply and demand difference; through the wireless communication technology, the charging state and energy production and consumption conditions of the over-charging station direct current micro-grid system are monitored and managed remotely in real time, and energy distribution is optimized. Power fluctuation can be effectively reduced, system stability is improved, charging efficiency is improved, charging time is shortened, user experience is improved, and energy distribution is optimized. And the charging station can be quickly switched to the emergency mode under the power grid fault or extreme condition, so that the continuous operation of the charging station is ensured.

Description

Technical Field

[0001] The present invention belongs to the technical field of microgrid control, and in particular relates to a method for controlling a DC microgrid system of a supercharging station. Background Art

[0002] With the popularity of electric vehicles, the demand for charging facilities is growing. Traditional charging stations mostly use AC power for charging, but in order to achieve fast charging, high-power DC charging equipment is required, especially for supercharging stations that can provide fast charging services. As an important infrastructure for fast charging of electric vehicles, the stability and efficiency of supercharging stations are crucial to the charging experience of electric vehicles. The existing DC microgrid system has the following main defects in the application of supercharging stations:

[0003] Large power fluctuations: Due to the uncertainty of electric vehicle charging demand, the power fluctuations of microgrids are large, affecting the stability of the system; the large grid load can also easily cause grid instability during peak charging periods.

[0004] Energy management is not sophisticated: The energy management strategies in existing technologies are relatively extensive and cannot accurately match charging demand with energy supply, nor can they reasonably allocate and dispatch energy in the energy storage system, resulting in low energy utilization.

[0005] Poor emergency response capabilities: In the event of grid failure or extreme weather conditions, the existing system lacks an effective emergency response mechanism, affecting the continuous operation of charging stations.

[0006] Low charging efficiency: Due to the lack of intelligent charging scheduling strategy, the charging efficiency cannot be maximized and the demand for fast charging cannot be met, which affects the user experience.

[0007] Lack of adaptive regulation capability: When faced with power grid fluctuations or changes in charging demand, the existing system has limited regulation capabilities and is difficult to respond quickly. The control system is complex and difficult to achieve intelligent management. Summary of the invention

[0008] In order to solve the above technical problems, the present invention provides a supercharging station DC microgrid system control method, comprising:

[0009] Model the DC microgrid system of the supercharging station and obtain system parameters;

[0010] Constructing an initial deep learning network model, and training the initial deep learning network model based on the system parameters to obtain a target deep learning network model;

[0011] Based on the target deep learning network model, the input ratio of different energy sources is dynamically adjusted according to the real-time energy supply and charging demand, and energy storage devices are introduced to balance the supply and demand differences;

[0012] Through wireless communication technology, the charging status, energy production and consumption of the DC microgrid system of the supercharging station can be remotely monitored and managed in real time to optimize energy distribution.

[0013] Preferably, the process of modeling the supercharging station DC microgrid system and obtaining system parameters includes:

[0014] By installing sensors and data acquisition equipment at various key nodes of the DC microgrid system of the supercharging station, the DC microgrid of the supercharging station is monitored in real time to obtain the required parameter data; and the collected required parameter data is transmitted to the central control unit via a wireless or wired network; wherein the required parameter data includes but is not limited to the voltage, current, power demand and supply of the DC bus, the charging and discharging status of the energy storage system, the remaining power of the battery, the charging and discharging rate and temperature.

[0015] Preferably, the process of constructing the initial deep learning network model includes:

[0016] A long short-term memory network model and a convolutional neural network model are constructed, and the long short-term memory network model and the convolutional neural network model are fused to obtain a fused neural network model.

[0017] Preferably, the process of fusing the long short-term memory network model with the convolutional neural network model includes:

[0018] Standardize the required parameter data to eliminate the dimensional effects between different features and obtain time series data;

[0019] Performing feature extraction on the time series data through a convolutional neural network model, adding a pooling layer after the convolution layer, and automatically extracting spatial features of the time series data using the convolution layer and the pooling layer to obtain a feature map;

[0020] Downsampling is performed in the spatial dimension through the pooling layer, and the pooled feature map is converted into a one-dimensional vector and input into the LSTM layer. The gating mechanism of LSTM is used to capture the dynamic characteristics in the time series.

[0021] Add one or more fully connected layers after the LSTM layer to output the final prediction results;

[0022] After selecting appropriate activation functions and loss functions according to task requirements, a fusion layer is designed to integrate the local features of CNN and the global temporal features of LSTM to obtain a fusion neural network model;

[0023] The fusion neural network model is trained using a back propagation algorithm, and the prediction error is minimized by adjusting the network parameters.

[0024] Preferably, based on the target deep learning network model, according to the real-time energy supply situation and charging demand, the process of dynamically adjusting the input ratio of different energy sources includes:

[0025] Predict the charging power demand of the charging station based on the fusion neural network model and monitor the power demand of the charging station in real time;

[0026] The prediction results are compared with the power demand of the real-time monitoring charging station. When it is detected that the power demand exceeds the supply, the energy storage system is automatically started to discharge and fill the power gap; otherwise, charging is carried out; according to the real-time monitoring data and power demand prediction, the charging and discharging state of the energy storage system is dynamically adjusted through fuzzy logic control;

[0027] According to the charging demand forecast of electric vehicles and the current status of the energy storage system, a charging plan is formulated through the particle swarm optimization algorithm to carry out refined energy management, giving priority to using electricity during off-peak hours for charging and reducing the charging load during peak hours;

[0028] When a grid fault is detected, it automatically switches to a backup power source or energy storage system for power supply.

[0029] Preferably, the process of predicting the charging power demand of the charging station based on the fusion neural network model and monitoring the power demand of the charging station in real time includes:

[0030] Use historical charging data and machine learning algorithms to predict the charging demand of electric vehicles and obtain prediction results;

[0031] The prediction results are used to guide the charging and discharging strategy of the energy storage system and the operation scheduling of the charging station.

[0032] Preferably, the process of dynamically adjusting the charging and discharging state of the energy storage system through fuzzy logic control includes:

[0033] Real-time monitoring of the status of the energy storage system, including but not limited to key parameters such as battery state of charge, charge and discharge current, and voltage;

[0034] According to the characteristics of the energy storage system and the charging and discharging requirements, fuzzy control rules are designed to deal with the uncertainty and nonlinear problems in the charging and discharging process;

[0035] Convert the monitored system state parameters into fuzzy values, perform logical reasoning on the input fuzzy values ​​according to the fuzzy control rules, and determine the charging and discharging strategy;

[0036] The result of fuzzy reasoning is then converted into a specific numerical value to determine the specific charging and discharging power or current setting, and the charging and discharging state of the energy storage system is dynamically adjusted according to the defuzzified control instructions;

[0037] The system status after execution is used as feedback and compared with the expected target to further correct the fuzzy control rules and optimize the charging and discharging process.

[0038] Preferably, according to the prediction of charging demand of electric vehicles and the current state of the energy storage system, the process of formulating a charging plan through a particle swarm optimization algorithm for refined energy management includes:

[0039] Construct an optimization model, take the charging demand forecast of electric vehicles and the current state of the energy storage system as input, minimize the total cost or maximize the energy utilization efficiency as the objective function, and consider constraints such as power balance and charging and discharging rate limit, randomly initialize the position and speed of the particle swarm, and each particle represents a potential charging and discharging plan;

[0040] Calculate the fitness value of each particle, based on the objective function of total cost or energy efficiency, update the individual extreme value and global extreme value of the particle according to the fitness value, adjust the position and speed of the particle, simulate the foraging behavior of the bird flock, and find the optimal solution;

[0041] By repeatedly iterating and updating the particle position and velocity, the optimal charge and discharge plan is gradually approached. The iteration stop condition is set when the maximum number of iterations or the quality of the solution meets the preset threshold. After the stop condition is met, the optimal charge and discharge plan is extracted from the particle swarm.

[0042] The optimal charging and discharging plan is applied to the actual energy storage system, and the performance is evaluated and adjusted in real time based on feedback.

[0043] Compared with the prior art, the present invention has the following advantages and technical effects:

[0044] The present invention effectively reduces power fluctuation and improves system stability through dynamic power balance control.

[0045] The present invention can improve the utilization rate of the energy storage system and optimize energy distribution through a refined energy management strategy.

[0046] The present invention can improve charging efficiency, shorten charging time and enhance user experience through intelligent charging scheduling strategy.

[0047] The intelligent emergency response mechanism of the present invention can quickly switch to emergency mode in the event of a power grid failure or extreme conditions to ensure the continuous operation of the charging station. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0049] Figure 1The figure is a schematic diagram of a method flow of an embodiment of the present invention. DETAILED DESCRIPTION

[0050] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0051] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0052] like Figure 1 As shown, this embodiment provides a supercharging station DC microgrid system control method, including:

[0053] Model the DC microgrid system of the supercharging station and obtain system parameters;

[0054] Construct an initial deep learning network model, train the initial deep learning network model based on system parameters, and obtain a target deep learning network model;

[0055] Based on the target deep learning network model, the input ratio of different energy sources is dynamically adjusted according to the real-time energy supply and charging demand, and energy storage devices are introduced to balance the supply and demand differences;

[0056] Through wireless communication technology, the charging status, energy production and consumption of the DC microgrid system of the supercharging station can be remotely monitored and managed in real time to optimize energy distribution.

[0057] Further optimization scheme, the DC microgrid system of the supercharging station is modeled, and the process of obtaining system parameters includes:

[0058] By installing sensors and data acquisition equipment at various key nodes of the DC microgrid system of the supercharging station, the DC microgrid of the supercharging station is monitored in real time to obtain the required parameter data; and the collected required parameter data is transmitted to the central control unit via a wireless or wired network; the required parameter data includes but is not limited to the voltage, current, power demand and supply of the DC bus, the charging and discharging status of the energy storage system, the remaining power of the battery, the charging and discharging rate and temperature.

[0059] Further optimization scheme, the process of building the initial deep learning network model includes:

[0060] A long short-term memory network model and a convolutional neural network model are constructed, and the long short-term memory network model and the convolutional neural network model are fused to obtain a fused neural network model.

[0061] A further optimization scheme is to fuse the long short-term memory network model with the convolutional neural network model, including:

[0062] Standardize the required parameter data to eliminate the dimensional effects between different features and obtain time series data;

[0063] The convolutional neural network model is used to extract features from time series data. A pooling layer is added after the convolution layer. The convolution layer and the pooling layer are used to automatically extract the spatial features of the time series data and obtain a feature map.

[0064] Downsampling is performed in the spatial dimension through the pooling layer, and the pooled feature map is converted into a one-dimensional vector and input into the LSTM layer. The gating mechanism of LSTM is used to capture the dynamic characteristics in the time series.

[0065] Add one or more fully connected layers after the LSTM layer to output the final prediction results;

[0066] After selecting appropriate activation functions and loss functions according to task requirements, a fusion layer is designed to integrate the local features of CNN and the global temporal features of LSTM to obtain a fusion neural network model;

[0067] The back propagation algorithm is used to train the fusion neural network model, and the prediction error is minimized by adjusting the network parameters.

[0068] Further optimization solutions, based on the target deep learning network model, dynamically adjust the input ratio of different energy sources according to the real-time energy supply and charging demand. The process includes:

[0069] Predict the charging power demand of the charging station based on the fusion neural network model and monitor the power demand of the charging station in real time;

[0070] The prediction results are compared with the power demand of the real-time monitoring charging station. When it is detected that the power demand exceeds the supply, the energy storage system is automatically started to discharge and fill the power gap; otherwise, charging is carried out; according to the real-time monitoring data and power demand prediction, the charging and discharging state of the energy storage system is dynamically adjusted through fuzzy logic control;

[0071] According to the charging demand forecast of electric vehicles and the current status of the energy storage system, a charging plan is formulated through the particle swarm optimization algorithm to carry out refined energy management, giving priority to using electricity during off-peak hours for charging and reducing the charging load during peak hours;

[0072] When a grid fault is detected, it automatically switches to a backup power source or energy storage system for power supply.

[0073] A further optimization scheme predicts the charging power demand of the charging station based on the fusion neural network model and monitors the power demand of the charging station in real time, including:

[0074] Use historical charging data and machine learning algorithms to predict the charging demand of electric vehicles and obtain prediction results;

[0075] The prediction results are used to guide the charging and discharging strategies of the energy storage system and the operation and scheduling of the charging stations.

[0076] Further optimization scheme, the process of dynamically adjusting the charging and discharging state of the energy storage system through fuzzy logic control includes:

[0077] Real-time monitoring of the status of the energy storage system, including but not limited to key parameters such as battery state of charge, charge and discharge current, and voltage;

[0078] According to the characteristics of the energy storage system and the charging and discharging requirements, fuzzy control rules are designed to deal with the uncertainty and nonlinear problems in the charging and discharging process;

[0079] The monitored system state parameters are converted into fuzzy values, and the input fuzzy values ​​are logically inferred according to the fuzzy control rules to determine the charging and discharging strategy;

[0080] The result of fuzzy reasoning is then converted into a specific numerical value to determine the specific charging and discharging power or current setting, and the charging and discharging state of the energy storage system is dynamically adjusted according to the defuzzified control instructions;

[0081] The system status after execution is used as feedback and compared with the expected target to further correct the fuzzy control rules and optimize the charging and discharging process.

[0082] Further optimization scheme, based on the charging demand forecast of electric vehicles and the current state of the energy storage system, the process of formulating charging plans through particle swarm optimization algorithm for refined energy management includes:

[0083] Construct an optimization model, take the charging demand forecast of electric vehicles and the current state of the energy storage system as input, minimize the total cost or maximize the energy utilization efficiency as the objective function, and consider constraints such as power balance and charging and discharging rate limit, randomly initialize the position and speed of the particle swarm, and each particle represents a potential charging and discharging plan;

[0084] Calculate the fitness value of each particle, based on the objective function of total cost or energy efficiency, update the individual extreme value and global extreme value of the particle according to the fitness value, adjust the position and speed of the particle, simulate the foraging behavior of the bird flock, and find the optimal solution;

[0085] By repeatedly iterating and updating the particle position and velocity, the optimal charge and discharge plan is gradually approached. The iteration stop condition is set when the maximum number of iterations or the quality of the solution meets the preset threshold. After the stop condition is met, the optimal charge and discharge plan is extracted from the particle swarm.

[0086] Apply the optimal charging and discharging plan to the actual energy storage system, evaluate the performance and make real-time adjustments based on feedback.

[0087] The control method of the present invention can realize efficient, stable and intelligent operation of the DC microgrid system of the supercharging station, while providing users with a high-quality charging service experience.

[0088] The DC microgrid system control method for supercharging stations of the present invention can not only improve the operating efficiency and stability of the charging station, but also provide users with safe, reliable and efficient charging services. At the same time, the method also has good adaptability, scalability and economy, and can meet the development trend of future microgrid and electric vehicle charging technology.

[0089] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A control method for a DC microgrid system of a supercharging station, It is characterized in that include: Model the DC microgrid system of the supercharging station and obtain system parameters; Constructing an initial deep learning network model, and training the initial deep learning network model based on the system parameters to obtain a target deep learning network model; Based on the target deep learning network model, the input ratio of different energy sources is dynamically adjusted according to the real-time energy supply and charging demand, and energy storage devices are introduced to balance the supply and demand differences; Through wireless communication technology, the charging status, energy production and consumption of the DC microgrid system of the supercharging station can be remotely monitored and managed in real time to optimize energy distribution.

2. The supercharging station DC microgrid system control method according to claim 1, It is characterized in that The process of modeling the DC microgrid system of the supercharging station and obtaining the system parameters includes: By installing sensors and data acquisition equipment at various key nodes of the DC microgrid system of the supercharging station, the DC microgrid of the supercharging station is monitored in real time to obtain the required parameter data; and the collected required parameter data is transmitted to the central control unit via a wireless or wired network; wherein the required parameter data includes but is not limited to the voltage, current, power demand and supply of the DC bus, the charging and discharging status of the energy storage system, the remaining power of the battery, the charging and discharging rate and temperature.

3. The supercharging station DC microgrid system control method according to claim 1, It is characterized in that The process of building the initial deep learning network model includes: A long short-term memory network model and a convolutional neural network model are constructed, and the long short-term memory network model and the convolutional neural network model are fused to obtain a fused neural network model.

4. The supercharging station DC microgrid system control method according to claim 3, It is characterized in that The process of fusing the long short-term memory network model with the convolutional neural network model includes: Standardize the required parameter data to eliminate the dimensional effects between different features and obtain time series data; Performing feature extraction on the time series data through a convolutional neural network model, adding a pooling layer after the convolution layer, and automatically extracting spatial features of the time series data using the convolution layer and the pooling layer to obtain a feature map; Downsampling is performed in the spatial dimension through the pooling layer, and the pooled feature map is converted into a one-dimensional vector and input into the LSTM layer. The gating mechanism of LSTM is used to capture the dynamic characteristics in the time series. Add one or more fully connected layers after the LSTM layer to output the final prediction results; After selecting appropriate activation functions and loss functions according to task requirements, a fusion layer is designed to integrate the local features of CNN and the global temporal features of LSTM to obtain a fusion neural network model; The fusion neural network model is trained using a back propagation algorithm, and the prediction error is minimized by adjusting the network parameters.

5. The supercharging station DC microgrid system control method according to claim 1, It is characterized in that Based on the target deep learning network model, the process of dynamically adjusting the input ratio of different energy sources according to the real-time energy supply and charging demand includes: Predict the charging power demand of the charging station based on the fusion neural network model and monitor the power demand of the charging station in real time; The prediction results are compared with the power demand of the real-time monitoring charging station. When it is detected that the power demand exceeds the supply, the energy storage system is automatically started to discharge and fill the power gap; otherwise, charging is carried out; according to the real-time monitoring data and power demand prediction, the charging and discharging state of the energy storage system is dynamically adjusted through fuzzy logic control; According to the charging demand forecast of electric vehicles and the current status of the energy storage system, a charging plan is formulated through the particle swarm optimization algorithm to carry out refined energy management, giving priority to using electricity during off-peak hours for charging and reducing the charging load during peak hours; When a grid fault is detected, it automatically switches to a backup power source or energy storage system for power supply.

6. The supercharging station DC microgrid system control method according to claim 5, It is characterized in that The process of predicting the charging power demand of the charging station based on the fusion neural network model and monitoring the power demand of the charging station in real time includes: Use historical charging data and machine learning algorithms to predict the charging demand of electric vehicles and obtain prediction results; The prediction results are used to guide the charging and discharging strategy of the energy storage system and the operation scheduling of the charging station.

7. The supercharging station DC microgrid system control method according to claim 5, It is characterized in that The process of dynamically adjusting the charging and discharging state of the energy storage system through fuzzy logic control includes: Real-time monitoring of the status of the energy storage system, including but not limited to key parameters such as battery state of charge, charge and discharge current, and voltage; According to the characteristics of the energy storage system and the charging and discharging requirements, fuzzy control rules are designed to deal with the uncertainty and nonlinear problems in the charging and discharging process; Convert the monitored system state parameters into fuzzy values, perform logical reasoning on the input fuzzy values ​​according to the fuzzy control rules, and determine the charging and discharging strategy; The result of fuzzy reasoning is then converted into a specific numerical value to determine the specific charging and discharging power or current setting, and the charging and discharging state of the energy storage system is dynamically adjusted according to the defuzzified control instructions; The system status after execution is used as feedback and compared with the expected target to further correct the fuzzy control rules and optimize the charging and discharging process.

8. The supercharging station DC microgrid system control method according to claim 5, It is characterized in that Based on the prediction of electric vehicle charging demand and the current status of the energy storage system, the process of developing a charging plan through the particle swarm optimization algorithm for refined energy management includes: Construct an optimization model, take the charging demand forecast of electric vehicles and the current state of the energy storage system as input, minimize the total cost or maximize the energy utilization efficiency as the objective function, and consider constraints such as power balance and charging and discharging rate limit, randomly initialize the position and speed of the particle swarm, and each particle represents a potential charging and discharging plan; Calculate the fitness value of each particle, based on the objective function of total cost or energy efficiency, update the individual extreme value and global extreme value of the particle according to the fitness value, adjust the position and speed of the particle, simulate the foraging behavior of the bird flock, and find the optimal solution; By repeatedly iterating and updating the particle position and velocity, the optimal charge and discharge plan is gradually approached. The iteration stop condition is set when the maximum number of iterations or the quality of the solution meets the preset threshold. After the stop condition is met, the optimal charge and discharge plan is extracted from the particle swarm. The optimal charging and discharging plan is applied to the actual energy storage system, and the performance is evaluated and adjusted in real time based on feedback.

Images