An intelligent battery replacement control method and system based on auxiliary power supply of an electric vehicle

Abstract

The application provides a kind of intelligent battery replacement control method and system based on auxiliary power supply of electric vehicle, belongs to the cross technical field of new energy vehicle and new type power system, its method includes: when electric vehicle enters battery replacement station and establishes high voltage electrical connection, determine to enter battery replacement state, and the high voltage bus of electric vehicle in battery replacement state, the battery pack to be replaced, station energy system and grid interface are constructed as multi-port real-time energy scheduling network;In the duration of battery replacement state, the power distribution equation is solved in real time to determine the real-time power instruction of each energy port, and the power output or input of each energy port is coordinated and controlled according to the real-time power instruction;When battery replacement mechanical operation is completed and the high voltage system of electric vehicle is confirmed to be stable, exit battery replacement state.Solve the problem of large power impact and low energy utilization rate in the prior art, realize the response of second-level power grid, reduce the power impact, improve the energy utilization rate of battery replacement process, and at the same time guarantee the battery replacement safety and battery life.

Description

Technical Field

[0001] This invention relates to the field of new energy management and smart grid intersection technology, and in particular to an intelligent battery swapping control method and system based on electric vehicle auxiliary power supply. Background Technology

[0002] With the promotion of battery swapping for electric vehicles, intelligent energy management of battery swapping stations has become a research hotspot. Existing technologies mainly focus on two directions: one is to optimize the charging strategy of backup battery packs within the station to take advantage of peak and off-peak electricity prices or participate in grid demand response; the other is to schedule vehicles entering the station in advance to smooth the battery swapping load. However, these solutions all share a common technical blind spot: they treat the mechanical operation process of battery swapping—the physical process of removing a depleted battery from the vehicle and installing a fully charged battery—as an uninterrupted period unrelated to energy scheduling.

[0003] During this period (usually lasting 3-10 minutes), the vehicle's high-voltage system is in a special connection state. The vehicle load (such as air conditioning and in-vehicle electrical appliances) is still running, and both the batteries to be removed and installed have a certain charging and discharging capacity. However, their energy flow is completely spontaneous or disordered, which leads to the following problems:

[0004] Energy dispatch timeliness loss: The inability to respond to the grid's real-time adjustment commands (such as primary frequency regulation) within this time window weakens the potential of battery swapping stations as a fast and flexible energy storage resource.

[0005] Risk of power surge superposition: Before and after the battery swapping operation, the connection state between the battery pack and the grid changes abruptly, which can easily be superimposed with the charging and discharging behavior of other batteries in the station, resulting in a large power step surge on the distribution network side.

[0006] Rigid vehicle load power supply mode: In this process, the vehicle load is usually powered by the depleted battery to be removed or the fully charged battery to be installed, without combining the grid status and station resources for coordinated power supply.

[0007] In summary, breaking down the timing barriers between battery swapping operations and energy dispatch, and transforming the mechanical process into a controllable and adjustable energy interaction window, is a technical challenge that urgently needs to be addressed.

[0008] Therefore, this invention proposes an intelligent battery swapping control method and system based on auxiliary power supply for electric vehicles. Summary of the Invention

[0009] This invention provides an intelligent battery swapping control method and system based on auxiliary power supply for electric vehicles, in order to solve the aforementioned technical problems.

[0010] This invention provides an intelligent battery swapping control method based on auxiliary power supply for electric vehicles, comprising:

[0011] Step 1: After the electric vehicle enters the battery swapping station and establishes a high-voltage electrical connection, it is determined that it has entered the battery swapping state, and the high-voltage bus of the electric vehicle in the battery swapping state, the battery pack to be replaced, the station's energy system and the grid interface are constructed into a multi-port real-time energy dispatch network.

[0012] Step 2: Solve the power allocation equation in real time during the duration of the switching state. To determine the real-time power command for each energy port, where, For power command vector, As a vector of real-time status and demand, To normalize the mechanical process time Let be the switching state coupling matrix of the independent variable, and let the matrix elements be... For weighting functions;

[0013] Step 3: Based on the real-time power command, coordinate and control the power output or input of each energy port to achieve continuous energy scheduling synchronized with the operation of the battery swapping machinery;

[0014] Step 4: Once the battery swapping operation is complete and the high-voltage system of the electric vehicle is confirmed to be stable, exit the battery swapping state.

[0015] Preferably, the normalized mechanical process time It is a dimensionless time parameter that maps the absolute time of the battery swapping mechanical operation from start to finish to the interval [0,1].

[0016] Preferably, the weighting function The expression is:

[0017] ;

[0018] ;

[0019] ;

[0020] ;

[0021] ;

[0022] in, For mechanical constraint envelope function, This represents the normalized time point of the k-th critical mechanical separation action; This is the step function after smoothing approximation; The tolerance parameter is related to the duration of the physical process of the k-th action; n is the total number of critical mechanical separation actions; The steady-state gain coefficient is determined by the physical characteristics of port i and the baseline weight of scheduling target j. This is the normalized dynamic contact resistance function; This represents the median value within the range where the connector is in a stable contact state. The resistance change sensitivity coefficient; This is a switching function triggered by a small change in contact resistance signal monitored in real time. This is the dynamic margin function for battery state; The current optimal scheduling reference state of charge is dynamically calculated based on the battery's current state of health (SOH). , These are the shape adjustment parameters; This is the thermal safety boundary function; Real-time temperature rise of battery pack or connection points; , These are the safe temperature rise threshold and the maximum permissible threshold, respectively. , These are the attenuation coefficient and the power exponent, respectively; , These are the dynamic balancing weights, and ; The preset standard health status value.

[0023] Preferably, the process of determining whether an electric vehicle has entered the battery swapping station and a high-voltage electrical connection has been established includes:

[0024] After the vehicle enters the battery swapping station, the battery swapping equipment performs a mechanical docking process that includes at least two independent action sequences. Each action sequence is monitored in real time by an integrated force-displacement-temperature three-parameter fusion sensor to generate a corresponding dynamic physical feature vector.

[0025] Meanwhile, during the mechanical docking process of the battery swapping equipment, the broadband vibration exciter integrated on the locking or docking mechanism is controlled to inject a preset non-stationary broadband excitation signal into the mechanical docking interface. Simultaneously, the multi-dimensional dynamic response signal generated by the propagation of the excitation signal in the mechanical structure is collected by the sensor array arranged on the battery swapping equipment side and the vehicle interface side.

[0026] Based on the collected multidimensional dynamic response signals, the physical identity transfer function feature vector characterizing the dynamic characteristics of the mechanical structure is extracted, and dynamic matching and verification are performed with the pre-stored benchmark physical identity fingerprint map of the corresponding vehicle model.

[0027] When all verification results reach their respective set confidence thresholds, physical identity verification is deemed successful; otherwise, a first warning is issued.

[0028] Once the physical authentication is verified, the dynamic physical feature vector is matched with the pre-stored baseline physical vector of the corresponding vehicle model. When the matching degree reaches the first threshold, the physical connection is determined to be truly established; otherwise, a second warning is issued.

[0029] Preferably, after determining that the physical connection has been truly established, the process further includes:

[0030] An excitation signal containing multiple frequency points is applied to the high-voltage circuit of the electric vehicle, and the response signal of each frequency point is collected simultaneously.

[0031] The collected impedance spectrum data is compared with the historical impedance spectrum fingerprint database associated with the pre-stored VIN code of the corresponding vehicle model, and the stability of the impedance spectrum changes within a set time window is detected in real time.

[0032] The electrical connection is deemed genuine and stable only if the impedance spectrum shape matching degree reaches the second threshold and the spectral deformation coefficient is lower than the set threshold; otherwise, a second warning is issued.

[0033] Preferably, after determining that the electrical connection is genuine and stable, the process further includes:

[0034] A set of dynamically modulated power waveforms is injected into the battery of the electric vehicle, and a first verification code is extracted based on the power waveform detected by the electric vehicle, wherein the power waveform carries the first verification code;

[0035] The electric vehicle injects another set of dynamically modulated current waveforms into the battery swapping station, and the battery swapping station extracts a second verification code from the current waveform, wherein the current waveform carries the second verification code.

[0036] The power circuit and communication channel are deemed to be genuine and reliable only if the verification codes in both directions are successfully matched and the dynamic response characteristics of the power waveform and current waveform are consistent with the pre-stored model; otherwise, a third early warning is issued.

[0037] Preferably, dynamic matching and verification are performed with the pre-stored baseline physical identity fingerprint map of the corresponding vehicle model, including:

[0038] Based on multidimensional dynamic response signals, a multi-scale nonlinear state-space equation describing the dynamic characteristics of mechanical docking structures is established:

[0039] ;

[0040] ;

[0041] in, It is a high-order state vector that includes structural displacement, velocity, and internal state. For the excitation signal vector, For the observed response vector; For a nonlinear time-varying system matrix, its elements include nonlinear function terms for structural stiffness and damping; , , To incentivize positions and observation location Related input / output matrices; and These represent model uncertainty and observation noise, respectively.

[0042] Solving the multi-scale nonlinear state-space equations yields the nonlinear modal characteristic tensor that characterizes the essential dynamic properties of the mechanical structure. ,in, It is a nonlinear function of the norm of the i-th order state vector; Let be a nonlinear function of the i-th order displacement and velocity; Let i be the left nonlinear mode shape vector of order i; Let i be the right nonlinear mode shape vector of order i; Let i be the kernel of the i-th order nonlinear frequency response function; Let m be the i-th order nonlinear coupling operator; m is the order.

[0043] Based on the nonlinear modal feature tensor Extract the feature set to construct the physical identity transfer function feature vector F:

[0044] ,in, These are linear approximate eigenvectors; These are nonlinear eigenvectors; These are topological feature vectors;

[0045] The physical identity transfer function feature vector F is decomposed and quantized to generate the physical identity fingerprint (PIF).

[0046] ,in, To use the mother wavelet Continuous wavelet transform of F is used to extract multi-scale features in the time-frequency domain; The gradient of the physical identity transfer function feature vector F as a function of time t1; The integral is the characteristic loop along the time path C of the connection process; , , These are quantization, hashing, and signature encoding functions, respectively.

[0047] The generated physical identity fingerprint (PIF) is dynamically matched and verified with the pre-stored benchmark physical identity fingerprint map.

[0048] Physical authentication is considered successful if and only if all authentication results reach their respective confidence thresholds.

[0049] This invention provides an intelligent battery swapping control system based on auxiliary power supply for electric vehicles, comprising:

[0050] The status entry module is used to determine that the electric vehicle has entered the battery swapping state after entering the battery swapping station and establishing a high-voltage electrical connection, and to construct the high-voltage bus of the electric vehicle in the battery swapping state, the battery pack to be replaced, the energy system in the station and the grid interface into a multi-port real-time energy dispatch network.

[0051] The instruction determination module is used to solve the power allocation equation in real time during the duration of the battery swapping state. To determine the real-time power command for each energy port, where, For power command vector, As a vector of real-time status and demand, To normalize the mechanical process time Let be the switching state coupling matrix of the independent variable, and let the matrix elements be... For weighting functions;

[0052] The input / output coordination module is used to coordinate and control the power output or input of each energy port according to the real-time power command, so as to realize continuous energy scheduling synchronized with the operation of the battery swapping machinery.

[0053] The status exit module is used to exit the battery swapping state after the battery swapping mechanical operation is completed and the high voltage system of the electric vehicle is confirmed to be stable.

[0054] Compared with the prior art, the beneficial effects of this application are as follows:

[0055] 1. Maximize the value of time resources: Transform the service time of each battery swapping vehicle at the station into a high-value real-time auxiliary service window for the power grid, greatly improving asset utilization and revenue potential.

[0056] 2. Suppressing power surges at the source: By continuously controlling the power at the battery port with soft start and soft stop throughout the entire battery swapping process, sudden power surges when the battery is connected to / disconnected from the grid are avoided, fundamentally improving the grid friendliness of the battery swapping station.

[0057] 3. Enhance the economic efficiency and resilience of system operation: Optimal economic dispatch and optimal grid support can be achieved during the battery swapping process, while providing seamless and high-quality power supply guarantee for vehicle load.

[0058] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0059] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0060] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0061] Figure 1 This is a flowchart of an intelligent battery swapping control method based on auxiliary power supply for electric vehicles, as described in an embodiment of the present invention.

[0062] Figure 2 This is a structural diagram of an intelligent battery swapping control system based on auxiliary power supply for electric vehicles, as described in an embodiment of the present invention. Detailed Implementation

[0063] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0064] This invention provides an intelligent battery swapping control method based on auxiliary power supply for electric vehicles, such as... Figure 1 As shown, it includes:

[0065] Step 1: After the electric vehicle enters the battery swapping station and establishes a high-voltage electrical connection, it is determined that it has entered the battery swapping state, and the high-voltage bus of the electric vehicle in the battery swapping state, the battery pack to be replaced, the station's energy system and the grid interface are constructed into a multi-port real-time energy dispatch network.

[0066] Step 2: Solve the power allocation equation in real time during the duration of the switching state. To determine the real-time power command for each energy port, where, For power command vector, As a vector of real-time status and demand, To normalize the mechanical process time Let be the switching state coupling matrix of the independent variable, and let the matrix elements be... For weighting functions;

[0067] Step 3: Based on the real-time power command, coordinate and control the power output or input of each energy port to achieve continuous energy scheduling synchronized with the operation of the battery swapping machinery;

[0068] Step 4: Once the battery swapping operation is complete and the high-voltage system of the electric vehicle is confirmed to be stable, exit the battery swapping state.

[0069] Preferably, the normalized mechanical process time It is a dimensionless time parameter that maps the absolute time of the battery swapping mechanical operation from start to finish to the interval [0,1].

[0070] In this embodiment, it is assumed that a vehicle enters the station for battery swapping, and the total operation time of the battery swapping machinery is 5 minutes.

[0071] The vehicle is located and high-voltage connection is completed, entering battery swapping mode. The controller identifies the vehicle model and loads the corresponding mechanical process timeline and default matrix. Function set.

[0072] At the instant the battery swap begins (t0, =0), the following data was collected: the power grid frequency is low, the electricity price is at its peak, the SOC of the battery to be removed is 45%, and the vehicle air conditioning load is 4kW.

[0073] The controller is based on real-time and Solve the equations. Assume the solution indicates that the battery to be removed should discharge at -3kW (to the power grid), and the battery to be installed should discharge at -1kW (to the vehicle), together meeting the vehicle load and supporting the power grid.

[0074] The control system executes commands. The energy flow is: battery to be removed (3kW) → vehicle bus → power grid interface; simultaneously, battery to be installed (1kW) → vehicle bus → vehicle load. The robotic arm begins unlocking the battery to be removed.

[0075] With the advancement of mechanical processes ( Increase), such as when When it reaches 0.3 (i.e., 1.5 minutes), According to its function curve, it has decayed to 60% of the initial value. At this time, if the grid demand remains unchanged, the controller will automatically and smoothly reduce the discharge power of the battery to be removed to about 1.8kW, and may increase the output of the battery to be installed or the energy storage in the station to maintain the total power balance.

[0076] when When the value approaches 0.8, the schedulable weight of the battery port to be removed is close to zero, and its power command smoothly returns to zero. The vehicle load is then fully taken over by the securely connected batteries to be installed or by resources within the station.

[0077] Battery swap complete ( =1), and the vehicle drove away. The removed battery's SOC dropped to approximately 43%, and some of its energy had already been effectively utilized for grid service and load power supply during the battery swapping process.

[0078] In this embodiment, when the battery management system of the electric vehicle sends a high-voltage ready signal to the battery swapping station control system, it is determined that the battery swapping state has been entered. This state continues until the fully charged battery pack is installed in place, the locking mechanism is locked, and the vehicle's high-voltage bus voltage is stable at 390V for 3 seconds, at which point it is determined that the battery swapping state has been exited.

[0079] In this embodiment, the multi-port real-time energy dispatch network refers to a network structure in the battery swapping state that consists of the electric vehicle high-voltage bus port, the port of the depleted battery pack to be removed, the port of the fully charged battery pack to be installed, the station's energy system port, and the grid interface port, enabling bidirectional energy flow and real-time dispatch. For example, the multi-port real-time energy dispatch network constructed in the battery swapping state includes 5 ports: the electric vehicle high-voltage bus port (load port, outputting the power demand of loads such as vehicle air conditioning and navigation, e.g., 8kW), the depleted battery pack port (energy port, SOC=30%, output power ≤15kW), the fully charged battery pack port (energy port, SOC=95%, input / output power ≤50kW), the station's energy storage battery port (energy port, SOC=70%, input / output power ≤100kW), and the grid interface port (energy port, absorbable / injectable power ≤200kW). Each port is electrically connected through the bidirectional converter and high-voltage bus switchgear of the battery swapping station, and energy dispatch is achieved through the control system.

[0080] In this embodiment, the power command vector This refers to the set of power control commands for each energy port in a multi-port real-time energy dispatch network at a given time t. Each element in the vector corresponds to a real-time power command for one energy port; positive values ​​represent the port's output power, and negative values ​​represent the port's input power. For example, at t=10 seconds, the power command vector... The elements correspond to the following in sequence: electric vehicle high-voltage bus port (-8kW, input 8kW power to meet load requirements), depleted battery pack port (0kW, no output / input power), fully charged battery pack port (10kW, output 10kW power), station energy storage battery port (-2kW, input 2kW power), and grid interface port (0kW, no absorption / injection of power). This vector indicates that the fully charged battery pack outputs 10kW power at this time, of which 8kW supplies the vehicle load and 2kW charges the station energy storage battery.

[0081] In this embodiment, the real-time state and the demand vector This refers to the set of parameters reflecting the operating status and scheduling requirements of each port in a multi-port real-time energy dispatch network at a specific time t. It includes core parameters such as grid dispatch commands, the status of each battery pack (SOC, SOH), contact resistance, and port temperature rise. For example, at t=10 seconds, the real-time state and demand vector... The elements correspond to the following in sequence: primary frequency regulation demand of the power grid (+5kW, requiring 5kW power injection from the battery swapping station), SOC of the depleted battery pack, SOH of the depleted battery pack, SOC of the fully charged battery pack, SOH of the fully charged battery pack, contact resistance at the connection point between the depleted battery pack and the bus, contact resistance at the connection point between the fully charged battery pack and the bus, temperature rise of the depleted battery pack, temperature rise of the fully charged battery pack, and vehicle load power demand. This vector comprehensively reflects the current power grid demand, battery status, electrical connection status, and load demand.

[0082] In this embodiment, =0 corresponds to the start time of mechanical operation. =1 corresponds to the time when the mechanical operation is completed, and Where t_start is the start time of the mechanical operation and t_end is the end time of the mechanical operation.

[0083] In this embodiment, the swapping state coupling matrix Refers to normalized mechanical process time The matrix is ​​a set of independent variables, with dimensions of: number of energy ports × dimensions of real-time status and demand vectors. Matrix elements... Let be the weighting function, used to characterize the influence weight of the j-th parameter in the real-time state and demand vector on the power command of the i-th energy port. For example, in a multi-port real-time energy dispatch network containing 5 energy ports and 10 parameters in the real-time state and demand vector, the switching state coupling matrix is... For a 5×10 matrix, such as when When =0.5 (mid-stage of mechanical operation, connector in stable contact state), in the matrix =0.2 indicates that the weight of the impact of the depleted battery pack SOC on the port power command of the fully charged battery pack is 0.2.

[0084] In this embodiment, the objective function of the power allocation equation is: ,in, This is the smoothing coefficient, with a value of 0.1. The power command value at time t; This is the optimal reference value for power at time t.

[0085] The beneficial effects of the above technical solution are as follows: by constructing a multi-port real-time energy dispatch network, the mechanical operation process of battery swapping is transformed into a controllable energy dispatch window, realizing the synchronous coordination of battery swapping actions and energy dispatch; by solving the power allocation equation in real time and controlling the power of each port in a closed loop, the power surge during battery connection / disconnection is effectively mitigated, and the grid response capability is improved; at the same time, it provides continuous and stable power supply for vehicle load, maximizes the utilization of energy resources during battery swapping, and improves the operational economy and grid friendliness of the battery swapping station.

[0086] This invention provides an intelligent battery swapping control method based on auxiliary power supply for electric vehicles, wherein the weighting function The expression is:

[0087] ;

[0088] ;

[0089] ;

[0090] ;

[0091] ;

[0092] in, For mechanical constraint envelope function, This represents the normalized time point of the k-th critical mechanical separation action; This is the step function after smoothing approximation; The tolerance parameter is related to the duration of the physical process of the k-th action; n is the total number of critical mechanical separation actions; The steady-state gain coefficient is determined by the physical characteristics of port i and the baseline weight of scheduling target j. This is the normalized dynamic contact resistance function; This represents the median value within the range where the connector is in a stable contact state. The resistance change sensitivity coefficient; This is a switching function triggered by a small change in contact resistance signal monitored in real time. This is the dynamic margin function for battery state; The current optimal scheduling reference state of charge is dynamically calculated based on the battery's current state of health (SOH). , These are the shape adjustment parameters; This is the thermal safety boundary function; Real-time temperature rise of battery pack or connection points; , These are the safe temperature rise threshold and the maximum permissible threshold, respectively. , These are the attenuation coefficient and the power exponent, respectively; , These are the dynamic balancing weights, and ; The preset standard health status value.

[0093] In this embodiment, calibration was performed using battery swapping experiments on 10 different vehicle models (including sedans, SUVs, and commercial vehicles). The statistical mean ± 0.1 error range was used to determine the accuracy of the calibration. The value ranges from 0.1 to 0.5;

[0094] Based on battery SOC-schedule margin characteristic curve fitting, ensure It exhibits a smooth distribution when SOC∈[20%,80%], thus yielding The value range is 2 to 5;

[0095] Referring to the industry standard battery SOH degradation curve, with =90% as the baseline calibration, resulting in The value ranges from 0.8 to 1.2;

[0096] Based on battery pack thermal safety test data, when near hour, The decay rate is controlled to be below 10% / ℃. The value ranges from 0.3 to 0.7;

[0097] Thermal shock experiments verified that the power command decays non-linearly and rapidly under over-temperature conditions, preventing thermal runaway. The value ranges from 1.5 to 2.5;

[0098] against The value is set at the upper limit (0.6-0.8) for high-frequency regulation scenarios of the power grid and at the lower limit (0.4-0.6) for economic dispatch scenarios. ;

[0099] against The value ranges from 0.02 to 0.05, and is converted based on the physical duration (1-3 seconds) of key mechanical actions (unlocking, alignment, locking). interval;

[0100] The value of n ranges from 3 to 5. For conventional battery swapping processes, n=3 (unlocking, alignment, locking) and for complex battery swapping mechanisms, n=5 (adding pre-alignment and re-inspection).

[0101] In this embodiment, the The product form ensures that once any critical mechanical separation action begins, the schedulability of the relevant ports is safely and continuously suppressed to zero.

[0102] In this embodiment, for By incorporating the microscopic physical state of electrical connections into scheduling decisions, power commands are reduced smoothly and in advance when contact resistance begins to increase abnormally, far exceeding the safety design of traditional systems that rely solely on time or position sensors.

[0103] In this embodiment, for For example, for batteries with low SOH, its To reduce stress, we will move closer to 50%. This function achieves nonlinear, adaptive matching between battery dispatchability and its real-time health and state of charge, optimizing economic dispatch while proactively protecting battery life.

[0104] In this embodiment, When the temperature approaches the safety boundary, the function nonlinearly decays the power command, upgrading thermal management from passive alarm protection to proactive and flexible power constraint.

[0105] In this embodiment, , Determined by the current dominant dispatch mode: when responding to high-frequency regulation commands from the power grid, Increase, emphasize based on Its rapid and safe response capabilities; when performing optimized scheduling focused on economy or battery life, Increase, emphasize based on State adaptability.

[0106] In this embodiment, based on dynamic contact resistance Battery health status Adaptive optimal state of charge and real-time temperature rise These three key physical quantities, combined with embedded weights, ensure the realization of energy port schedulability during battery swapping.

[0107] Through the product function and Discrete mechanical action sequences and continuous thermal states are seamlessly embedded into the power decision core in the form of soft constraints, which surpasses the traditional hard cut-off protection logic and achieves performance maximization within the safety boundary.

[0108] In this embodiment, the unlocking action is as follows: =0.2, corresponding to an absolute time of 1-2 minutes, based on a standard battery swapping time of 5 minutes;

[0109] Separation action: =0.5, corresponding to an absolute time of 2.5 minutes;

[0110] Locking action: =0.8, corresponding to an absolute time of 4 minutes.

[0111] During the middle stage of the battery swapping process, the contact resistance is minimal and stable.

[0112] The beneficial effects of the above technical solution are: through and The function design achieves early warning of sub-optimal conditions and proactive power suppression for poor contact and overheating risks, through... The function implements differentiated scheduling strategies for batteries with different state of health (SOH) to match their health status, thereby achieving refined management of battery assets and extending the overall service life of battery packs within the battery swapping system. The design can dynamically balance the grid's rapid response needs with the equipment's own safety and lifespan constraints within the short window of battery swapping, and obtain the globally optimal instantaneous power command, thus comprehensively improving the scheduling quality and economic value of the battery swapping station as a distributed resource.

[0113] This invention provides an intelligent battery swapping control method based on auxiliary power supply for electric vehicles. The process of determining whether an electric vehicle has entered the battery swapping station and established a high-voltage electrical connection includes:

[0114] After the vehicle enters the battery swapping station, the battery swapping equipment performs a mechanical docking process that includes at least two independent action sequences. Each action sequence is monitored in real time by an integrated force-displacement-temperature three-parameter fusion sensor to generate a corresponding dynamic physical feature vector.

[0115] Meanwhile, during the mechanical docking process of the battery swapping equipment, the broadband vibration exciter integrated on the locking or docking mechanism is controlled to inject a preset non-stationary broadband excitation signal into the mechanical docking interface. Simultaneously, the multi-dimensional dynamic response signal generated by the propagation of the excitation signal in the mechanical structure is collected by the sensor array arranged on the battery swapping equipment side and the vehicle interface side.

[0116] Based on the collected multidimensional dynamic response signals, the physical identity transfer function feature vector characterizing the dynamic characteristics of the mechanical structure is extracted, and dynamic matching and verification are performed with the pre-stored benchmark physical identity fingerprint map of the corresponding vehicle model.

[0117] When all verification results reach their respective set confidence thresholds, physical identity verification is deemed successful; otherwise, a first warning is issued.

[0118] Once the physical authentication is verified, the dynamic physical feature vector is matched with the pre-stored baseline physical vector of the corresponding vehicle model. When the matching degree reaches the first threshold, the physical connection is determined to be truly established; otherwise, a second warning is issued.

[0119] In this embodiment, a three-parameter fusion sensor is installed on the docking head of the battery swapping robotic arm. The force sensor is a strain gauge type with a measurement range of 0-500N; the displacement sensor is a laser displacement meter with a measurement range of 0-100mm; and the temperature sensor is an NTC thermistor with a measurement range of -20℃-85℃. During the docking process, the sensors collect data at a sampling frequency of 100Hz. At a certain moment, the force value of 20N, the displacement value of 50mm, and the temperature value of 25℃ are collected. After data fusion, a dynamic physical feature vector [20N, 50mm, 25℃, 0.4N / mm (force-displacement ratio), 0.2℃ / s (temperature rise rate)] is generated, containing original parameters and derived parameters. In this embodiment, an independent action sequence refers to a set of action steps that are independent of each other and have no logical dependence in the battery swapping robotic arm docking process. Each action sequence has clear start and end conditions and is completed by the corresponding actuator. For example, the battery swapping mechanical docking process includes two independent action sequences. Action sequence 1 is: battery latch unlocking: completed by the electromagnetic lock actuator, the starting condition is that the vehicle positioning is qualified, and the ending condition is that the latch is fully unlocked (triggered by the limit switch), and the action duration is 10-15 seconds; Action sequence 2 is: robotic arm alignment: completed by the robotic arm driven by the servo motor, the starting condition is that the vehicle positioning is qualified, and the ending condition is that the alignment error between the robotic arm and the battery pack is ≤±1mm (detected by the laser positioning sensor), and the action duration is 15-20 seconds. The two action sequences can be executed in parallel to improve docking efficiency.

[0120] In this embodiment, the broadband vibration exciter refers to a device integrated on the locking or docking mechanism of the battery swapping equipment, capable of generating a non-stationary broadband vibration signal within a preset frequency range and injecting it into the mechanical docking interface to excite the dynamic response of the mechanical structure in order to extract the physical identity characteristics of the structure. For example, the broadband vibration exciter is installed on the locking mechanism of the battery swapping robotic arm, with a frequency range of 100Hz-10kHz, an amplitude of 0.1g-0.5g (g is the acceleration due to gravity), and a non-stationary random signal type. The exciter consists of a signal generator, a power amplifier, and a vibration motor. The signal generator generates a preset broadband signal, the power amplifier amplifies the signal and drives the vibration motor, and the vibration motor transmits the vibration to the docking interface through the mechanical structure, exciting the structural vibration on the vehicle interface side and the battery swapping equipment side.

[0121] In this embodiment, the sensor array refers to a collection of multiple sensors arranged on the battery swapping equipment side and the vehicle interface side. These sensors are used to synchronously acquire multi-dimensional dynamic response signals generated by the propagation of broadband vibration excitation signals within the mechanical structure. Sensor types include accelerometers and displacement sensors. For example, the sensor array may contain three accelerometers and two displacement sensors. On the battery swapping equipment side, two accelerometers are arranged (mounted on the robotic arm connector and locking mechanism, respectively) and one displacement sensor (mounted on the robotic arm guide rail). On the vehicle interface side, one accelerometer is arranged (mounted on the battery pack interface frame) and one displacement sensor (mounted at the battery pack locking slot). All sensors have a sampling frequency of 1kHz, synchronously acquiring the vibration acceleration and displacement response signals of the mechanical structure to form a multi-dimensional dynamic response signal set.

[0122] In this embodiment, the physical identity transfer function feature vector refers to the set of characteristic parameters extracted based on multidimensional dynamic response signals, which can characterize the dynamic characteristics of the mechanical docking structure (such as natural frequency, damping ratio, and mode shape). For example, based on the multidimensional dynamic response signals collected by the sensor array, the natural frequency (such as 1kHz, 3kHz, 5kHz), damping ratio (such as 0.05, 0.08, 0.10), and mode shape vector (such as [0.1, 0.3, 0.5], [0.2, 0.4, 0.6]) of the structure are extracted through frequency domain analysis (Fourier transform). The peak value (such as 2g, 1.5g) and rise time (such as 0.2s, 0.3s) of the response signal are extracted through time domain analysis. After standardizing these parameters, the physical identity transfer function feature vector [1kHz, 3kHz, 5kHz, 0.05, 0.08, 0.10, 0.1, 0.3, 0.5, 0.2, 0.4, 0.6, 2g, 1.5g, 0.2s, 0.3s] are formed.

[0123] In this embodiment, the benchmark physical identity fingerprint spectrum refers to the standard set of physical identity transfer function feature vectors corresponding to different vehicle models, which are pre-stored in the battery swapping station control system. It serves as the benchmark for physical identity verification and is obtained by collecting feature vectors of the same vehicle model through numerous experiments and statistical analysis. For example, for a certain brand of Model Y vehicle, experiments were conducted on the battery swapping docking process of 100 vehicles of the same model to collect the physical identity transfer function feature vectors of each vehicle. Outliers were removed using a clustering algorithm, and the mean and standard deviation of each feature parameter were calculated to generate the benchmark physical identity fingerprint spectrum for that vehicle model. For example, the benchmark range of natural frequency [0.95kHz-1.05kHz, 2.9kHz-3.1kHz, 4.9kHz-5.1kHz], the benchmark range of damping ratio [0.04-0.06, 0.07-0.09, 0.09-0.11], and the benchmark range of modal shape vector [[0.08-0.12, 0.28-0.32, 0.48-0.52], etc.].

[0124] In this embodiment, dynamic matching verification refers to comparing the physical identity transfer function feature vector extracted in real time with the pre-stored benchmark physical identity fingerprint map to verify whether the similarity between the two reaches the set confidence threshold, so as to determine whether the currently docked mechanical structure matches the vehicle model.

[0125] The first early warning signal is a warning signal issued by the battery swapping station when physical identity verification fails. It is used to alert operators to risks such as unauthorized vehicle access or abnormal mechanical structure. The warning methods include audible and visual alarms, upper computer prompts, and SMS notifications.

[0126] In this embodiment, the reference physical vector refers to a standard set of dynamic physical feature vectors corresponding to different vehicle models, pre-stored in the battery swapping station control system. It serves as the benchmark for physical connection verification and is obtained through extensive experimental collection and statistical analysis of dynamic physical feature vectors of the same vehicle model. For example, for a certain brand of Han EV vehicle, the parameter range of the reference physical vector is: contact force 25N-35N, docking displacement 75mm-85mm, interface temperature 24℃-28℃, force-displacement ratio 0.3N / mm-0.4N / mm, temperature rise rate 0-0.2℃ / s, and displacement change rate 1mm / s-3mm / s.

[0127] In this embodiment, the first threshold refers to the matching degree threshold between the dynamic physical feature vector and the reference physical vector. When the matching degree reaches or exceeds this threshold, it is determined that the physical connection is truly established; otherwise, a second warning reminder is triggered. The first threshold is determined through experimental calibration. For example, if the first threshold for the physical connection matching degree of a certain vehicle model is set to 95%, the matching degree between the real-time dynamic physical feature vector [30N, 80mm, 26℃, 0.375N / mm, 0.1℃ / s, 2mm / s] and the reference physical vector is calculated to be 97% (all parameters are within the reference range and the deviation is small), which exceeds the first threshold of 95%, thus determining that the physical connection is truly established.

[0128] In this embodiment, the second early warning reminder refers to the early warning signal issued by the battery swapping station when the physical connection matching does not reach the first threshold, indicating to the operators that there are problems such as loose physical connection or incomplete connection. The early warning method is similar to the first early warning reminder.

[0129] The beneficial effects of the above technical solution are as follows: By introducing a dual verification mechanism of force-displacement-temperature three-parameter fusion sensing and monitoring, broadband vibration excitation and multi-dimensional dynamic response acquisition into the battery swapping mechanical docking process, the mechanical docking efficiency is improved through the parallel execution of independent action sequences. Furthermore, by leveraging the dynamic matching of physical identity transfer function feature vectors and benchmark fingerprint maps, and the precise comparison of dynamic physical feature vectors and benchmark physical vectors, a comprehensive and highly reliable verification of vehicle physical identity and connection status is achieved. This effectively identifies risks such as unauthorized vehicle access, incomplete mechanical docking, or loose connections. Through tiered early warning reminders, abnormal processes are promptly blocked, ensuring the safety and accuracy of battery swapping status establishment from the source. This lays a solid physical connection foundation for the stable operation of the subsequent multi-port energy dispatching network.

[0130] This invention provides an intelligent battery swapping control method based on auxiliary power supply for electric vehicles, which, after determining that a physical connection has been truly established, further includes:

[0131] An excitation signal containing multiple frequency points is applied to the high-voltage circuit of the electric vehicle, and the response signal of each frequency point is collected simultaneously.

[0132] The collected impedance spectrum data is compared with the historical impedance spectrum fingerprint database associated with the pre-stored VIN code of the corresponding vehicle model, and the stability of the impedance spectrum changes within a set time window is detected in real time.

[0133] The electrical connection is deemed genuine and stable only if the impedance spectrum shape matching degree reaches the second threshold and the spectral deformation coefficient is lower than the set threshold; otherwise, a second warning is issued.

[0134] In this embodiment, the excitation signal with multiple frequencies refers to an electrical signal that covers a specific frequency range and contains multiple discrete frequency values, applied to detect the electrical characteristics of the high-voltage circuit of an electric vehicle. For example, the excitation signal applied to the high-voltage circuit contains 5 frequencies, namely 1kHz, 3kHz, 5kHz, 8kHz, and 10kHz. The signal type is a sine wave, and the amplitude is 5% of the rated voltage of the high-voltage circuit (e.g., when the rated voltage is 390V, the amplitude is 19.5V). The duration of each frequency is 0.2 seconds, and the total signal duration is 1 second.

[0135] In this embodiment, impedance spectrum data refers to a dataset of high-voltage circuit impedance values ​​varying with frequency, calculated based on excitation signals and corresponding response signals at different frequencies. For example, for excitation signals at five frequencies, the collected response currents are 4.875mA, 6.5mA, 3.9mA, 2.4375mA, and 1.95mA, respectively. Based on impedance = voltage / current, the impedances at each frequency are calculated as follows: Composition of impedance spectrum data .

[0136] In this embodiment, the pre-stored historical impedance spectrum fingerprint database associated with the VIN code refers to the set of impedance spectrum data collected during the historical battery swapping of each registered battery swapping vehicle, stored in the server of the battery swapping station, which corresponds one-to-one with the VIN code of that vehicle. This data forms the electrical identity fingerprint of the vehicle's high-voltage circuit, used for subsequent comparison and verification. For example, if a vehicle's VIN code is LVVDC21B8NF000001, its historical impedance spectrum fingerprint database stores impedance spectrum data from three battery swaps. The impedance spectrum of the first battery swap is... The second time was The third time was This reflects the stable electrical characteristics of the vehicle's high-voltage circuit.

[0137] In this embodiment, the waveform matching degree refers to the degree of similarity between the waveform of the impedance spectrum data collected in real time and the data in the historical impedance spectrum fingerprint database.

[0138] The spectral deformation coefficient refers to the statistical value of the change in real-time impedance spectrum data relative to historical impedance spectrum data. It is calculated as the maximum value of the ratio of the absolute value of the impedance difference at each frequency point to the historical impedance value.

[0139] The second threshold refers to the acceptable threshold for spectral matching. When the spectral matching degree of the real-time impedance spectrum reaches or exceeds this threshold, and the spectral deformation coefficient is lower than the set threshold, the electrical connection is deemed genuine and stable. The second threshold is determined through experimental calibration. For example, if the second threshold for spectral matching degree of a certain vehicle model is set to 90% (0.9), and the threshold for the spectral deformation coefficient is set to 5%; the real-time spectral matching degree is 0.97, and the spectral deformation coefficient is 3.3%, both meeting the threshold requirements, the electrical connection is deemed genuine and stable.

[0140] The beneficial effects of the above technical solution are as follows: by using multi-frequency impedance spectrum detection and historical fingerprint database comparison, the electrical connection status of high-voltage circuits can be accurately verified. The dual threshold judgment mechanism of spectrum matching degree and spectrum deformation coefficient can effectively identify potential risks such as electrical loose connection and insulation degradation, avoid energy dispatching failures or safety hazards caused by unstable electrical connection, and further improve the electrical safety and reliability of the battery swapping process.

[0141] This invention provides an intelligent battery swapping control method based on auxiliary power supply for electric vehicles, which, after determining that the electrical connection is real and stable, further includes:

[0142] A set of dynamically modulated power waveforms is injected into the battery of the electric vehicle, and a first verification code is extracted based on the power waveform detected by the electric vehicle, wherein the power waveform carries the first verification code;

[0143] The electric vehicle injects another set of dynamically modulated current waveforms into the battery swapping station, and the battery swapping station extracts a second verification code from the current waveform, wherein the current waveform carries the second verification code.

[0144] The power circuit and communication channel are deemed to be genuine and reliable only if the verification codes in both directions are successfully matched and the dynamic response characteristics of the power waveform and current waveform are consistent with the pre-stored model; otherwise, a third early warning is issued.

[0145] In this embodiment, the dynamically modulated power waveform refers to a dynamic power signal that carries specific verification code information and is modulated by changing the power amplitude, frequency, or phase. It is used by the battery swapping station to transmit verification information to the electric vehicle and at the same time detect the transmission characteristics of the power circuit.

[0146] The first verification code refers to the verification information transmitted by the battery swapping station to the electric vehicle through dynamic modulation of the power waveform. It is a string of binary numbers or character sequences used to verify the reliability of the forward transmission of the power circuit and the vehicle's receiving capability. For example, the first verification code is an 8-bit binary number "11001100".

[0147] Dynamically modulated current waveforms refer to the dynamic current signals injected by electric vehicles into the battery swapping station in response to the station's verification request. These signals, carrying a second verification code, are modulated by changing the current amplitude, frequency, or phase. This is used to verify the reliability of the reverse transmission of the power circuit and the receiving capability of the battery swapping station.

[0148] The second verification code refers to the verification information transmitted by the electric vehicle to the battery swapping station through a dynamically modulated current waveform. It is a string of binary numbers or character sequences, which, together with the first verification code, form a two-way verification to verify the bidirectional reliability of the power circuit and the communication channel. For example, the second verification code is an 8-bit binary number "00110011".

[0149] In this embodiment, the pre-stored model refers to the standard model of the dynamic response characteristics of the power waveform and current waveform pre-stored in the battery swapping station and electric vehicle. It includes parameters such as the amplitude variation range of the waveform, frequency response speed, and allowable phase deviation range, and is used to verify the transmission quality of the actual waveform.

[0150] The third early warning signal is issued by the battery swapping station when the two-way verification code fails to match or the waveform dynamic response characteristics do not conform to the pre-stored model. It indicates to the operators that there is a fault in the power circuit or communication channel. The early warning methods include audible and visual alarms, upper computer prompts, and SMS notifications. The alarm level is higher than the first and second early warnings.

[0151] The beneficial effects of the above technical solution are as follows: through bidirectional verification code transmission and waveform dynamic response characteristic verification, dual reliability verification of power circuit and communication channel is realized. The bidirectional verification mechanism ensures bidirectional smooth energy transmission and signal communication. Pre-stored model comparison further guarantees waveform transmission quality, effectively avoids energy scheduling out of control due to power circuit failure or communication interruption, and provides a stable and reliable hardware and communication foundation for subsequent continuous energy scheduling.

[0152] This invention provides an intelligent battery swapping control method based on auxiliary power supply for electric vehicles, which dynamically matches and verifies the battery swapping with a pre-stored baseline physical identity fingerprint map of the corresponding vehicle model, including:

[0153] Based on multidimensional dynamic response signals, a multi-scale nonlinear state-space equation describing the dynamic characteristics of mechanical docking structures is established:

[0154] ;

[0155] ;

[0156] in, It is a high-order state vector that includes structural displacement, velocity, and internal state. For the excitation signal vector, For the observed response vector; For a nonlinear time-varying system matrix, its elements include nonlinear function terms for structural stiffness and damping; , , To incentivize positions and observation location Related input / output matrices; and These represent model uncertainty and observation noise, respectively.

[0157] Solving the multi-scale nonlinear state-space equations yields the nonlinear modal characteristic tensor that characterizes the essential dynamic properties of the mechanical structure. ,in, It is a nonlinear function of the norm of the i-th order state vector; Let be a nonlinear function of the i-th order displacement and velocity; Let i be the left nonlinear mode shape vector of order i; Let i be the right nonlinear mode shape vector of order i; Let i be the kernel of the i-th order nonlinear frequency response function; Let m be the i-th order nonlinear coupling operator; m is the order.

[0158] Based on the nonlinear modal feature tensor Extract the feature set to construct the physical identity transfer function feature vector F:

[0159] ,in, These are linear approximate eigenvectors; These are nonlinear eigenvectors; These are topological feature vectors;

[0160] The physical identity transfer function feature vector F is decomposed and quantized to generate the physical identity fingerprint (PIF).

[0161] ,in, To use the mother wavelet Continuous wavelet transform of F is used to extract multi-scale features in the time-frequency domain; The gradient of the physical identity transfer function feature vector F as a function of time t1; The integral is the characteristic loop along the time path C of the connection process; , , These are quantization, hashing, and signature encoding functions, respectively.

[0162] The generated physical identity fingerprint (PIF) is dynamically matched and verified with the pre-stored benchmark physical identity fingerprint map.

[0163] Physical authentication is considered successful if and only if all authentication results reach their respective confidence thresholds.

[0164] In this embodiment, the verification includes:

[0165] Wavelet domain correlation test: Calculate the cross-correlation coefficients between wavelet coefficients;

[0166] Loop integral conservation test: verify the deviation between the characteristic loop integral and the reference path integral;

[0167] Gradient manifold consistency test: Verify the Hausdorff distance between the manifold with characteristic gradients in phase space and the reference manifold.

[0168] In this embodiment, the steps for solving the multi-scale nonlinear state-space equations are as follows:

[0169] Initialization parameters:

[0170] higher-order state vector Initial values: based on static detection data of the mechanical structure before the start of battery swapping (displacement=0, velocity=0, internal state=baseline value).

[0171] Wideband vibration excitation signal (100Hz-10kHz, amplitude 0.1g-0.5g).

[0172] Set to Gaussian white noise, variance ;

[0173] Solution method: Fourth-order Runge-Kutta method is used.

[0174] Iteration step size: 0.01s (to ensure that the high-frequency components of the dynamic response signal are not lost);

[0175] Convergence condition: The difference in the norm of the state vectors between two consecutive iterations ≤ ;

[0176] Solution range: t1∈[0,T], where T is the total mechanical docking time, 30-60 seconds;

[0177] Output results: After solving, extract the 6 core components of the nonlinear modal feature tensor. .

[0178] In this embodiment, the Physical Identity Fingerprint (PIF) generation process is as follows:

[0179] Continuous wavelet transform (WT):

[0180] Mother wavelet Select Morlet wavelet, center frequency 5kHz;

[0181] Scale range: 1-8 scales, extracting multi-scale features in the time and frequency domain;

[0182] Quantization function :

[0183] 8-bit binary quantization is used, and the quantization interval is adaptively divided based on the maximum and minimum values ​​of the feature vector F. It is divided into 256 intervals.

[0184] Hash function :

[0185] The SHA-256 algorithm is used, with the time-varying gradient of the feature vector F as the input. Output a 256-bit hash value;

[0186] Signature encoding function :

[0187] The RSA algorithm is used, with the private key stored in the battery swapping station control system and the public key pre-stored in the vehicle battery management system (BMS).

[0188] Characteristic loop integral Along the mechanical docking time path C ( Integrate at [0,1], with an integration step size of 0.01.

[0189] In this embodiment, the physical identity matching confidence threshold is ≥90% (based on experimental data from 100 vehicles of the same model, with a false positive rate of ≤0.5%).

[0190] First threshold for dynamic physical feature vector matching degree: ≥95% (deviation of each parameter ≤±5%)

[0191] Wavelet domain correlation test threshold: ≥0.85 (cross-correlation coefficient between real-time wavelet coefficients and benchmark coefficients);

[0192] Loop integral conservation test threshold: deviation ≤ ±3% (relative error between real-time integral value and benchmark integral value);

[0193] Second threshold for impedance spectrum shape matching: ≥90% (calculated using cosine similarity);

[0194] Spectral deformation coefficient threshold setting: ≤5% (maximum value of impedance change rate at each frequency point);

[0195] High-voltage circuit excitation signal frequencies: 1kHz, 3kHz, 5kHz, 8kHz, 10kHz (covering the main resonant frequencies of the high-voltage circuit);

[0196] Verification code format: 8-bit binary number (the first and second verification codes are pre-stored in the battery swapping station and the vehicle BMS, and are updated synchronously in both directions).

[0197] Waveform dynamic response characteristic thresholds: amplitude deviation ≤ ±10%, phase deviation ≤ ±5° (compared with the pre-stored model);

[0198] Verification timeout: ≤2 seconds (if timeout occurs, a third warning will be triggered).

[0199] The beneficial effects of the above technical solution are as follows: by establishing multi-scale nonlinear state-space equations, the essential dynamic characteristics of the mechanical docking structure are deeply explored, and a comprehensive physical identity fingerprint code containing linear, nonlinear, and topological features is extracted. Combined with three multi-dimensional checks, high-precision and high-reliability verification of physical identity is achieved. This verification method can effectively distinguish the differences between different vehicle models and different mechanical structures, resist external interference and forgery risks, provide solid technical support for physical identity verification in the battery swapping process, and further improve the security and anti-misoperation capabilities of the battery swapping system.

[0200] This invention provides an intelligent battery swapping control method based on auxiliary power supply for electric vehicles. The method constructs a multi-port real-time energy dispatch network comprising the high-voltage busbar of the electric vehicle in the battery swapping state, the battery pack to be replaced, the on-site energy system, and the grid interface.

[0201] The high-voltage busbar of the electric vehicle in the battery swapping state, the battery pack to be replaced, the on-site energy system, and the grid interface are defined as dynamic energy nodes, and the set of all nodes is denoted as . To describe the time-varying topology and coupling strength of energy interactions between nodes, a normalized mechanical process time is constructed. Dynamically coupled tensor with independent variable ,in, The total number of nodes. For coupling feature dimensions; tensor elements Representation at time From node To the node Time-varying connection weights on the m0th coupling feature dimension:

[0202] ,in, It is a process window function that is synchronized with the m0th mechanical sub-process (such as unlocking, lifting, aligning, locking). It is non-zero within the effective time window of the corresponding sub-process, and zero otherwise. For activation functions; , These are the learnable weight matrix and bias vector; , They are nodes , At any moment The real-time state vector includes voltage, current, power, SOC, temperature, and health status; The difference between the state vectors of the two nodes;

[0203] Based on the dynamic coupling tensor A tensor contraction equation describing the real-time power balance of the network is constructed to solve for the power commands of each node. :

[0204] ,in, Inject vectors into node power; The external incentives and demand vectors for nodes include grid dispatch instructions, real-time electricity prices, and vehicle load demand; Fixed load vector for nodes; symbol Represents the tensor product. A vector of all ones for the appropriate dimension; This represents a tensor contraction operation along a specific dimension, and its physical meaning is to aggregate the influence from other nodes based on the dynamic coupling weights.

[0205] To ensure the physical realizability and security of energy dispatch networks, dynamic coupling tensors are used. With power solution Apply the following constraints:

[0206] Topological reachability constraint: For any pair of nodes that are not physically connected in the mechanical process Its corresponding coupling weight components It is always zero;

[0207] Power balance integrity constraints: This ensures real-time balance of total power within the network.

[0208] Node secure operation domain constraints: ,in, It is a time-varying convex polyhedron composed of the real-time maximum / minimum schedulable power, voltage / current limits, and temperature boundaries of each node;

[0209] In each control cycle, based on the real-time collected node status... and external vector Online updates of dynamically coupled tensors Furthermore, by solving the tensor contraction equation under constraints, a feasible power command satisfying all constraints is obtained. This completes the construction and parameterization of the multi-port real-time energy scheduling network, providing a network model foundation for subsequent real-time power allocation.

[0210] In this embodiment, when constructing a battery swapping energy network, the existing technology usually uses a fixed or simply switched adjacency matrix, which cannot accurately characterize the continuous, high-dimensional, nonlinear time-varying characteristics of the connection state and coupling strength between nodes caused by the gradual advancement of the mechanical process. Therefore, it is particularly important to establish a dynamic network model that can closely fit the mechanical physical process of battery swapping and accurately quantify its intermittent and staged connection characteristics.

[0211] In this embodiment, tensors of order three and above are introduced. As the core of the network model, the third dimension, L1, is used to encode multidimensional coupling features (such as resistive coupling, inductive coupling, power transmission efficiency, thermal coupling, etc.), which goes beyond the description of single scalar weights in traditional graph theory. Tensor elements The generation formula (including flow window functions) The discrete-stage signals of the mechanical process and the continuous electrical states of the nodes are deeply integrated by the state vector, so that the network topology and connection strength can evolve in strict time and physical synchronization with the mechanical action.

[0212] Upgrading the classic nodal power balance equations to tensor contraction operations naturally enables dynamic topology to be transformed. , external incentives It integrates load requirements into a single mathematical framework and supports efficient numerical solutions.

[0213] The proposed topological reachability constraints are directly derived from the physical reality of mechanical connections, ensuring the physical authenticity of the network model, while the time-varying convex polyhedron safe operating domain... The concept extends the node security boundary from static to dynamic, changing with state and time. Its definition and solution complexity become feasible due to the tensor model.

[0214] The beneficial effects of the above technical solution are as follows: the dynamic coupling tensor model can accurately characterize the transient and steady-state characteristics of the entire process of connection establishment-energy transfer-connection disconnection during battery swapping, especially details such as contact resistance changes and pre-charging path switching. The model fidelity is much higher than that of the fixed topology model. Since the network model accurately reflects the physical constraints, the power command obtained based on this model is theoretically closer to the global optimum. This tensor network model provides an idealized environment with a regular structure and continuously differentiable parameters for advanced algorithms such as model predictive control and deep reinforcement learning, which greatly facilitates the development and application of subsequent advanced energy management strategies.

[0215] This invention provides an intelligent battery swapping control system based on auxiliary power supply for electric vehicles, such as... Figure 2 As shown, it includes:

[0216] The status entry module is used to determine that the electric vehicle has entered the battery swapping state after entering the battery swapping station and establishing a high-voltage electrical connection, and to construct the high-voltage bus of the electric vehicle in the battery swapping state, the battery pack to be replaced, the energy system in the station and the grid interface into a multi-port real-time energy dispatch network.

[0217] The instruction determination module is used to solve the power allocation equation in real time during the duration of the battery swapping state. To determine the real-time power command for each energy port, where, For power command vector, As a vector of real-time status and demand, To normalize the mechanical process time Let be the switching state coupling matrix of the independent variable, and let the matrix elements be... For weighting functions;

[0218] The input / output coordination module is used to coordinate and control the power output or input of each energy port according to the real-time power command, so as to realize continuous energy scheduling synchronized with the operation of the battery swapping machinery.

[0219] The status exit module is used to exit the battery swapping state after the battery swapping mechanical operation is completed and the high voltage system of the electric vehicle is confirmed to be stable.

[0220] The beneficial effects of the above technical solution are as follows: by constructing a multi-port real-time energy dispatch network, the mechanical operation process of battery swapping is transformed into a controllable energy dispatch window, realizing the synchronous coordination of battery swapping actions and energy dispatch; by solving the power allocation equation in real time and controlling the power of each port in a closed loop, the power surge during battery connection / disconnection is effectively mitigated, and the grid response capability is improved; at the same time, it provides continuous and stable power supply for vehicle load, maximizes the utilization of energy resources during battery swapping, and improves the operational economy and grid friendliness of the battery swapping station.

[0221] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A smart battery swapping control method based on auxiliary power supply for electric vehicles, characterized in that, include: Step 1: After the electric vehicle enters the battery swapping station and establishes a high-voltage electrical connection, it is determined that it has entered the battery swapping state, and the high-voltage bus of the electric vehicle in the battery swapping state, the battery pack to be replaced, the station's energy system and the grid interface are constructed into a multi-port real-time energy dispatch network. Step 2: Solve the power allocation equation in real time during the duration of the switching state. To determine the real-time power command for each energy port, where, For power command vector, As a vector of real-time status and demand, To normalize the mechanical process time Let be the switching state coupling matrix of the independent variable, and let the matrix elements be... For weighting functions; Step 3: Based on the real-time power command, coordinate and control the power output or input of each energy port to achieve continuous energy scheduling synchronized with the operation of the battery swapping machinery; Step 4: Once the battery swapping operation is complete and the high-voltage system of the electric vehicle is confirmed to be stable, exit the battery swapping state; Wherein, the normalized mechanical process time It is a dimensionless time parameter that maps the absolute time of the battery swapping mechanical operation from start to finish to the interval [0,1]. Wherein, the weight function The expression is: ； ； ； ； ； in, For mechanical constraint envelope function, This represents the normalized time point of the k-th critical mechanical separation action; This is the step function after smoothing approximation; The tolerance parameter is related to the duration of the physical process of the k-th action; n is the total number of critical mechanical separation actions; The steady-state gain coefficient is determined by the physical characteristics of port i and the baseline weight of scheduling target j. This is the normalized dynamic contact resistance function; This represents the median value within the range where the connector is in a stable contact state. The resistance change sensitivity coefficient; This is a switching function triggered by a small change in contact resistance signal monitored in real time. This is the dynamic margin function for battery state; The current optimal scheduling reference state of charge is dynamically calculated based on the battery's current state of health (SOH). , These are the shape adjustment parameters; This is the thermal safety boundary function; Real-time temperature rise of battery pack or connection points; , These are the safe temperature rise threshold and the maximum permissible threshold, respectively. , These are the attenuation coefficient and the power exponent, respectively; , These are the dynamic balancing weights, and ; The preset standard health status value.

2. The intelligent battery swapping control method based on auxiliary power supply for electric vehicles according to claim 1, characterized in that, Once an electric vehicle enters the battery swapping station and a high-voltage electrical connection is established, the process of determining whether it has entered the battery swapping state includes: After the vehicle enters the battery swapping station, the battery swapping equipment performs a mechanical docking process that includes at least two independent action sequences. Each action sequence is monitored in real time by an integrated force-displacement-temperature three-parameter fusion sensor to generate a corresponding dynamic physical feature vector. Meanwhile, during the mechanical docking process of the battery swapping equipment, the broadband vibration exciter integrated on the locking or docking mechanism is controlled to inject a preset non-stationary broadband excitation signal into the mechanical docking interface. Simultaneously, the multi-dimensional dynamic response signal generated by the propagation of the excitation signal in the mechanical structure is collected by the sensor array arranged on the battery swapping equipment side and the vehicle interface side. Based on the collected multidimensional dynamic response signals, the physical identity transfer function feature vector characterizing the dynamic characteristics of the mechanical structure is extracted, and dynamic matching and verification are performed with the pre-stored benchmark physical identity fingerprint map of the corresponding vehicle model. When all verification results reach their respective set confidence thresholds, physical identity verification is deemed successful; otherwise, a first warning is issued. Once the physical authentication is verified, the dynamic physical feature vector is matched with the pre-stored baseline physical vector of the corresponding vehicle model. When the matching degree reaches the first threshold, the physical connection is determined to be truly established; otherwise, a second warning reminder is issued.

3. The intelligent battery swapping control method based on auxiliary power supply for electric vehicles according to claim 2, characterized in that, After determining that the physical connection has been truly established, the following steps are also included: An excitation signal containing multiple frequency points is applied to the high-voltage circuit of the electric vehicle, and the response signal of each frequency point is collected simultaneously. The collected impedance spectrum data is compared with the historical impedance spectrum fingerprint database associated with the pre-stored VIN code of the corresponding vehicle model, and the stability of the impedance spectrum changes within a set time window is detected in real time. The electrical connection is deemed genuine and stable only if the impedance spectrum shape matching degree reaches the second threshold and the spectral deformation coefficient is lower than the set threshold; otherwise, a second warning is issued.

4. The intelligent battery swapping control method based on auxiliary power supply for electric vehicles according to claim 2, characterized in that, After determining that the electrical connection is genuine and stable, the following steps are also included: A set of dynamically modulated power waveforms is injected into the battery of the electric vehicle, and a first verification code is extracted based on the power waveform detected by the electric vehicle, wherein the power waveform carries the first verification code; The electric vehicle injects another set of dynamically modulated current waveforms into the battery swapping station, and the battery swapping station extracts a second verification code from the current waveform, wherein the current waveform carries the second verification code. The power circuit and communication channel are deemed to be genuine and reliable only if the verification codes in both directions are successfully matched and the dynamic response characteristics of the power waveform and current waveform are consistent with the pre-stored model; otherwise, a third early warning is issued.

5. The intelligent battery swapping control method based on auxiliary power supply for electric vehicles according to claim 2, characterized in that, Dynamic matching and verification with the pre-stored baseline physical identity fingerprint map of the corresponding vehicle model, including: Based on multidimensional dynamic response signals, a multi-scale nonlinear state-space equation describing the dynamic characteristics of mechanical docking structures is established: ； ； in, It is a high-order state vector that includes structural displacement, velocity, and internal state. For the excitation signal vector, For the observed response vector; For a nonlinear time-varying system matrix, its elements include nonlinear function terms for structural stiffness and damping; , , To incentivize positions and observation location Related input / output matrices; and These represent model uncertainty and observation noise, respectively. Solving the multi-scale nonlinear state-space equations yields the nonlinear modal characteristic tensor that characterizes the essential dynamic properties of the mechanical structure. ,in, It is a nonlinear function of the norm of the i-th order state vector; Let be a nonlinear function of the i-th order displacement and velocity; Let i be the left nonlinear mode shape vector of order i; Let i be the right nonlinear mode shape vector of order i; Let i be the kernel of the i-th order nonlinear frequency response function; Let m be the i-th order nonlinear coupling operator; m is the order. Based on the nonlinear modal feature tensor Extract the feature set to construct the physical identity transfer function feature vector F: ,in, These are linear approximate eigenvectors; These are nonlinear eigenvectors; These are topological feature vectors; The physical identity transfer function feature vector F is decomposed and quantized to generate the physical identity fingerprint (PIF). ,in, To use the mother wavelet Continuous wavelet transform of F is used to extract multi-scale features in the time-frequency domain; The gradient of the physical identity transfer function feature vector F as a function of time t1; The integral is the characteristic loop along the time path C of the connection process; , , These are quantization, hashing, and signature encoding functions, respectively. The generated physical identity fingerprint (PIF) is dynamically matched and verified with the pre-stored benchmark physical identity fingerprint map. Physical authentication is considered successful if and only if all authentication results reach their respective confidence thresholds.

6. An intelligent battery swapping control system based on auxiliary power supply for electric vehicles, characterized in that, include: The status entry module is used to determine that the electric vehicle has entered the battery swapping state after entering the battery swapping station and establishing a high-voltage electrical connection, and to construct the high-voltage bus of the electric vehicle in the battery swapping state, the battery pack to be replaced, the energy system in the station and the grid interface into a multi-port real-time energy dispatch network. The instruction determination module is used to solve the power allocation equation in real time during the duration of the battery swapping state. To determine the real-time power command for each energy port, where, For power command vector, As a vector of real-time status and demand, To normalize the mechanical process time Let be the switching state coupling matrix of the independent variable, and let the matrix elements be... For weighting functions; The input / output coordination module is used to coordinate and control the power output or input of each energy port according to the real-time power command, so as to realize continuous energy scheduling synchronized with the operation of the battery swapping machinery. The status exit module is used to exit the battery swapping state after the battery swapping mechanical operation is completed and the high voltage system of the electric vehicle is confirmed to be stable. Wherein, the normalized mechanical process time It is a dimensionless time parameter that maps the absolute time of the battery swapping mechanical operation from start to finish to the interval [0,1]. Wherein, the weight function The expression is: ； ； ； ； ； in, For mechanical constraint envelope function, This represents the normalized time point of the k-th critical mechanical separation action; This is the step function after smoothing approximation; The tolerance parameter is related to the duration of the physical process of the k-th action; n is the total number of critical mechanical separation actions; The steady-state gain coefficient is determined by the physical characteristics of port i and the baseline weight of scheduling target j. This is the normalized dynamic contact resistance function; This represents the median value within the range where the connector is in a stable contact state. The resistance change sensitivity coefficient; This is a switching function triggered by a small change in contact resistance signal monitored in real time. This is the dynamic margin function for battery state; The current optimal scheduling reference state of charge is dynamically calculated based on the battery's current state of health (SOH). , These are the shape adjustment parameters; This is the thermal safety boundary function; Real-time temperature rise of battery pack or connection points; , These are the safe temperature rise threshold and the maximum permissible threshold, respectively. , These are the attenuation coefficient and the power exponent, respectively; , These are the dynamic balancing weights, and ; The preset standard health status value.

Images