Electric vehicle micro-grid self-adaptive charging and discharging system based on multi-source information fusion and cooperation method

Abstract

The invention belongs to the technical field of electric vehicle self-adaptive discharging and intelligent power grids, and particularly relates to an electric vehicle micro-grid self-adaptive charging and discharging system based on multi-source information fusion and a cooperation method. The charging pile is provided with an electric vehicle end and a corresponding charging pile end; wherein the electric vehicle end is provided with a battery capable of charging and discharging, a main control unit, an environment sensing terminal and a scene architecture switching end; the main control unit is electrically connected with the power battery, the environment sensing terminal and the scene architecture switching end; the charging pile end is provided with a micro-grid cooperative controller, and the micro-grid cooperative controller can carry out data exchange and coordination with the main control unit of the electric vehicle end through a communication interface. The power battery and the main control unit are commercially available products. The environment sensing terminal comprises a dust sensing mechanism and / or a liquid cooling mechanism; the dust sensing mechanism and the liquid cooling mechanism are in signal connection with the main control unit through communication interfaces.

Description

Technical Field

[0001] This invention belongs to the field of adaptive discharge technology for electric vehicles and smart grid technology, specifically relating to an adaptive charging and discharging system and collaborative method for electric vehicle microgrids based on multi-source information fusion. Background Technology

[0002] Existing patents related to electric vehicle discharge technology have the following limitations: Most existing patents focus on single scenarios (such as maintenance or outdoor camping), lacking multi-scenario adaptive capabilities; V2G technology primarily addresses vehicle-to-grid interaction, failing to fully explore the potential of vehicles as the core of mobile microgrids; discharge control largely relies on preset parameters, lacking real-time collaborative optimization based on grid status, user behavior, and battery health; protocol compatibility solutions remain complex and could be further intelligentized; and thermal management is mostly passive, lacking predictive control. Therefore, it is necessary to design a collaborative management method that can meet the needs of multiple scenario switching and also deeply integrate with the electric vehicle as a microgrid for self-discharge. Summary of the Invention

[0003] The purpose of this invention is to address the shortcomings pointed out in the background technology above by providing an adaptive charging and discharging system and collaborative method for electric vehicle microgrids based on multi-source information fusion. It uses a three-mode hardware reconfiguration mode that is not a simple multi-socket design, has online self-learning capability without a preset protocol library, and combines a SOC reservation strategy based on calendar data to predict user behavior. This solves the problems of weak scenario adaptability, protocol fragmentation, and low energy utilization in existing technologies, and has good application and promotion effects.

[0004] To achieve the above objectives, the present invention adopts the following technical solution for an adaptive charging and discharging system for electric vehicle microgrids based on multi-source information fusion:

[0005] This invention relates to an adaptive charging and discharging system for electric vehicle microgrids based on multi-source information fusion, which has an electric vehicle end and a corresponding charging pile end;

[0006] The electric vehicle side includes a battery capable of charging and discharging, a main control unit (MCU), an environmental sensing terminal, and a scene architecture switching terminal; the main control unit (MCU) is electrically connected to the power battery, the environmental sensing terminal, and the scene architecture switching terminal respectively;

[0007] The charging pile has a microgrid co-controller, which can exchange and coordinate data with the main control unit (MCU) of the electric vehicle through a communication interface.

[0008] The power battery and main control unit (MCU) mentioned are both commercially available products.

[0009] The environmental sensing terminal includes a dust sensing mechanism and / or a liquid cooling mechanism; the dust sensing mechanism and the liquid cooling mechanism are respectively connected to the main control unit (MCU) via a communication interface.

[0010] Specifically, the dust sensing mechanism includes a dust sensor, a nitrogen cylinder, and an electronically controlled valve. The dust sensor is installed on the surface of the electric vehicle. The outlet of the nitrogen cylinder leads into the power battery. An electronically controlled valve is installed at the outlet of the nitrogen cylinder, and the electronically controlled valve is connected to the main control unit (MCU) via a signal line. When the dust sensor detects dust exceeding a set threshold, the electronically controlled valve is opened, and the high-pressure nitrogen in the nitrogen cylinder blows nitrogen towards the front of the power battery at a predetermined rate, blowing away the dust and preventing the dust from falling onto the power battery's circuitry, thereby avoiding electrical short circuits, eliminating the risk of explosion, and inhibiting electrochemical corrosion.

[0011] Specifically, the liquid cooling mechanism includes a liquid cooling pump and a temperature sensor. The temperature sensor is placed inside the power battery and can detect the temperature of the power battery in real time. A liquid cooling pipe is also embedded inside the power battery. Both ends of the liquid cooling pipe extend out of the power battery. The part of the liquid cooling pipe extending out of the power battery is equipped with heat dissipation fins. A liquid cooling pump for circulating the coolant inside the liquid cooling pipe is installed on the liquid cooling pipe.

[0012] Furthermore, an electric fan that matches the heat dissipation fins is also installed on the electric vehicle. The electric fan is electrically connected to the main control unit (MCU). When the electric fan is started, it can quickly blow airflow over the heat dissipation fins to accelerate the dissipation of heat.

[0013] The dust sensor described is a Sharp GP2Y1010AU0F laser dust sensor, which can detect PM2.5 and PM10 particles in the air. This sensor uses the principle of laser scattering to measure the concentration of dust in the air.

[0014] The dust sensor is connected to the main control unit (MCU) via I2C or UART serial communication.

[0015] When the ambient temperature is too high, the liquid cooling mechanism activates to improve the system's heat dissipation efficiency.

[0016] The conditions for starting the liquid cooling mechanism must be met:

[0017] ;

[0018] in, The heat required for cooling This refers to the coolant flow rate. This refers to the specific heat capacity of the coolant. This refers to temperature changes.

[0019] The liquid cooling pump is an ilo STRATOS Pico type liquid cooling pump, used for circulating coolant in the liquid cooling system.

[0020] The temperature sensor mentioned is a DS18B20 digital temperature sensor, which monitors the temperature changes of the power battery in real time.

[0021] In the aforementioned scenario architecture switching terminal, three modes can be automatically switched through electrical topology switching:

[0022] The three modes are AC mode, DC mode and V2G mode.

[0023] AC mode (220V / 380V): Output via matrix IGBT combination, compatible with camping appliances / industrial equipment.

[0024] ;

[0025] in, For AC power, AC voltage, For alternating current, The power factor.

[0026] DC mode (48V-1000V): Direct DC output, supporting direct connection of photovoltaic energy storage equipment (avoiding inverter losses).

[0027] ;

[0028] in, DC power DC voltage This is direct current.

[0029] V2G mode: Automatically switches to the grid synchronous phase when connected to the grid, and supports power supply response within 10ms.

[0030] ;

[0031] in, This refers to the feedback power from electric vehicles to the power grid. This is the grid voltage. This refers to the output current of the electric vehicle.

[0032] The collaborative method of the adaptive charging and discharging system for electric vehicle microgrids based on multi-source information fusion in this invention is as follows:

[0033] S1. Configure the dynamic protocol learning engine;

[0034] S2. Utilize battery life optimization algorithms and combine them with real-time data to dynamically adjust battery discharge parameters;

[0035] S3. Utilize a microgrid co-controller to coordinate the flow of electrical energy between electric vehicles and various devices in the microgrid, ensuring the stable operation of the microgrid.

[0036] Specifically, S1. Configure the dynamic protocol learning engine as follows:

[0037] A dynamic protocol learning engine is used to adapt to different electric vehicle discharge protocols, enhancing the system's protocol compatibility. A database of discharge protocols from mainstream electric vehicle brands (such as Tesla and BYD) is pre-trained and stored within the system as its foundation through protocol pre-training.

[0038] When the system receives an unknown electric vehicle, it activates online learning. The dynamic protocol learning engine automatically analyzes the signals sent by the battery management system (BMS) and generates a matching protocol through feature extraction and generative adversarial network (GAN) training. Finally, protocol matching is performed, and the system generates a compatible discharge strategy within 30 seconds and optimizes battery discharge management.

[0039] Furthermore, during the protocol matching process, a matching strategy is generated in real time using a Generative Adversarial Network (GAN). This process employs the following formula:

[0040] , The loss function of GAN, For the discriminator output, The matching strategy generated by the generator. This algorithm improves protocol compatibility from 80% to 99.6%. By enhancing system compatibility, it can identify and adapt to the discharge protocols of different electric vehicles, ensuring the system's versatility and scalability. Through dynamic learning and rapid matching, the system can quickly adjust its operating mode when connecting to electric vehicles of different brands, providing precise discharge management.

[0041] Specifically, S2. Utilizes a battery life optimization algorithm and dynamically adjusts the battery's discharge parameters based on real-time data;

[0042] The battery lifespan optimization algorithm dynamically adjusts the battery's discharge parameters based on real-time data, thereby extending the battery's lifespan. The system predicts battery degradation by establishing a three-dimensional degradation model based on charge / discharge frequency, SOC (state of charge) range, and temperature, and dynamically adjusts the discharge strategy based on the prediction results.

[0043] Based on the battery's real-time SOC and temperature, the discharge rate is optimized to avoid over-discharge under adverse conditions (such as high temperature or low SOC), thereby extending battery life. The system can adjust charging and discharging strategies according to real-time changes in battery state, reducing damage to the battery.

[0044] The battery life optimization algorithm includes:

[0045] Dynamically adjust discharge parameters: Adjust the discharge strategy according to SOC and temperature, using the formula:

[0046] ;

[0047] This refers to the battery discharge power. The temperature decay factor, For maximum power, The current ambient temperature. This represents the maximum temperature threshold.

[0048] Extend battery life and improve battery economy and sustainability through precise battery management algorithms. Maximize battery life by dynamically adjusting discharge strategies and optimizing charge / discharge frequencies to prevent overuse.

[0049] Specifically, S3. The microgrid co-controller coordinates the flow of electrical energy between electric vehicles and various devices in the microgrid to ensure the stable operation of the microgrid.

[0050] The microgrid coordinating controller coordinates the flow of electrical energy between various devices such as electric vehicles, energy storage devices, and photovoltaic systems to ensure the stable operation of the microgrid.

[0051] The microgrid controller monitors the power flow of electric vehicles, energy storage devices, and photovoltaic systems in real time through a hybrid communication protocol of LoRa and PLC.

[0052] Power is dynamically allocated based on device priority to ensure that critical loads (such as medical equipment) receive power first. In the event of a grid failure, electric vehicles will act as power sources to start the microgrid, restore power supply, and continue to provide power to the loads.

[0053] The functions of the microgrid co-controller include:

[0054] Intelligent power allocation: Dynamically allocates power based on load priority and SOC to ensure priority power supply to critical loads, using the formula:

[0055] ;

[0056] in, The power allocation weight is calculated based on the load type and SOC.

[0057] Off-grid black start: In the event of a grid failure, the electric vehicle starts the microgrid to restore power supply, using the following formula:

[0058] ;

[0059] in, For black start power, For the output power of electric vehicles, System efficiency is affected by temperature and load.

[0060] By coordinating various devices within the microgrid (such as electric vehicles, energy storage devices, and photovoltaic systems), the microgrid controller ensures efficient power distribution and management. Through intelligent power distribution and black-start functionality, the microgrid controller can provide a stable power supply when grid faults occur, while ensuring power coordination and safe operation among devices.

[0061] This technical solution aims to build a highly intelligent electric vehicle adaptive discharge and microgrid collaborative system with autonomous decision-making capabilities. Its core technological breakthroughs are reflected in the deep integration and innovation of hardware architecture, algorithm engine and system control.

[0062] The device provided by this invention achieves multiple significant benefits through multi-source information fusion and intelligent collaborative control: In terms of scenario adaptability, by leveraging the collaboration between the reconfigurable interface module and the environmental sensing terminal, it achieves millisecond-level (<5ms) seamless switching between AC / DC / V2G modes, and maintains IP68 protection level in extreme environments such as high temperature, high humidity, and vibration, solving the reliability problem of power supply in multiple scenarios; In terms of protocol compatibility, through the offline-online fusion learning mechanism of the dynamic protocol learning engine, the protocol compatibility is increased from the industry average of 80% to 99.6%, achieving seamless "plug and play" access. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of the charging and discharging system architecture in this invention;

[0064] Figure 2 This is a schematic diagram of the multi-scenario adaptive discharge and extreme environment protection architecture in Embodiment 1 of the present invention;

[0065] Figure 3 This is a schematic diagram of the dynamic protocol learning engine in Embodiment 2 of the present invention;

[0066] Figure 4 This is a schematic diagram of microgrid collaboration and V2G economic operation in Embodiment 3 of the present invention. Detailed Implementation

[0067] See attached document Figure 1-4 The present invention is an adaptive charging and discharging system for electric vehicle microgrids based on multi-source information fusion, which has an electric vehicle end and a corresponding charging pile end;

[0068] The electric vehicle side includes a battery capable of charging and discharging, a main control unit (MCU), an environmental sensing terminal, and a scene architecture switching terminal; the main control unit (MCU) is electrically connected to the power battery, the environmental sensing terminal, and the scene architecture switching terminal respectively;

[0069] The charging pile has a microgrid co-controller, which can exchange and coordinate data with the main control unit (MCU) of the electric vehicle through a communication interface.

[0070] The power battery and main control unit (MCU) mentioned are both commercially available products.

[0071] The environmental sensing terminal includes a dust sensing mechanism and / or a liquid cooling mechanism; the dust sensing mechanism and the liquid cooling mechanism are respectively connected to the main control unit (MCU) via a communication interface.

[0072] Specifically, the dust sensing mechanism includes a dust sensor, a nitrogen cylinder, and an electronically controlled valve. The dust sensor is installed on the surface of the electric vehicle. The outlet of the nitrogen cylinder leads into the power battery. An electronically controlled valve is installed at the outlet of the nitrogen cylinder, and the electronically controlled valve is connected to the main control unit (MCU) via a signal line. When the dust sensor detects dust exceeding a set threshold, the electronically controlled valve is opened, and the high-pressure nitrogen in the nitrogen cylinder blows nitrogen towards the front of the power battery at a predetermined rate, blowing away the dust and preventing the dust from falling onto the power battery's circuitry, thereby avoiding electrical short circuits, eliminating the risk of explosion, and inhibiting electrochemical corrosion.

[0073] Specifically, the liquid cooling mechanism includes a liquid cooling pump and a temperature sensor. The temperature sensor is placed inside the power battery and can detect the temperature of the power battery in real time. A liquid cooling pipe is also embedded inside the power battery. Both ends of the liquid cooling pipe extend out of the power battery. The part of the liquid cooling pipe extending out of the power battery is equipped with heat dissipation fins. A liquid cooling pump for circulating the coolant inside the liquid cooling pipe is installed on the liquid cooling pipe.

[0074] Furthermore, an electric fan that matches the heat dissipation fins is also installed on the electric vehicle. The electric fan is electrically connected to the main control unit (MCU). When the electric fan is started, it can quickly blow airflow over the heat dissipation fins to accelerate the dissipation of heat.

[0075] The dust sensor described is a Sharp GP2Y1010AU0F laser dust sensor, which can detect PM2.5 and PM10 particles in the air. This sensor uses the principle of laser scattering to measure the concentration of dust in the air.

[0076] The dust sensor is connected to the main control unit (MCU) via I2C or UART serial communication.

[0077] When the ambient temperature is too high, the liquid cooling mechanism activates to improve the system's heat dissipation efficiency.

[0078] The conditions for starting the liquid cooling mechanism must be met:

[0079] ;

[0080] in, The heat required for cooling This refers to the coolant flow rate. This refers to the specific heat capacity of the coolant. This refers to temperature changes.

[0081] The liquid cooling pump is an ilo STRATOS Pico type liquid cooling pump, used for circulating coolant in the liquid cooling system.

[0082] The temperature sensor mentioned is a DS18B20 digital temperature sensor, which monitors the temperature changes of the power battery in real time.

[0083] In the aforementioned scenario architecture switching terminal, three modes can be automatically switched through electrical topology switching:

[0084] The three modes are AC mode, DC mode and V2G mode.

[0085] AC mode (220V / 380V): Output via matrix IGBT combination, compatible with camping appliances / industrial equipment.

[0086] ;

[0087] in, For AC power, AC voltage, For alternating current, The power factor.

[0088] DC mode (48V-1000V): Direct DC output, supporting direct connection of photovoltaic energy storage equipment (avoiding inverter losses).

[0089] ;

[0090] in, DC power DC voltage This is direct current.

[0091] V2G mode: Automatically switches to the grid synchronous phase when connected to the grid, and supports power supply response within 10ms.

[0092] ;

[0093] in, This refers to the feedback power from electric vehicles to the power grid. This is the grid voltage. This refers to the output current of the electric vehicle.

[0094] The collaborative method of the adaptive charging and discharging system for electric vehicle microgrids based on multi-source information fusion in this invention is as follows:

[0095] S1. Configure the dynamic protocol learning engine;

[0096] S2. Utilize battery life optimization algorithms and combine them with real-time data to dynamically adjust battery discharge parameters;

[0097] S3. Utilize a microgrid co-controller to coordinate the flow of electrical energy between electric vehicles and various devices in the microgrid, ensuring the stable operation of the microgrid.

[0098] Specifically, S1. Configure the dynamic protocol learning engine as follows:

[0099] A dynamic protocol learning engine is used to adapt to different electric vehicle discharge protocols, enhancing the system's protocol compatibility. A database of discharge protocols from mainstream electric vehicle brands (such as Tesla and BYD) is pre-trained and stored within the system as its foundation through protocol pre-training.

[0100] When the system receives an unknown electric vehicle, it activates online learning. The dynamic protocol learning engine automatically analyzes the signals sent by the battery management system (BMS) and generates a matching protocol through feature extraction and generative adversarial network (GAN) training. Finally, protocol matching is performed, and the system generates a compatible discharge strategy within 30 seconds and optimizes battery discharge management.

[0101] Furthermore, during the protocol matching process, a matching strategy is generated in real time using a Generative Adversarial Network (GAN). This process employs the following formula:

[0102] , The loss function of GAN, For the discriminator output, The matching strategy generated by the generator. This algorithm improves protocol compatibility from 80% to 99.6%. By enhancing system compatibility, it can identify and adapt to the discharge protocols of different electric vehicles, ensuring the system's versatility and scalability. Through dynamic learning and rapid matching, the system can quickly adjust its operating mode when connecting to electric vehicles of different brands, providing precise discharge management.

[0103] The protocol database contains discharge protocols from various electric vehicle brands; here are a few examples:

[0104] Tesla:

[0105] Tesla electric vehicles' discharge protocols primarily include a CAN bus-based Battery Management System (BMS) communication protocol. This protocol supports rapid charge / discharge status reporting, SOC (State of Charge) updates, and temperature control.

[0106] For example, Tesla's BMS protocol automatically adjusts the maximum discharge power based on battery temperature and SOC.

[0107] BYD:

[0108] BYD's electric vehicle discharge protocol also uses the CAN bus for data communication, but BYD's BMS protocol supports higher frequency SOC adjustment and introduces a temperature decay algorithm to cope with battery health management under different ambient temperatures.

[0109] NIO:

[0110] NIO's electric vehicles support V2G (Vehicle-to-Grid) mode for their discharge protocol. This protocol not only supports synchronous switching with the power grid, but also dynamically optimizes charging and discharging strategies based on battery charging and discharging frequency and grid conditions.

[0111] BMW:

[0112] BMW's electric vehicle discharge protocol includes battery state monitoring, charging power control, and SOC prediction, and includes dynamic SOC adjustment and battery degradation models.

[0113] Specifically, S2. Utilizes a battery life optimization algorithm and dynamically adjusts the battery's discharge parameters based on real-time data;

[0114] The battery lifespan optimization algorithm dynamically adjusts the battery's discharge parameters based on real-time data, thereby extending the battery's lifespan. The system predicts battery degradation by establishing a three-dimensional degradation model based on charge / discharge frequency, SOC (state of charge) range, and temperature, and dynamically adjusts the discharge strategy based on the prediction results.

[0115] Based on the battery's real-time SOC and temperature, the discharge rate is optimized to avoid over-discharge under adverse conditions (such as high temperature or low SOC), thereby extending battery life. The system can adjust charging and discharging strategies according to real-time changes in battery state, reducing damage to the battery.

[0116] The battery life optimization algorithm includes:

[0117] Dynamically adjust discharge parameters: Adjust the discharge strategy according to SOC and temperature, using the formula:

[0118] ;

[0119] This refers to the battery discharge power. The temperature decay factor, For maximum power, The current ambient temperature. This represents the maximum temperature threshold.

[0120] Extend battery life and improve battery economy and sustainability through precise battery management algorithms. Maximize battery life by dynamically adjusting discharge strategies and optimizing charge / discharge frequencies to prevent overuse.

[0121] Specifically, S3. The microgrid co-controller coordinates the flow of electrical energy between electric vehicles and various devices in the microgrid to ensure the stable operation of the microgrid.

[0122] The microgrid coordinating controller coordinates the flow of electrical energy between various devices such as electric vehicles, energy storage devices, and photovoltaic systems to ensure the stable operation of the microgrid.

[0123] The microgrid controller monitors the power flow of electric vehicles, energy storage devices, and photovoltaic systems in real time through a hybrid communication protocol of LoRa and PLC.

[0124] Power is dynamically allocated based on device priority to ensure that critical loads (such as medical equipment) receive power first. In the event of a grid failure, electric vehicles will act as power sources to start the microgrid, restore power supply, and continue to provide power to the loads.

[0125] The functions of the microgrid co-controller include:

[0126] Intelligent power allocation: Dynamically allocates power based on load priority and SOC to ensure priority power supply to critical loads, using the formula:

[0127] ;

[0128] in, The power allocation weight is calculated based on the load type and SOC.

[0129] Off-grid black start: In the event of a grid failure, the electric vehicle starts the microgrid to restore power supply, using the following formula:

[0130] ;

[0131] in, For black start power, For the output power of electric vehicles, System efficiency is affected by temperature and load.

[0132] By coordinating various devices within the microgrid (such as electric vehicles, energy storage devices, and photovoltaic systems), the microgrid controller ensures efficient power distribution and management. Through intelligent power distribution and black-start functionality, the microgrid controller can provide a stable power supply when grid faults occur, while ensuring power coordination and safe operation among devices.

[0133] Example:

[0134] The multi-scenario adaptive discharge architecture integrates a reconfigurable connection module, supporting automatic switching between AC mode (220V / 380V), DC mode (48V–1000V), and V2G mode through electrical topology switching, with a switching time of less than 5 milliseconds. Its environmental sensing terminal embeds temperature, humidity, dust, and vibration sensors in the discharge gun head, which can trigger adaptive protection measures such as nitrogen sealing or liquid cooling pressurization in real time. The dynamic protocol learning engine possesses offline pre-training and online self-learning capabilities. Through pulse feature extraction and adversarial training, it improves protocol compatibility from 80% to 99.6%, and dynamically optimizes discharge parameters based on a battery degradation model to extend battery life by 40%.

[0135] The microgrid collaborative controller achieves dynamic networking of "vehicle-storage-photovoltaic" through hybrid communication of LoRa and PLC, supporting intelligent power distribution and off-grid black start; its V2G economic strategy combines grid electricity price API and user calendar data to charge to 90% SOC during off-peak hours (cost of electricity 0.2 yuan per kilowatt-hour) and discharge to 40% SOC during peak hours (revenue of 0.8 yuan per kilowatt-hour), and reserves 60% SOC for scenarios such as camping, significantly improving energy utilization efficiency and economic benefits.

[0136] Example 1:

[0137] Multi-scenario adaptive discharge and extreme environment protection, such as Figure 2 As shown, by using a hybrid topology of magnetic latching relay and SiCMOSFET, electrical isolation and topology reconfiguration of AC, DC and V2G modes can be completed within 5ms after receiving the mode switching command from the main control unit.

[0138] Specific application scenario: When a vehicle enters a campsite and connects to the discharge gun, the environmental sensing terminal detects, through the sensor integrated in the gun head, that the ambient dust concentration exceeds the standard (>50μg / m³) and the temperature rises to 45℃, thereby triggering two levels of protection:

[0139] First, control the nitrogen filling operation of the sealed cavity to ensure that the interface reaches the IP68 protection level;

[0140] At the same time, a signal is sent to the predictive thermal management system to increase the pump pressure of the liquid cooling system by 15% to cope with the decrease in heat dissipation efficiency caused by high temperature.

[0141] At this time, the system automatically identifies the connected camping equipment as an AC 220V load, seamlessly switches to AC mode, and outputs a stable sine wave through matrix IGBTs to power high-power equipment such as air conditioners and induction cookers.

[0142] The entire process requires no user intervention, achieving safe and adaptive discharge in harsh environments.

[0143] Example 2: Dynamic Protocol Learning and Battery Life Optimization Figure 3 As shown;

[0144] When a modified electric vehicle that has not been registered in the protocol library is connected to the system, the dynamic protocol learning engine initiates an online self-learning process.

[0145] First, the protocol learning engine sends a set of standard pulse sequences to the vehicle's battery management system (BMS) for detection. Then, the feature extraction unit performs time-domain and frequency-domain analysis on the BMS's response signal to extract its modulation characteristics and timing rules. Finally, the online learning and matching unit, based on a pre-trained generative adversarial network (GAN) model, generates and verifies a high-confidence communication strategy within 30 seconds, successfully establishing a discharge connection. During discharge, the engine calls a battery life optimization algorithm in real time. This algorithm, based on real-time monitored battery temperature (35℃) and current SOC (75%), queries a three-dimensional degradation model of charge / discharge frequency, SOC range, and temperature to calculate the optimal discharge rate of 0.3C. This parameter is then passed to the power allocation unit to ensure that battery degradation is minimized while meeting power requirements. Real-world testing data shows that after 2000 cycles, the battery pack using this algorithm maintains a state of health (SOH) of over 80%, extending its lifespan by approximately 40% compared to uncontrolled discharge battery packs.

[0146] Example 3: Microgrid Collaboration and Economical Operation of V2G Figure 4 As shown;

[0147] In a microgrid in a residential community, multiple electric vehicles equipped with this device are networked with rooftop photovoltaics and basement energy storage cabinets via a hybrid communication network of LoRa and PLC.

[0148] When the power grid experiences a power outage due to a fault, the coordinated controller immediately executes the off-grid black start protocol:

[0149] First, the electric vehicle with the highest SOC (SOC=85%) is designated as the main power source to establish the voltage and frequency reference;

[0150] Then, the energy storage cabinet and photovoltaic system are connected in sequence to form a stable islanded microgrid.

[0151] The controller flexibly allocates power according to preset priorities (medical equipment > lighting > air conditioning) to ensure continuous power supply to critical loads.

[0152] In normal V2G mode, the controller obtains the grid time-of-use electricity price information for the next 24 hours through the cloud API and synchronizes the user's calendar data to determine if the vehicle has no travel plans for the next day.

[0153] Based on this, the system automatically formulates an economical strategy: charging the vehicle battery to 90% SOC during the off-peak period in the early morning (electricity price 0.2 yuan / kWh); and feeding 40% SOC back to the grid during the peak period in the evening (electricity price 0.8 yuan / kWh).

[0154] This single cycle can generate a net profit of approximately 24 yuan for the user. If the user has a pre-set weekend camping trip, the system will intelligently reserve 60% of the SOC power for camping and mark that period as "power priority" in the calendar, suspending V2G discharge, thus achieving a perfect balance between user needs and economic benefits.

[0155] This technical solution aims to build a highly intelligent electric vehicle adaptive discharge and microgrid collaborative system with autonomous decision-making capabilities. Its core technological breakthroughs are reflected in the deep integration and innovation of hardware architecture, algorithm engine and system control.

[0156] At the hardware level, the system adopts an innovative multi-scenario adaptive discharge architecture. Its core is a reconfigurable interface module, which creatively combines the high current carrying capacity of a magnetic latching relay with the high-frequency characteristics of a silicon carbide MOSFET, forming a unique hybrid topology. This design enables the system to seamlessly and rapidly switch between three operating modes: AC mode (providing two standard outputs, 220V / 50Hz and 380V / 50Hz, with a maximum power of 22kW, compatible with all scenarios from household appliances to industrial equipment), DC mode (supporting a wide range of DC output from 48V to 1000V, particularly suitable for direct connection to photovoltaic energy storage systems, avoiding approximately 5-7% energy loss during traditional inverter processes), and V2G mode (compatible with the IEEE 1547-2018 standard, supporting phase synchronization and power feed to the grid within 10 milliseconds). The measured switching time can be controlled within 4.3 ± 0.5 milliseconds, far superior to the traditional solution's level of over 20 milliseconds. In terms of environmental adaptability, the system integrates a multi-parameter environmental sensing terminal inside the discharge gun head, including a MEMS-based temperature and humidity sensor (measurement range -40℃ to +125℃, accuracy ±0.5℃), a laser dust sensor (detects PM2.5 / PM10 concentration, resolution 1μg / m³), and a triaxial vibration sensor (range ±16g, frequency response 0.5Hz-2kHz). When the system detects an ambient dust concentration exceeding 150μg / m³, it automatically activates the nitrogen filling protection system to maintain the pressure inside the sealed chamber at 1.05-1.1 atmospheres. When the ambient temperature exceeds 45℃, the liquid cooling system increases the circulation rate to 1.5 times the standard value. When a continuous vibration acceleration exceeding 5g is detected, the system gradually reduces the output power according to a preset curve to ensure connection reliability.

[0157] At the algorithm level, the system is equipped with a dynamic protocol learning engine with continuous learning capabilities. This engine not only has a built-in protocol database covering major global brands (including Tesla, BYD, NIO, BMW, Volkswagen, etc.), but also innovatively adopts a deep learning-based protocol recognition method. When an unknown protocol is detected, the system sends a set of low-power probe pulses at a specific frequency (1kHz-1MHz). By analyzing the time and frequency domain characteristics of the response waveform and combining it with a recognition model trained by an adversarial generative network, the system can complete protocol parsing and matching strategy generation within 28±3 seconds. Laboratory and field tests have shown that the engine achieves 99.69% compatibility with 327 different vehicle models. Regarding battery life optimization, the system establishes a multi-physics-based battery degradation model. This model comprehensively considers the three-dimensional coupling effects of charge-discharge cycle count (accurate to the depth and rate of each cycle), SOC operating range (especially the dwell time in the critical 20-80% range), and temperature parameters (including instantaneous temperature and temperature change rate). Through model predictive control algorithms, the system can dynamically optimize discharge parameters, such as automatically limiting peak power in high-temperature environments and avoiding large-current discharge at low SOC. Actual test data shows that it can improve battery cycle life by 38.7-41.2%.

[0158] At the system control level, the microgrid co-controller adopts a heterogeneous communication architecture, combining the LoRaWAN protocol (transmission distance up to 5km, power consumption only 0.5W) and G3-PLC power line carrier technology (data transmission rate up to 300kbps) to build a reliable local communication network. The system supports self-organizing networking of up to 32 nodes (including electric vehicles, energy storage devices, photovoltaic inverters, etc.), features intelligent power allocation based on priority (8 configurable levels), and achieves microgrid black start capability within 3 seconds after a grid fault. The system's V2G economic strategy obtains grid time-of-use electricity price data in real time through a RESTful API interface (updated every minute), and combines this with multi-dimensional data such as user calendar trip schedules and vehicle GPS location information, employing a mixed-integer linear programming algorithm for optimization decisions. Typical application scenarios include: charging the vehicle to 90% SOC during off-peak electricity pricing (approximately RMB 0.2 / kWh) from 0:00 to 6:00 AM; discharging the vehicle to 40% SOC during peak electricity pricing (approximately RMB 0.8 / kWh) from 6:00 PM to 10:00 PM; and reserving at least 60% SOC 12 hours in advance when a user's calendar shows a "camping" plan. Through over six months of field verification, this strategy can generate an average annual V2G revenue of RMB 3200-4500 per electric vehicle.

[0159] Through the aforementioned innovations, this technical solution has achieved significant breakthroughs in key indicators such as response speed <5ms, protocol compatibility >99.6%, battery life protection (improved by approximately 40%), and V2G economics (revenue of 0.8 yuan per kilowatt-hour), providing a complete technical path and commercial solution for the deep participation of electric vehicles in the energy internet.

Claims

1. An adaptive charging and discharging system for electric vehicle microgrids based on multi-source information fusion, characterized in that, It has an electric vehicle terminal and a corresponding charging pile terminal; The electric vehicle side includes a battery capable of charging and discharging, a main control unit, an environmental sensing terminal, and a scene architecture switching terminal; the main control unit is electrically connected to the power battery, the environmental sensing terminal, and the scene architecture switching terminal respectively. The charging pile has a microgrid co-controller, which can exchange and coordinate data with the main control unit of the electric vehicle through a communication interface.

2. The adaptive charging and discharging system for electric vehicle microgrids based on multi-source information fusion according to claim 1, characterized in that, The environmental sensing terminal includes a dust sensing mechanism and / or a liquid cooling mechanism; the dust sensing mechanism and the liquid cooling mechanism are respectively connected to the main control unit via communication interfaces.

3. The adaptive charging and discharging system for electric vehicle microgrids based on multi-source information fusion according to claim 2, characterized in that, The dust sensing mechanism includes a dust sensor, a nitrogen cylinder, and an electronically controlled valve. The dust sensor is installed on the surface of the electric vehicle. The outlet of the nitrogen cylinder leads into the power battery. An electronically controlled valve is installed at the outlet of the nitrogen cylinder, and the electronically controlled valve is connected to the main control unit via a signal line. When the dust sensor detects dust exceeding a set threshold, the electronically controlled valve is opened, and the high-pressure nitrogen in the nitrogen cylinder is blown into the power battery at a predetermined rate to blow away the dust, preventing the dust from falling onto the power battery circuit, thereby avoiding electrical short circuits, eliminating the risk of explosion, and inhibiting electrochemical corrosion.

4. The adaptive charging and discharging system for electric vehicle microgrids based on multi-source information fusion according to claim 2, characterized in that, The liquid cooling mechanism includes a liquid cooling pump and a temperature sensor. The temperature sensor is placed inside the power battery and can detect the temperature of the power battery in real time. A liquid cooling pipe is also embedded inside the power battery. Both ends of the liquid cooling pipe extend out of the power battery. The part of the liquid cooling pipe extending out of the power battery is equipped with heat dissipation fins. A liquid cooling pump is installed on the liquid cooling pipe to circulate the coolant inside the liquid cooling pipe.

5. The adaptive charging and discharging system for electric vehicle microgrids based on multi-source information fusion according to claim 4, characterized in that, The electric vehicle is also equipped with an electric fan that matches the heat dissipation fins. The electric fan is electrically connected to the main control unit. When the electric fan is started, it can quickly blow airflow over the heat dissipation fins to accelerate the dissipation of heat.

6. The adaptive charging and discharging system for electric vehicle microgrids based on multi-source information fusion according to claim 4, characterized in that, The conditions for starting the liquid cooling mechanism must be met: ； in, To cool the required amount of heat, This refers to the coolant flow rate. The specific heat capacity of the coolant. This refers to temperature changes.

7. The adaptive charging and discharging system for electric vehicle microgrids based on multi-source information fusion according to claim 1, characterized in that, In the aforementioned scenario architecture switching terminal, three modes can be automatically switched through electrical topology switching: The three modes are AC mode, DC mode and V2G mode.

8. A collaborative method for an adaptive charging and discharging system of an electric vehicle microgrid based on multi-source information fusion, characterized in that, S1. Configure the dynamic protocol learning engine; S2. Utilize battery life optimization algorithms and combine them with real-time data to dynamically adjust battery discharge parameters; S3. Utilize a microgrid co-controller to coordinate the flow of electrical energy between electric vehicles and various devices in the microgrid, ensuring the stable operation of the microgrid.

9. The collaborative method for an adaptive charging and discharging system of an electric vehicle microgrid based on multi-source information fusion according to claim 8, characterized in that, S1. Configure the dynamic protocol learning engine as follows: The system adapts to different electric vehicle discharge protocols through a dynamic protocol learning engine, enhancing the system's protocol compatibility. It also pre-trains and stores a database of discharge protocols for mainstream electric vehicles as the foundation of the system.

10. The collaborative method for an adaptive charging and discharging system of an electric vehicle microgrid based on multi-source information fusion according to claim 9, characterized in that, If the system receives an unknown electric vehicle, it will activate online learning. The dynamic protocol learning engine will automatically analyze the signals sent by the battery management system, generate a matching protocol through feature extraction and generative adversarial network training; finally, the protocol will be matched, and the system will generate a compatible discharge strategy within a predetermined time and optimize battery discharge management. During the protocol matching process, a matching strategy is generated in real time using a generative adversarial network (GAN). This process uses the following formula: , The loss function of GAN, For the discriminator output, The matching strategy generated by the generator.

11. The collaborative method for an adaptive charging and discharging system of an electric vehicle microgrid based on multi-source information fusion according to claim 8, characterized in that, S2. Utilize battery life optimization algorithms and combine them with real-time data to dynamically adjust battery discharge parameters; The battery lifespan optimization algorithm dynamically adjusts the battery's discharge parameters based on real-time data, thereby extending the battery's lifespan. The system predicts battery degradation by establishing a three-dimensional degradation model based on charge / discharge frequency, state of charge range, and temperature, and dynamically adjusts the discharge strategy according to the prediction results. Based on the battery's real-time SOC and temperature, the discharge rate is optimized to avoid over-discharge under adverse conditions, thereby extending battery life. The system can adjust charging and discharging strategies based on real-time changes in battery status to reduce damage to the battery; The battery life optimization algorithm includes: Dynamically adjust discharge parameters: Adjust the discharge strategy according to SOC and temperature, using the formula: ； This refers to the battery discharge power. It is the temperature decay factor. For maximum power, The current ambient temperature. This represents the maximum temperature threshold.

12. The collaborative method for an adaptive charging and discharging system of an electric vehicle microgrid based on multi-source information fusion according to claim 8, characterized in that, S3. Utilize a microgrid co-controller to coordinate the flow of electrical energy between electric vehicles and various devices in the microgrid, ensuring the stable operation of the microgrid; The microgrid coordinating controller coordinates the flow of electrical energy between various devices in electric vehicles, energy storage devices, and photovoltaic systems to ensure the stable operation of the microgrid; The microgrid controller monitors the power flow of electric vehicles, energy storage devices and photovoltaic systems in real time through a hybrid communication protocol of LoRa and PLC; Power is dynamically allocated according to the priority of the equipment to ensure that critical loads receive power first. In the event of a grid failure, electric vehicles will act as a power source to start the microgrid, restore power supply, and continue to provide power to the loads. The microgrid co-controller enables intelligent power distribution and off-grid black start; Intelligent power allocation works by dynamically adjusting power based on load priority and SOC to ensure that critical loads are supplied with priority, using the following formula: ； in, The power allocation weights are calculated based on load type and SOC; Off-grid black start refers to the process where, in the event of a grid failure, an electric vehicle starts the microgrid to restore power, using the following formula: ； in, Black starting power, For the output power of electric vehicles, System efficiency is affected by temperature and load.

Images