Full-flexible power distribution and switching charging pile

Abstract

The invention discloses a full-flexible power distribution and switching charging pile, and relates to the technical field of charging piles, and a full-flexible power distribution method is applied to the full-flexible power distribution and switching charging pile. The full-flexible power distribution method comprises the following steps: step 1, collecting voltage and current of a charging pile in real time through a sensor network to obtain temperature data, performing trial update on a digital twin model in combination with a power grid load and a user demand, performing simulation prediction by using the digital twin model, optimizing a power distribution strategy, and obtaining a power distribution result; potential faults are pre-judged in advance and active maintenance is realized; 2, each charging terminal is provided with an edge calculation module; through cooperative control based on the digital twinborn technology and edge calculation, the system can monitor and predict the charging demand and the power grid load in real time and dynamically adjust the power distribution strategy, the charging efficiency and the power grid stability are remarkably improved, and the problem of resource waste or power grid overload caused by fixed power of a traditional charging pile is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of charging piles, and in particular to a fully flexible power distribution and switching charging pile. Background Art

[0002] With the popularization of electric vehicles, the contradiction between their charging demand and grid load has become increasingly prominent.

[0003] For example, a distributed charging pile power scheduling method and a distributed charging pile disclosed in publication number CN118438914A, adjacent charging piles send power resource information to each other through communication lines; when the current charging pile starts to charge the power-consuming device, it obtains the current charging demand of the power-consuming device and allocates the output capacity of the current charging pile based on the current charging demand; when the output capacity of the current charging pile is less than the current charging demand, the current charging pile searches for a first target charging pile that meets the preset conditions among the charging piles adjacent to the current charging pile according to the power resource information; if there is a first target charging pile that meets the preset conditions, the current charging pile sends a first power control instruction to the target charging pile, so that the first target charging pile switches the corresponding output capacity to the current charging pile according to the instruction of the first power control instruction. This application method does not require the reconstruction of new high-power charging piles, and realizes the transformation and upgrading of existing charging stations that have been built.

[0004] However, there are many problems with traditional charging piles. First, the power is fixed and cannot be flexibly adjusted according to the actual charging demand and the state of the power grid, resulting in resource waste or power grid overload. Second, the traditional power module has low energy conversion efficiency, large energy loss, large equipment size, high construction and operation costs, and is difficult to meet the growing demand for electric vehicle charging and the requirements for stable operation of the power grid, and is not convenient to use. Therefore, we propose a fully flexible power distribution and switching charging pile to solve the above technical problems. Summary of the invention

[0005] The object of the present invention is to provide a fully flexible power distribution and switching charging stack to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention adopts the following technical solution: the fully flexible power distribution method is applied to the fully flexible power distribution switching charging pile, and the fully flexible power distribution method includes the following steps:

[0007] Step 1: Use the sensor network to collect the voltage, current and temperature data of the charging pile in real time, and try to update the digital twin model based on the grid load and user needs. Use the digital twin model for simulation prediction, optimize the power allocation strategy, predict potential failures in advance and implement proactive maintenance.

[0008] Step 2: Equip each charging terminal with an edge computing module to process local data in real time and execute control commands. Through collaborative communication between edge computing nodes, dynamic power balance and fault isolation within the charging stack can be achieved.

[0009] Step 3: Build the state space, action space and reward function of the charging environment, train the reinforcement learning model, and dynamically adjust the charging power and sequence according to the real-time grid load and electricity price fluctuation information;

[0010] Step 4: Collect the voltage, current and temperature data of the battery, train the AI ​​model to predict SOH, and dynamically adjust the charging current and voltage based on the SOH prediction results to avoid overcharging or over-discharging the battery;

[0011] Step 5: Deploy 5G communication modules to achieve high-speed data transmission between the charging pile and the cloud and user terminals, support remote monitoring, fault diagnosis and software upgrades, and improve the maintainability of the system;

[0012] In the step 1, the power allocation strategy is optimized based on the simulation results of the digital twin model, and the power allocation strategy optimization formula is:

[0013] Among them, P alloc is the optimized power allocation scheme, P i is the allocated power of the i-th charging terminal, P i，req is the required power of the i-th charging terminal, λ is the weight coefficient, which is used to balance power distribution and load balancing, and Load_Imbalance(P) is the load imbalance function, which measures the balance of power distribution;

[0014] In step three, the state space, action space and reward function of the component charging environment are constructed to train a reinforcement learning model. The training formula of the reinforcement learning model is:

[0015] Wherein, Q(s, a) is the value function of executing action α in state s, r is the instant reward of charging efficiency and user satisfaction, γ is the discount factor, s' is the next state, α' is the next action. In step 3, the charging power and sequence are dynamically adjusted according to the real-time grid load and electricity price fluctuations. The formula for dynamically adjusting the charging power is: Where P charge is the optimized charging power, α is the adjusted charging power P i Or charging sequence O i action.

[0016] As a preferred solution, in the step 1, more specifically, the voltage, current and temperature data of the charging stack are collected in real time through the sensor network, and the collected data are uploaded to the digital twin platform. Based on the topological structure and operating parameters of the physical charging stack, a digital twin model is constructed, and the digital twin model is used for simulation to predict future charging demand and grid load, and the power allocation strategy of the charging stack is dynamically adjusted according to the simulation results;

[0017] The digital twin model update formula is: t+1 =f(M t , S t , U t , G t ), where M t is the state of the digital twin model at time t, S t is the voltage V at time t t 、Current I t and temperature T t Sensor data, U t is the charging power P at time t user And the charging time P user User needs, G t is the grid power P at time t grid and the grid frequency f grid The grid load, f(M t , S t , U t , G t ) is the update function of the digital twin model.

[0018] As a preferred solution, in step 2, the edge computing module of each charging terminal processes local data in real time, and the edge computing nodes exchange data through the communication network to achieve collaborative control, and dynamically adjust the output power of each charging terminal according to the collaborative communication results. When a fault is detected, the faulty node is automatically isolated to ensure stable operation;

[0019] The edge extreme module of each charging terminal executes the control command according to the local data. The edge calculation formula is expressed as: C i =g(S i , M i ), where C i is the charging power P of the i-th charging terminal i And the charging state S i The control command, S i is the local sensor data of the i-th charging terminal, M i is the local digital twin model state of the i-th charging terminal, g(S i , M i ) is the decision function of the edge computing module.

[0020] As a preferred solution, in step 2, dynamic power balance is achieved through cooperative communication between edge computing nodes. The dynamic power balance formula is: Where P balance is the total power balance value of the charging stack, P i is the output power of the i-th charging terminal, P j，loss is the jth power loss.

[0021] As a preferred solution, in step four, the health state of the battery, i.e., SOH, is predicted by an AI model, and the prediction model formula is: SOH=h(V, I, T), wherein SOH is the health state of the battery, and its health state output value is 0 to 1, wherein 1 indicates complete health, V is the battery voltage, I is the battery current, T is the battery temperature, and h(V, I, T) is the SOH prediction model.

[0022] As a preferred solution, in step 4, it is also necessary to dynamically adjust the charging current and voltage according to the SOH prediction result. The process is expressed as: charge =k1·SOH,V charge =k2·SOH, where L charge is the optimized charging current, V charge is the optimized charging voltage, k1 and k2 are proportional coefficients.

[0023] As a preferred solution, in step five, the 5G communication module supports high-speed data transmission, and realizes remote monitoring and fault diagnosis through 5G communication.

[0024] As a preferred solution, it includes a sensor unit, which includes a voltage sensor, a current sensor, a temperature sensor, a humidity sensor and a battery status detector;

[0025] A power conversion unit, the power conversion unit comprising an AD / DC converter and a DC / DC converter;

[0026] An edge computing node unit, comprising an ARM processor, an SSD memory and a communication interface component;

[0027] A digital twin platform hardware unit, which includes a high-performance server for running a digital twin model, a GPU accelerator card for supporting large-scale data simulation and machine learning calculations, and a data storage device for storage;

[0028] A communication module unit, wherein the communication module unit includes a 5G communication module;

[0029] A user interaction unit, the user interaction unit comprising at least one touch screen;

[0030] A distributed energy interface unit, wherein the distributed energy interface unit includes a photovoltaic inverter, a wind power converter and an energy management system (EMS).

[0031] As a preferred solution, the communication interface components include Ethernet, Wi-Fi and 5G modules.

[0032] Compared with the prior art, the present invention has obvious advantages and beneficial effects. Specifically, it can be seen from the above technical solution that it mainly has the following advantages:

[0033] The present invention can achieve a comprehensive improvement in efficiency, flexibility, safety and economy of charging infrastructure. Through collaborative control based on digital twin technology and edge computing, the system can monitor and predict charging demand and grid load in real time, dynamically adjust power allocation strategy, significantly improve charging efficiency and grid stability, and avoid resource waste or grid overload caused by fixed power of traditional charging piles; by introducing reinforcement learning and AI algorithms, the system can adaptively optimize charging strategy, which not only prolongs the service life of electric vehicle batteries, but also realizes peak shaving and valley filling, reducing users' charging costs; and adopts high-efficiency power modules with wide bandgap semiconductor technology, which greatly improves energy conversion efficiency, reduces energy loss, and reduces equipment size, reducing construction and operation costs; more importantly, through two-way energy interaction, electric vehicles become distributed energy storage units of the power grid, further enhancing the flexibility of the power grid and the absorption capacity of renewable energy; through 5G / 6G ultra-low latency communication and quantum key distribution technology, the system realizes high-speed and secure remote control and data transmission, ensuring user privacy and data security.

[0034] In order to more clearly illustrate the structural features and effects of the present invention, the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a schematic flow chart of a fully flexible power allocation method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and implementation examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0037] It should be noted that when an element is referred to as being "fixed to" another element, it may be directly on the other element or there may be a central element. When an element is considered to be "connected to" another element, it may be directly connected to the other element or there may be a central element at the same time. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are for illustrative purposes only.

[0038] See also Figure 1 The embodiment of the present invention provides a fully flexible power distribution and switching charging stack. The fully flexible power distribution method is applied to the fully flexible power distribution switching charging stack. The fully flexible power distribution method includes the following steps:

[0039] Step 1: Use the sensor network to collect the voltage, current and temperature data of the charging pile in real time, and try to update the digital twin model in combination with the grid load and user needs. Use the digital twin model for simulation prediction, optimize the power allocation strategy, predict potential failures in advance and implement proactive maintenance; introduce digital twin technology, construct a virtual mirror model of the charging pile, and map the operating status of the physical system in real time;

[0040] Step 2: Equip each charging terminal with an edge computing module to process local data in real time and execute control commands. Through collaborative communication between edge computing nodes, dynamic power balance and fault isolation within the charging pile can be achieved. This reduces dependence on cloud computing, improves the real-time and robustness of the system, and thus adapts to large-scale distributed charging scenarios.

[0041] Step 3: Build the state space, action space and reward function of the charging environment, train the reinforcement learning model, and dynamically adjust the charging power and sequence according to the real-time grid load and electricity price fluctuation information; this can improve the economy of the charging pile and user satisfaction, realize the peak and valley shaving of the grid load, and have a more balanced effect;

[0042] Step 4: Collect the voltage, current and temperature data of the battery, train the AI ​​model to predict SOH, and dynamically adjust the charging current and voltage according to the SOH prediction results to avoid overcharging or over-discharging the battery; this step can improve the battery life and reduce the user's cost of use;

[0043] Step 5: Deploy 5G communication modules to achieve high-speed data transmission between the charging stack and the cloud and user terminals, support remote monitoring, fault diagnosis and software upgrades, and improve the maintainability of the system; this step can improve the response speed and user experience of the charging stack to meet the needs of modern smart power grids.

[0044] In the step 1, more specifically, the voltage, current and temperature data of the charging pile are collected in real time through the sensor network, and the collected data are uploaded to the digital twin platform. Based on the topological structure and operating parameters of the physical charging pile, a digital twin model is constructed, and the digital twin model is used for simulation to predict future charging demand and grid load, and the power allocation strategy of the charging pile is dynamically adjusted according to the simulation results;

[0045] The digital twin model update formula is: t+1 =f(M t , S t , U t , G t ), where M t is the state of the digital twin model at time t, S t is the voltage V at time t t 、Current I t and temperature T t Sensor data, U t is the charging power P at time t user And charging time T user User needs, G t is the grid power P at time t grid and the grid frequency f grid The grid load, f(M t , S t , U t , G t ) is the update function of the digital twin model.

[0046] The power allocation strategy is optimized based on the simulation results of the digital twin model. The power allocation strategy optimization formula is:

[0047] Among them, P alloc is the optimized power allocation scheme, P i is the allocated power of the i-th charging terminal, P i，req is the required power of the i-th charging terminal, λ is the weight coefficient, which is used to balance power distribution and load balancing, and Load_Imbalance(P) is the load imbalance function, which measures the balance of power distribution.

[0048] In step 2, the edge computing module of each charging terminal processes local data in real time, and the edge computing nodes exchange data through the communication network to achieve collaborative control. According to the collaborative communication results, the output power of each charging terminal is dynamically adjusted. When a fault is detected, the faulty node is automatically isolated to ensure stable operation.

[0049] The edge extreme module of each charging terminal executes the control command according to the local data. The edge calculation formula is expressed as: C i =g(S i , M i ), where C i is the charging power P of the i-th charging terminal i And the charging state S i The control command, S i is the local sensor data of the i-th charging terminal, M i is the local digital twin model state of the i-th charging terminal, g(S i , M i ) is the decision function of the edge computing module;

[0050] Through the collaborative communication between edge computing nodes, dynamic power balance is achieved. The dynamic power balance formula is: Where P balance is the total power balance value of the charging stack, P i is the output power of the i-th charging terminal, P j，loss For the jth power loss, deploying edge computing nodes in the charging stack can realize localized intelligent decision-making and distributed collaborative control, thereby reducing extreme reliance on the cloud, improving the real-time and robustness of the system, and adapting to large-scale distributed charging scenarios.

[0051] In step three, the state space, action space and reward function of the component charging environment are constructed to train a reinforcement learning model. The training formula of the reinforcement learning model is:

[0052] Wherein, Q(s, a) is the value function of executing action α in state s, r is the instant reward of charging efficiency and user satisfaction, γ is the discount factor, s' is the next state, α' is the next action. In step 3, the charging power and sequence are dynamically adjusted according to the real-time grid load and electricity price fluctuations. The formula for dynamically adjusting the charging power is: Where P charge is the optimized charging power, α is the adjusted charging power P i Or charging sequence O i This step realizes the adaptive optimization of the charging strategy through the reinforcement learning algorithm.

[0053] In the step 4, the health state of the battery, i.e., SOH, is predicted by the AI ​​model, and the prediction model formula is: SOH=h(V, I, T), where SOH is the health state of the battery, and its health state output value is 0 to 1, where 1 indicates complete health, V is the battery voltage, I is the battery current, T is the battery temperature, and h(V, I, T) is the SOH prediction model, which is a neural network that can reduce energy loss, reduce equipment size, improve convenient performance, and improve the overall performance of the charging stack;

[0054] In step 4, it is also necessary to dynamically adjust the charging current and voltage according to the SOH prediction result. The process is expressed as: charge =k1·SOH,V charge = k2·SOH, where I charge is the optimized charging current, V charge is the optimized charging voltage, k1 and k2 are proportional coefficients.

[0055] In step five, the 5G communication module supports high-speed data transmission and realizes remote monitoring and fault diagnosis through 5G communication.

[0056] A fully flexible power distribution and switching charging stack, comprising a sensor unit, wherein the sensor unit comprises a voltage sensor for detecting voltage, a current sensor for detecting charging stack current, a temperature sensor for detecting charging stack temperature, a humidity sensor for detecting charging stack humidity, and a battery status detector for detecting battery status;

[0057] A power conversion unit, the power conversion unit comprising an AD / DC converter and a DC / DC converter;

[0058] An edge computing node unit, the edge computing node unit comprising an ARM processor for data processing, an SSD memory for storing data, and a communication interface component, the communication interface component comprising Ethernet, Wi-Fi, and 5G modules;

[0059] A digital twin platform hardware unit, comprising a high-performance server for running a digital twin model, a GPU accelerator card for supporting large-scale data simulation and machine learning calculations, and a data storage device for storage;

[0060] A communication module unit, wherein the communication module unit includes a 5G communication module;

[0061] A user interaction unit, the user interaction unit includes at least one touch screen, and further includes a QR code scanner for user payment, and also has a microphone and a speaker for language interaction with the user;

[0062] A distributed energy interface unit, wherein the distributed energy interface unit includes a photovoltaic inverter, a wind power converter and an energy management system (EMS).

[0063] In summary, the present invention can achieve comprehensive improvements in efficiency, flexibility, safety and economy of charging infrastructure;

[0064] Specifically, based on the collaborative control of digital twin technology and edge computing, the system can monitor and predict charging demand and grid load in real time, dynamically adjust the power allocation strategy, significantly improve charging efficiency and grid stability, and avoid resource waste or grid overload caused by fixed power of traditional charging piles;

[0065] Secondly, by introducing reinforcement learning and AI algorithms, the system can adaptively optimize charging strategies, which not only prolongs the service life of electric vehicle batteries, but also achieves peak load shifting and valley filling, reducing users' charging costs. In addition, the high-efficiency power module using wide bandgap semiconductor (SiC / GaN) technology has greatly improved energy conversion efficiency, reduced energy loss, and reduced equipment size, reducing construction and operating costs;

[0066] More importantly, this solution supports V2G (Vehicle-to-Grid) bidirectional energy interaction, making electric vehicles a distributed energy storage unit of the power grid, further enhancing the flexibility of the grid and the ability to absorb renewable energy;

[0067] Through 5G / 6G ultra-low latency communication and quantum key distribution technology, the system achieves high-speed and secure remote control and data transmission, ensuring user privacy and data security.

[0068] 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 and improvements made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A fully flexible power distribution, characterized in that: The invention comprises a fully flexible power distribution method, which is applied to a fully flexible power distribution switching charging pile, and the fully flexible power distribution method comprises the following steps: Step 1: Use the sensor network to collect the voltage, current and temperature data of the charging pile in real time, and try to update the digital twin model based on the grid load and user needs. Use the digital twin model for simulation prediction, optimize the power allocation strategy, predict potential failures in advance and implement proactive maintenance. Step 2: Equip each charging terminal with an edge computing module to process local data in real time and execute control commands. Through collaborative communication between edge computing nodes, dynamic power balance and fault isolation within the charging stack can be achieved. Step 3: Build the state space, action space and reward function of the charging environment, train the reinforcement learning model, and dynamically adjust the charging power and sequence according to the real-time grid load and electricity price fluctuation information; Step 4: Collect the voltage, current and temperature data of the battery, train the AI ​​model to predict SOH, and dynamically adjust the charging current and voltage based on the SOH prediction results to avoid overcharging or over-discharging the battery; Step 5: Deploy 5G communication modules to achieve high-speed data transmission between the charging pile and the cloud and user terminals, support remote monitoring, fault diagnosis and software upgrades, and improve the maintainability of the system; In the step 1, the power allocation strategy is optimized based on the simulation results of the digital twin model, and the power allocation strategy optimization formula is: Among them, P alloc is the optimized power allocation scheme, P i is the allocated power of the i-th charging terminal, P i，req is the required power of the i-th charging terminal, λ is the weight coefficient, which is used to balance power distribution and load balancing, and Load_Imbalance(P) is the load imbalance function, which measures the balance of power distribution; In step three, the state space, action space and reward function of the component charging environment are constructed to train a reinforcement learning model. The training formula of the reinforcement learning model is: Wherein, Q(s, a) is the value function of executing action α in state s, r is the instant reward of charging efficiency and user satisfaction, γ is the discount factor, s′ is the next state, α′ is the next action. In step 3, the charging power and sequence are dynamically adjusted according to the real-time grid load and electricity price fluctuations. The formula for dynamically adjusting the charging power is: Where P charge is the optimized charging power, α is the adjusted charging power P i Or charging sequence O i action.

2. A fully flexible power distribution according to claim 1, characterized in that: In the step 1, more specifically, the voltage, current and temperature data of the charging pile are collected in real time through the sensor network, and the collected data are uploaded to the digital twin platform. Based on the topological structure and operating parameters of the physical charging pile, a digital twin model is constructed, and the digital twin model is used for simulation to predict future charging demand and grid load, and the power allocation strategy of the charging pile is dynamically adjusted according to the simulation results; The digital twin model update formula is: t+1 =f(M t , S t , U t , G t ), where M t is the state of the digital twin model at time t, S t is the voltage V at time t t 、Current I t and temperature T t Sensor data, U t is the charging power P at time t user And charging time T user User needs, G t is the grid power P at time t grid and the grid frequency f grid The grid load, f(M t , S t , U t , G t ) is the update function of the digital twin model.

3. The fully flexible power distribution according to claim 1, characterized in that: In step 2, the edge computing module of each charging terminal processes local data in real time, and the edge computing nodes exchange data through the communication network to achieve collaborative control. According to the collaborative communication results, the output power of each charging terminal is dynamically adjusted. When a fault is detected, the faulty node is automatically isolated to ensure stable operation. The edge extreme module of each charging terminal executes the control command according to the local data. The edge calculation formula is expressed as: C i =g(S i , M i ), where C i is the charging power P of the i-th charging terminal i And the charging state S i The control command, S i is the local sensor data of the i-th charging terminal, M i is the local digital twin model state of the i-th charging terminal, g(S i , M i ) is the decision function of the edge computing module.

4. A fully flexible power distribution according to claim 3, characterized in that: In step 2, dynamic power balance is achieved through collaborative communication between edge computing nodes. The dynamic power balance formula is: Where P balance is the total power balance value of the charging stack, P i is the output power of the i-th charging terminal, P j，loss is the jth power loss.

5. The fully flexible power distribution according to claim 1, characterized in that: In step 4, the AI ​​model is used to predict the battery's state of health, i.e., SOH, and the prediction model formula is: SOH=h(V, I, T), where SOH is the battery's state of health, V is the battery voltage, I is the battery current, T is the battery temperature, and h(V, I, T) is the SOH prediction model.

6. A fully flexible power distribution according to claim 5, characterized in that: In step 4, it is also necessary to dynamically adjust the charging current and voltage according to the SOH prediction result. The process is expressed as: charge =k1·SOH,V charge = k2·SOH, where I charge is the optimized charging current, V charge is the optimized charging voltage, k1 and k2 are proportional coefficients.

7. The fully flexible power distribution according to claim 1, characterized in that: In step five, the 5G communication module supports high-speed data transmission and realizes remote monitoring and fault diagnosis through 5G communication.

8. A fully flexible power distribution switching charging stack, applied to a fully flexible power distribution according to any one of claims 1 to 7, characterized in that: The sensor unit includes a voltage sensor, a current sensor, a temperature sensor, a humidity sensor and a battery status detector; A power conversion unit, the power conversion unit comprising an AD / DC converter and a DC / DC converter; An edge computing node unit, comprising an ARM processor, an SSD memory and a communication interface component; A digital twin platform hardware unit, which includes a high-performance server for running a digital twin model, a GPU accelerator card for supporting large-scale data simulation and machine learning calculations, and a data storage device for storage; A communication module unit, wherein the communication module unit includes a 5G communication module; A user interaction unit, the user interaction unit comprising at least one touch screen; A distributed energy interface unit, wherein the distributed energy interface unit includes a photovoltaic inverter, a wind power converter and an energy management system (EMS).

9. A fully flexible power distribution switching charging stack according to claim 8, characterized in that: The communication interface components include Ethernet, Wi-Fi and 5G modules.

Images