Unified control method for intelligent charging station based on electric Hongyu distributed soft bus

Abstract

The invention discloses an intelligent charging station unified control method based on an electric Hongyu distributed soft bus, relates to the technical field of charging station control, and realizes efficient response and resource optimization configuration of power grid fluctuation by constructing a charging station digital twin and fusing the electric Hongyu distributed soft bus. A predictive scheduling algorithm based on digital twinning can sense instantaneous fluctuation of a power grid in real time, an emergency optimization model is started immediately, energy storage discharging power and charging pile output limitation are adjusted dynamically, and millisecond-level response is achieved. The electric red-mounted distributed soft bus synchronizes the state of global equipment in a microsecond-level time delay mode, it is ensured that digital twin accurately maps the dynamic change of a physical field station, the single-point bottleneck of traditional centralized communication is broken through, and the reliability is improved; meanwhile, the prediction algorithm generates an optimization decision instruction through global modeling, the optimization decision instruction is issued to each device for execution through a soft bus, it is ensured that the charging pile and the energy storage unit strictly follow optimal solution operation, and the problems of extensive resource allocation and scheduling lag in a traditional scheme are effectively solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of charging station control, and particularly relates to a unified control method for an intelligent charging station based on a distributed soft bus of electric Hong. BACKGROUND

[0002] The existing charging station control system based on the electric Hong soft bus realizes plug-and-play and unified management of the equipment, but the control strategy is mostly reactive or rule-based, for example, adjustment is only performed after receiving a power grid dispatching instruction, or time-of-use electricity pricing is performed according to the current simple load condition; this control mode has a lag and cannot cope with transient fluctuations of the power grid, and it is also difficult to optimally plan and allocate resources for complex charging demands in the station that have time-space coupling characteristics. SUMMARY

[0003] Therefore, the present application provides a unified control method for an intelligent charging station based on a distributed soft bus of electric Hong, which can effectively solve the defects that the existing technology cannot cope with transient fluctuations of the power grid and cannot optimally plan and allocate resources for charging demands.

[0004] The technical scheme of the present application is implemented as follows:

[0005] A unified control method for an intelligent charging station based on a distributed soft bus of electric Hong, specifically comprising the following steps:

[0006] Constructing a digital twin of the charging station based on static attributes of physical equipment of the charging station and real-time states of the physical equipment of the charging station that are continuously synchronized through the distributed soft bus of electric Hong;

[0007] Running a predictive scheduling algorithm based on the digital twin to obtain an optimal decision of the charging station;

[0008] Downlinking the optimal decision instruction to the physical equipment of the charging station through the distributed soft bus of electric Hong for execution, to realize global collaborative control of the charging station.

[0009] As a further optional solution of the unified control method for an intelligent charging station based on a distributed soft bus of electric Hong, the step of running a predictive scheduling algorithm based on the digital twin to obtain an optimal decision of the charging station specifically comprises the following steps:

[0010] Obtaining real-time states of the physical equipment of the charging station;

[0011] Performing load prediction and energy side prediction according to the real-time states of the physical equipment of the charging station to obtain load prediction results and energy side prediction results;

[0012] Based on load forecasting results and energy-side forecasting results, a mixed-integer linear programming model is constructed to solve the problem with the objective function of maximizing the operating revenue of power plants or minimizing the grid impact.

[0013] The optimal power allocation plan for the future time span is generated based on the solution results.

[0014] As a further optional solution to the unified control method for intelligent charging stations based on the Dianhong distributed soft bus, the load prediction based on the real-time status of the physical equipment in the charging station specifically includes:

[0015] Based on the real-time status of physical equipment at charging stations, historical data, time characteristics, real-time characteristics, and external event characteristics are acquired.

[0016] The hybrid model predicts charging demand curves for charging stations over a future period based on historical data, time features, real-time features, and external event features. The hybrid model includes an STL decomposition model, an LSTM model, and a LightGBM model. The STL decomposition model is used to decompose historical load sequences into trend terms, seasonal terms, and residual terms. The LSTM model is used to capture complex periodic patterns of the seasonal terms. The LightGBM model is used to predict the residual terms.

[0017] As a further optional solution to the unified control method for intelligent charging stations based on the Dianhong distributed soft bus, the step of predicting energy levels based on the real-time status of the physical equipment at the charging station specifically includes:

[0018] Based on the real-time status of the physical equipment at charging stations, obtain weather forecast data and grid electricity price data;

[0019] Photovoltaic power generation forecasts are made based on weather forecast data to obtain a photovoltaic power generation prediction sequence;

[0020] Based on grid electricity price data, grid electricity prices are predicted to obtain a series of electricity price levels.

[0021] As a further optional scheme of the unified control method for intelligent charging stations based on the Dianhong distributed soft bus, the method, based on load forecasting results and energy-side forecasting results, constructs a mixed-integer linear programming model to solve the problem with the objective function of maximizing station operating revenue or minimizing grid impact. Specifically, this includes:

[0022] The objective function is to maximize the operating revenue of the power station or minimize the grid impact. The objective function to maximize the operating revenue of the power station comprehensively considers the revenue from charging services, photovoltaic power generation, and the revenue from interaction with the grid. The objective function to minimize the grid impact considers the degree of influence of the power station power fluctuation on the grid frequency and voltage.

[0023] Based on the load forecast results, energy-side forecast results, and various preset constraints, a mixed-integer linear programming model is constructed, wherein the constraints include user constraints, power grid constraints, equipment constraints, and spatiotemporal constraints.

[0024] As a further optional scheme of the unified control method for intelligent charging stations based on the Dianhong distributed soft bus, the step of generating the optimal power allocation plan for the future time span based on the solution results specifically includes:

[0025] The mixed-integer linear programming model is solved using a solver to obtain the model solution results;

[0026] Based on the solution results, an optimal power allocation plan for the charging station is generated over a future time span. The optimal power allocation plan includes the charging power of each charging pile in the charging station at each time point, the charging and discharging power of the energy storage system, and the power exchange power with the power grid.

[0027] As a further optional solution to the unified control method for intelligent charging stations based on the Elec-Hong distributed soft bus, the step of issuing optimization decision commands to the physical devices of the charging station via the Elec-Hong distributed soft bus to achieve full-domain collaborative control of the charging station specifically includes:

[0028] Establish communication connections between the charging station control center and various physical devices using the Dianhong distributed soft bus;

[0029] The optimization decision command is sent to the physical equipment in the charging station through the distributed soft bus of the Elec-Hong;

[0030] Each physical device receives and executes the optimization decision command, and adjusts its own operating status according to the command requirements to achieve full-domain collaborative control of the charging station.

[0031] A unified control system for intelligent charging stations based on the Dianhong distributed soft bus includes:

[0032] The digital twin construction module is used to construct a digital twin of the charging station based on the static attributes of the charging station's physical equipment and the real-time status of the charging station's physical equipment continuously synchronized through the Dianhong distributed soft bus.

[0033] The predictive decision module is used to run predictive scheduling algorithms based on digital twins to obtain optimized decisions for charging stations.

[0034] The execution coordination module is used to send optimization decision instructions to the physical equipment of the charging station for execution via the Elec-Power distributed soft bus, so as to realize the whole-domain coordinated control of the charging station.

[0035] A computing device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the unified control method for intelligent charging stations based on the E-Hong distributed soft bus described above.

[0036] A computer-readable storage medium storing a computer program, wherein when executed by a processor, the computer program implements the steps of the unified control method for intelligent charging stations based on the E-Hong distributed soft bus described above.

[0037] The beneficial effects of this invention are as follows: By constructing a digital twin of the charging station and running a predictive scheduling algorithm based on the digital twin, a dual-drive mode of predictive forecasting plus real-time correction is achieved. When the power grid experiences instantaneous fluctuations, an emergency optimization model can be immediately activated to dynamically adjust the energy storage discharge power and charging pile output limits, achieving millisecond-level response. At the same time, the distributed soft bus achieves real-time synchronization of the status of all devices with microsecond-level latency, ensuring that the digital twin can accurately map the instantaneous changes of the physical station. Compared with the traditional centralized communication architecture, the distributed bus eliminates single-point communication bottlenecks and solves the defects of existing technologies that cannot cope with instantaneous fluctuations in the power grid. In addition, by running a predictive scheduling algorithm based on the digital twin, the optimization decision of the charging station is obtained, and the optimization decision command is issued to the physical equipment of the charging station for execution, ensuring that each charging pile and energy storage unit strictly follows the global optimal solution, thus solving the defects of existing technologies that make it difficult to optimally plan and allocate resources for charging demand. Attached Figure Description

[0038] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Fig. 1 This is a flowchart illustrating a unified control method for intelligent charging stations based on the Dianhong distributed soft bus according to the present invention.

[0040] Fig. 2 This is a schematic diagram of the composition of a unified control system for intelligent charging stations based on the Dianhong distributed soft bus according to the present invention.

[0041] Fig. 3 This is a schematic diagram of the composition of a computing device according to the present invention. Detailed Implementation

[0042] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. 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.

[0043] refer to Figs. 1 to 3 A unified control method for intelligent charging stations based on the Dianhong distributed soft bus, specifically including:

[0044] A digital twin of the charging station is constructed based on the static attributes of the physical equipment at the charging station and the real-time status of the physical equipment at the charging station that is continuously synchronized through the Dianhong distributed soft bus.

[0045] Specifically, based on the static attributes of the physical equipment in the charging station (such as equipment power, location information, etc.) and the real-time status of the physical equipment in the charging station (such as battery SOC status, equipment temperature, real-time power, etc.) continuously synchronized through the Elec-Tech distributed soft bus, a high-fidelity digital twin of the charging station is built on the station controller or in the cloud. This digital twin not only reflects the current status of the equipment, but also has the ability to predict the future status.

[0046] Based on the predictive scheduling algorithm of digital twin operation, the optimal decision for charging stations is obtained, specifically including:

[0047] Obtain the real-time status of physical equipment at charging stations;

[0048] Based on the real-time status of the physical equipment at the charging station, load forecasting and energy-side forecasting are performed to obtain load forecasting results and energy-side forecasting results.

[0049] Based on load forecasting results and energy-side forecasting results, a mixed-integer linear programming model is constructed to solve the problem with the objective function of maximizing the operating revenue of power plants or minimizing the grid impact.

[0050] The optimal power allocation plan for the future time span is generated based on the solution results.

[0051] Specifically, by acquiring the real-time status of physical equipment (such as charging pile power, battery SOC, photovoltaic output, etc.), the digital twin can dynamically map the operation of the power station, ensuring that optimization decisions are based on the latest data and avoiding scheduling deviations caused by information lag; combined with load forecasting and energy-side forecasting (photovoltaic power generation, electricity price fluctuations, etc.), power allocation strategies can be planned in advance to effectively cope with instantaneous fluctuations in the power grid (such as sudden load increases or unstable output of new energy sources) and reduce emergency response costs;

[0052] With the goal of maximizing operational revenue or minimizing grid impact, a mixed-integer linear programming (MILP) model is used to balance electricity purchase costs, battery depreciation costs, and dispatch deviation penalty costs to achieve optimal resource allocation. For example, during periods of low electricity prices, priority is given to charging or energy storage, while profits are generated through V2G electricity sales during peak periods; photovoltaic power generation and charging demand are dynamically coordinated to reduce dependence on the traditional power grid.

[0053] Based on the twin model and optimization results, power is accurately allocated to charging piles, energy storage systems, photovoltaic equipment, etc., to avoid equipment overload or inefficient operation. Considering constraints such as vehicle dwell time and the timing of charging demand, the charging behavior of multiple vehicles is coordinated through mechanisms such as virtual charging queues to ensure both user satisfaction (SOC compliance) and site efficiency.

[0054] In some embodiments, the load forecasting based on the real-time status of the physical equipment at the charging station specifically includes:

[0055] Based on the real-time status of physical equipment at charging stations, historical data, time characteristics, real-time characteristics, and external event characteristics are acquired.

[0056] The hybrid model predicts charging demand curves for charging stations over a future period based on historical data, time features, real-time features, and external event features. The hybrid model includes an STL decomposition model, an LSTM model, and a LightGBM model. The STL decomposition model is used to decompose historical load sequences into trend terms, seasonal terms, and residual terms. The LSTM model is used to capture complex periodic patterns of the seasonal terms. The LightGBM model is used to predict the residual terms.

[0057] Specifically, the STL decomposition model breaks down historical load sequences into trend, seasonal, and residual terms, enabling the model to employ differentiated prediction strategies for different component characteristics. For example: the trend term (reflects long-term changes in charging demand (e.g., the increasing adoption of new energy vehicles), which reduces noise interference through smoothing); the seasonal term (captures daily / weekly / yearly periodic patterns (e.g., morning / evening peak hours, holiday effects), where the LSTM model effectively models complex periodic dependencies through memory units); and the residual term (LightGBM, based on the Gradient Boosting Decision Tree (GBDT) framework, efficiently fits nonlinear residuals, reducing prediction errors). Combining the structured decomposition capabilities of STL, the time-series modeling capabilities of LSTM, and the nonlinear fitting capabilities of LightGBM, the overall prediction error (e.g., MAE, RMSE) is significantly lower than that of a single model, especially under conditions of severe fluctuations or external event interference.

[0058] By accessing real-time status data of charging equipment (such as current charging power and number of connected vehicles) and external event characteristics (such as electricity price adjustments and weather changes), the model can dynamically correct prediction results. For example, when sudden weather causes a decrease in photovoltaic output, the model adjusts charging demand forecasts in real time to match available energy; when grid peak-shaving instructions are issued, the model responds quickly and optimizes the load curve. The model's prediction sensitivity to sudden events (such as a surge in charging demand caused by large-scale events) is improved, avoiding prediction biases caused by data lag in traditional static models.

[0059] STL decomposition preprocessing reduces data complexity, LSTM only needs to process seasonal terms (relative rules), LightGBM focuses on residual fitting, and its overall computational efficiency is better than end-to-end deep learning models; LightGBM supports parallel computing and feature parallelism, making it suitable for edge computing scenarios; LSTM can be adapted to local computing resources at the site through model compression techniques (such as knowledge distillation).

[0060] By combining traditional time series decomposition (STL) methods with deep learning (LSTM) and machine learning (LightGBM), this approach balances model interpretability and predictive performance. It breaks through the limitations of single historical load data, integrates real-time equipment status and external event information, and constructs a comprehensive prediction system.

[0061] In some embodiments, the energy-side prediction based on the real-time status of the physical equipment at the charging station specifically includes:

[0062] Based on the real-time status of the physical equipment at charging stations, obtain weather forecast data and grid electricity price data;

[0063] Photovoltaic power generation forecasts are made based on weather forecast data to obtain a photovoltaic power generation prediction sequence;

[0064] Based on grid electricity price data, grid electricity prices are predicted to obtain a series of electricity price levels.

[0065] Specifically, photovoltaic power generation forecasting involves obtaining high-precision numerical weather prediction (NWP) data from the meteorological bureau's API, including irradiance, cloud cover, temperature, and humidity for the next T hours. First, based on physical parameters such as the tilt angle and efficiency of the photovoltaic panels, the theoretical power generation is calculated from the irradiance. Then, a lightweight neural network (such as an MLP) is used to correct this theoretical value. This network is trained using temperature, cloud cover, and humidity as inputs and the deviation between historical actual power generation and the theoretical value as labels to eliminate factors that physical models cannot cover, such as dust obstruction and equipment aging. The output is a photovoltaic power generation forecast sequence for the next T hours.

[0066] Electricity price forecasting: By acquiring historical time-of-use electricity prices, the day-ahead electricity price announcements issued by the power grid, and the predicted regional load demand, the XGBoost model is used to predict the price range level (such as valley, flat, peak, peak) of electricity prices in future time periods, rather than specific values. This method is more stable and reliable than regression forecasting and outputs the electricity price level sequence for the next T hours.

[0067] Based on real-time operating data of photovoltaic equipment in charging stations (such as module temperature, dirt and shading, and inverter efficiency), the model dynamically corrects the prediction bias of weather forecasts. For example, when equipment overheats abnormally or is covered in dust, the model can reduce the theoretical power generation estimate and avoid over-reliance on idealized weather data. By integrating multi-dimensional weather data such as irradiance, temperature, and cloud cover, and combining it with the equipment's historical output patterns, the model captures the nonlinear power generation characteristics under complex weather conditions (such as intermittent output on cloudy days), improving the granularity of predictions. This significantly reduces prediction errors caused by sudden weather changes (such as sudden cloud shading) or changes in equipment status, thereby improving the reliability of photovoltaic power generation prediction sequences.

[0068] By continuously accessing real-time electricity price data from the power grid and combining it with historical electricity price fluctuation patterns (such as peak-valley periods and price spikes during sudden events), the model can quickly identify changes in electricity price trends. For example, when there is an imbalance between power grid supply and demand, real-time electricity price data drives the model to adjust its predictions in real time. By combining weather data (such as extreme high temperatures increasing air conditioning load) with electricity price predictions, the model can capture the coupling relationship between electricity prices and the external environment. For example, electricity prices may rise due to a surge in demand during hot weather, and the model can predict this in advance and optimize charging strategies. The model can also improve the predictive sensitivity of electricity price tiers, especially during periods of drastic price fluctuations (such as a drop in electricity prices due to a surge in renewable energy generation, or a spike in electricity prices during peak consumption periods), providing accurate data for power station economic decisions.

[0069] Based on photovoltaic power generation forecasts, the charging and discharging strategies of the energy storage system are dynamically adjusted (when the predicted power generation exceeds the demand of the power plant, excess energy is stored first or sold to the grid; when the power generation is insufficient, energy storage or electricity is purchased during low-price periods to supplement the supply and reduce electricity costs). In conjunction with the electricity price forecast results, charging or energy storage is increased during low-price periods (such as the peak photovoltaic generation period or the off-peak of grid load), and charging is reduced or electricity is sold to the grid through V2G (vehicle-to-grid) technology during high-price periods to maximize economic benefits.

[0070] In some embodiments, the construction of a mixed-integer linear programming model based on load forecasting results and energy-side forecasting results, with the objective function of maximizing power station operating revenue or minimizing grid impact, specifically includes:

[0071] The objective function is to maximize the operating revenue of the power station or minimize the grid impact. The objective function to maximize the operating revenue of the power station comprehensively considers the revenue from charging services, photovoltaic power generation, and the revenue from interaction with the grid. The objective function to minimize the grid impact considers the degree of influence of the power station power fluctuation on the grid frequency and voltage.

[0072] Based on the load forecast results, energy-side forecast results, and various preset constraints, a mixed-integer linear programming model is constructed, wherein the constraints include user constraints, power grid constraints, equipment constraints, and spatiotemporal constraints.

[0073] Specifically, a two-stage mixed-integer linear programming (MILP) model is adopted, and the objective function used for optimization is to minimize the total operating cost, as follows:

[0074] ;

[0075] in, The cost of electricity purchase is calculated by subtracting the revenue from selling electricity to the grid (V2G) from the cost of purchasing electricity from the grid. The cost of battery degradation, and the cost of loss introduced by V2G and energy storage battery cycle charging and discharging, can be estimated by simplifying the model using the rainflow counting method. The cost of dispatch deviation penalty represents the penalty for the deviation between the actual dispatch and the dispatch instructions issued by the power grid. , , These are the weighting coefficients for each item, and the sum of the three equals 1;

[0076] The constraints of the objective function include power balance constraints, user satisfaction constraints, equipment operation constraints, transformer capacity constraints, and spatiotemporal constraints, among which:

[0077] Power balance constraints:

[0078]

[0079] The above formula represents that grid power + photovoltaic power + energy storage discharge power = total charging power + energy storage charging power + system losses.

[0080] User satisfaction constraint: For each vehicle i, its final SOC must reach the minimum value preset by the user. The details are as follows:

[0081] ;

[0082] in, The initial charge is dimensionless and represents the initial state of charge of the battery when vehicle i starts charging. This is a known input value that is read by the onboard BMS and reported to the control system via the Dianhong soft bus. The overall charging efficiency is dimensionless (between 0 and 1), representing the efficiency loss throughout the entire process from the output of the charging pile to the final charging of the battery, including charger efficiency, line loss, battery charging and discharging efficiency, etc. Let , kilowatts (kW), represent the average charging power allocated to vehicle i within a discrete time step t. The optimization algorithm optimizes the charging power of all vehicles across all time periods. To find the optimal scheduling scheme; To optimize the time interval, the basic step size for discretizing time in the model is used. For example, if the model uses 15-minute intervals, then... =0.25 hours; The total battery capacity, in kilowatt-hours (kWh), represents the maximum usable capacity of the power battery of vehicle i. It is a known vehicle parameter that can be used to convert the charged energy (kWh) into the change in the percentage of charge (SOC). Total time steps, dimensionless, represents the total number of discrete time intervals covered by the optimization problem. For example, if optimizing the next 24 hours, with each interval being 15 minutes, then... =96;

[0083] Equipment operating constraints: Charging power and energy storage charging and discharging power shall not exceed their rated upper and lower limits;

[0084] Transformer capacity constraint: Total power consumption shall not exceed the upper limit of the transformer capacity of the power station;

[0085] Spatiotemporal constraints include the following:

[0086] Define a "virtual charging queue": Define a binary variable for each vehicle i. This indicates whether charging is in progress during time period t;

[0087] Stay time constraints:

[0088] ;

[0089] Ensure that the total charging time during the vehicle's stay at the station meets the demand;

[0090] Charging continuity constraints: Constraints can be added to make the charging process as continuous as possible, avoiding the impact of frequent start-stop cycles on the battery and power grid;

[0091] For the mixed-integer linear programming model consisting of the above objective function and constraints, a two-stage robust optimization is performed. Specifically, the first stage involves solving the mixed-integer linear programming model based on point prediction values ​​to generate a basic, economically optimal scheduling plan, including a power allocation schedule for each vehicle, energy storage, and photovoltaic system.

[0092] Phase Two (Real-time Scrolling Optimization):

[0093] A model predictive control (MPC) framework is adopted. At the beginning of each rolling optimization window, the prediction model is refreshed with the latest measured data (such as the actual vehicle SOC and the updated weather forecast), and the uncertainty set of the prediction error is taken into account.

[0094] Solve a robust optimization problem with the goal of adjusting the power allocation in the short term so that all constraints are met and the plan is as close as possible to the day-ahead plan of the first phase, even under the worst prediction error.

[0095] The output is the optimal control instruction to be executed immediately for the next time step. After executing all optimal control instructions, the model optimization is complete.

[0096] In some embodiments, generating the optimal power allocation plan for the future time span based on the solution results specifically includes:

[0097] The mixed-integer linear programming model is solved using a solver to obtain the model solution results;

[0098] Based on the solution results, an optimal power allocation plan for the charging station is generated over a future time span. The optimal power allocation plan includes the charging power of each charging pile in the charging station at each time point, the charging and discharging power of the energy storage system, and the power exchange power with the power grid.

[0099] Specifically, after optimization, the MILP model is solved using commercial solvers (such as Gurobi, CPLEX) or open-source solvers (such as SCIP). To improve real-time performance, optimization strategy libraries for different scenarios can be pre-calculated and the optimal strategy can be matched in real time.

[0100] The solver outputs an optimal power allocation plan spanning the future, such as: "Command A1 pile to operate at 30kW from 10:00-12:00, pause from 12:00-13:00, and operate at 50kW from 13:00-14:00." "Command the energy storage system to charge from 10:00-12:00 when electricity prices are low, and discharge from 18:00-20:00 when electricity prices are high." "Predicting a load peak at 14:00, notify vehicles about to arrive in advance to discharge via V2G to participate in peak shaving."

[0101] The process of sending optimization decision commands to the physical devices of the charging station via the distributed soft bus to achieve full-domain collaborative control of the charging station includes:

[0102] Establish communication connections between the charging station control center and various physical devices using the Dianhong distributed soft bus;

[0103] The optimization decision command is sent to the physical equipment in the charging station through the distributed soft bus of the Elec-Hong;

[0104] Each physical device receives and executes the optimization decision command, and adjusts its own operating status according to the command requirements to achieve full-domain collaborative control of the charging station.

[0105] Specifically, the Elec-Hong distributed soft bus adopts a distributed virtual bus mechanism to achieve microsecond-level communication latency between the control center and physical devices, ensuring that optimization decision commands (such as power adjustment and energy storage charging and discharging control) are issued in real time, avoiding device response delays caused by communication delays; through multi-path redundant transmission and dynamic topology optimization, it ensures that commands can still reliably reach the devices in complex electromagnetic environments or local network failures, reducing the risk of control failure caused by communication interruptions;

[0106] The Elec-Tech Soft Bus supports dynamic access and exit of devices. Physical devices such as charging piles, energy storage systems, and photovoltaic inverters can be plugged in and used, and automatically added to the collaborative control network. For example, when adding a new charging pile, there is no need to manually configure network parameters. It can receive and execute global optimization commands in real time. Through the high concurrency processing capability of the distributed bus, the control center can issue commands to hundreds of devices at the same time (such as adjusting charging power and switching energy storage modes), realize the synchronous update of the status of all equipment in the site, and avoid system-level efficiency loss caused by local optimization.

[0107] The distributed soft bus does not rely on a single central node. Local device failures or network fluctuations will not cause the entire site control to be paralyzed. The remaining devices can automatically reorganize the communication path based on the bus protocol to maintain collaborative control capabilities. It supports dynamic changes in the number, type and topology of devices (such as energy storage unit expansion and photovoltaic array adjustment). The control strategy does not need to be reconstructed. Seamless compatibility can be achieved simply by updating the device list through the soft bus.

[0108] Optimization decision commands are transmitted directly to the equipment control layer via the soft bus, and the equipment execution results (such as actual power output and SOC status) are fed back to the control center in real time, forming a closed loop of "decision-issuance-execution-verification" to ensure the accurate implementation of control strategies. When there are grid frequency fluctuations, sudden changes in electricity prices, or sudden charging demands, the control center can issue adjustment commands (such as suspending non-emergency charging or starting energy storage discharge) instantly via the soft bus to achieve millisecond-level response and ensure the economic efficiency and grid friendliness of the site.

[0109] A unified control system for intelligent charging stations based on the Dianhong distributed soft bus includes:

[0110] The digital twin construction module is used to construct a digital twin of the charging station based on the static attributes of the charging station's physical equipment and the real-time status of the charging station's physical equipment continuously synchronized through the Dianhong distributed soft bus.

[0111] The predictive decision module is used to run predictive scheduling algorithms based on digital twins to obtain optimized decisions for charging stations.

[0112] The execution coordination module is used to send optimization decision instructions to the physical equipment of the charging station for execution via the Elec-Power distributed soft bus, so as to realize the whole-domain coordinated control of the charging station.

[0113] A computing device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the unified control method for intelligent charging stations based on the E-Hong distributed soft bus described above.

[0114] A computer-readable storage medium storing a computer program, wherein when executed by a processor, the computer program implements the steps of the unified control method for intelligent charging stations based on the E-Hong distributed soft bus described above.

[0115] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A unified control method for intelligent charging stations based on the Dianhong distributed soft bus, characterized in that, Specifically, it includes: Based on the static attributes of the physical equipment in the charging station and the real-time status of the physical equipment in the charging station that is continuously synchronized through the Dianhong distributed soft bus, a digital twin of the charging station is constructed. Based on the predictive scheduling algorithm of digital twin operation, the optimal decision of charging station is obtained; The optimized decision-making instructions are sent to the physical equipment of the charging station through the distributed soft bus of the electric vehicle, so as to realize the whole-domain collaborative control of the charging station.

2. The unified control method for intelligent charging stations based on the Dianhong distributed soft bus according to claim 1, characterized in that, The predictive scheduling algorithm based on digital twins obtains optimized decisions for charging stations, specifically including: Obtain the real-time status of physical equipment at charging stations; Based on the real-time status of the physical equipment at the charging station, load forecasting and energy-side forecasting are performed to obtain load forecasting results and energy-side forecasting results. Based on load forecasting results and energy-side forecasting results, a mixed-integer linear programming model is constructed to solve the problem with the objective function of maximizing the operating revenue of power plants or minimizing the grid impact. The optimal power allocation plan for the future time span is generated based on the solution results.

3. The unified control method for intelligent charging stations based on the Dianhong distributed soft bus according to claim 2, characterized in that, The load forecasting based on the real-time status of the physical equipment at the charging station specifically includes: Based on the real-time status of physical equipment at charging stations, historical data, time characteristics, real-time characteristics, and external event characteristics are acquired. The hybrid model predicts charging demand curves for charging stations over a future period based on historical data, time features, real-time features, and external event features. The hybrid model includes an STL decomposition model, an LSTM model, and a LightGBM model. The STL decomposition model is used to decompose historical load sequences into trend terms, seasonal terms, and residual terms. The LSTM model is used to capture complex periodic patterns of the seasonal terms. The LightGBM model is used to predict the residual terms.

4. The unified control method for intelligent charging stations based on the Dianhong distributed soft bus according to claim 2, characterized in that, The energy-side prediction based on the real-time status of the physical equipment at the charging station specifically includes: Based on the real-time status of the physical equipment at charging stations, obtain weather forecast data and grid electricity price data; Photovoltaic power generation forecasts are made based on weather forecast data to obtain a photovoltaic power generation prediction sequence; Based on grid electricity price data, grid electricity prices are predicted to obtain a series of electricity price levels.

5. The unified control method for intelligent charging stations based on the Dianhong distributed soft bus according to claim 2, characterized in that, Based on load forecasting and energy-side forecasting results, a mixed-integer linear programming model is constructed to solve the problem, with the objective function of maximizing power station operating revenue or minimizing grid impact. Specifically, this includes: The objective function is to maximize the operating revenue of the power station or minimize the grid impact. The objective function to maximize the operating revenue of the power station comprehensively considers the revenue from charging services, photovoltaic power generation, and the revenue from interaction with the grid. The objective function to minimize the grid impact considers the degree of influence of the power station power fluctuation on the grid frequency and voltage. Based on the load forecast results, energy-side forecast results, and various preset constraints, a mixed-integer linear programming model is constructed, wherein the constraints include user constraints, power grid constraints, equipment constraints, and spatiotemporal constraints.

6. The unified control method for intelligent charging stations based on the Dianhong distributed soft bus according to claim 2, characterized in that, The step of generating the optimal power allocation plan for the future time span based on the solution results specifically includes: The mixed-integer linear programming model is solved using a solver to obtain the model solution results; Based on the solution results, an optimal power allocation plan for the charging station is generated over a future time span. The optimal power allocation plan includes the charging power of each charging pile in the charging station at each time point, the charging and discharging power of the energy storage system, and the power exchange power with the power grid.

7. The unified control method for intelligent charging stations based on the Dianhong distributed soft bus according to claim 1, characterized in that, The process of sending optimization decision commands to the physical devices of the charging station via the distributed soft bus to achieve full-domain collaborative control of the charging station includes: Establish communication connections between the charging station control center and various physical devices using the Dianhong distributed soft bus; The optimization decision command is sent to the physical equipment in the charging station through the distributed soft bus of the Elec-Hong; Each physical device receives and executes the optimization decision command, and adjusts its own operating status according to the command requirements to achieve full-domain collaborative control of the charging station.

8. A unified control system for intelligent charging stations based on the Dianhong distributed soft bus, characterized in that, include: The digital twin construction module is used to construct a digital twin of the charging station based on the static attributes of the charging station's physical equipment and the real-time status of the charging station's physical equipment continuously synchronized through the Dianhong distributed soft bus. The predictive decision module is used to run predictive scheduling algorithms based on digital twins to obtain optimized decisions for charging stations. The execution coordination module is used to send optimization decision instructions to the physical equipment of the charging station for execution via the Elec-Power distributed soft bus, so as to realize the whole-domain coordinated control of the charging station.

9. A computing device, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the unified control method for intelligent charging stations based on the Dianhong distributed soft bus as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the steps of the unified control method for intelligent charging stations based on the distributed soft bus of any one of claims 1-7.

Images