Contact arc suppression control method and system for charging gun microswitch

Abstract

The application relates to the technical field of arc extinguishing of electric switches, and particularly discloses a contact arc suppression control method and system for a charging gun micro switch, the method comprising: acquiring real-time composite sensing data of the charging gun micro switch, wherein the real-time composite sensing data comprises a piezoelectric voltage signal collected by a micro-nano piezoelectric sensing unit, a contact end voltage signal collected by a pole end voltage sensor, and a loop current signal collected by a loop current sampling unit. According to the technical scheme, the pre-arcing precursor is identified before the physical separation of the moving and static contacts, and the loop current is transferred, so that the mechanical bounce arcing risk caused by the non-standard pressing action is reduced. In the actual plug-in operation of the charging gun, the signals collected by the multi-dimensional sensing device are decoupled through orthogonal projection, and the pre-arcing precursor features can be separated from the composite data containing hand shaking and spring vibration.

Description

Technical Field

[0001] This invention relates to the field of arc extinguishing technology for electrical switches, and particularly to a method and system for controlling and suppressing contact arcs in a micro switch for a charging gun. Background Technology

[0002] In new energy vehicle charging systems, microswitches inside the charging gun plug are used to control the electrical connection and disconnection of the circuit. Existing microswitches typically employ an internal spring and a transmission mechanism working together, relying on mechanical force to separate the moving and stationary contacts. Under standard operation, the transmission mechanism provides the breaking speed, causing the moving and stationary contacts to physically separate, thereby cutting off the electrical signal.

[0003] In practical applications, operators often press slowly, hold halfway, or experience hand tremors, causing the microswitch's internal transmission mechanism to fail to establish a normal breaking speed. At this point, the moving and stationary contacts repeatedly bounce mechanically within the critical distance. Under the influence of circuit current and voltage, the contact gap can break down, triggering an electric arc. The high temperature of the arc can erode the contact surface, causing contact adhesion and hindering the power-off operation of the control circuit. Simultaneously, the electromagnetic disturbance generated by the arc can affect the stability of the charging system's communication signal. Summary of the Invention

[0004] This invention aims to at least partially solve one of the technical problems in related technologies. Therefore, the purpose of this invention is to propose a contact arc suppression control method and system for a micro switch in a charging gun, so as to improve the electrical service life of the switch and the operational reliability of the charging control circuit.

[0005] To achieve the above objectives, a first aspect of the present invention provides a method for controlling the contact arc suppression of a micro switch in a charging gun, comprising the following steps:

[0006] The real-time composite sensing data of the charging gun micro switch is acquired. The real-time composite sensing data includes the piezoelectric voltage signal collected by the micro-nano piezoelectric sensing unit, the contact terminal voltage signal collected by the electrode terminal voltage sensor, and the loop current signal collected by the loop current sampling unit.

[0007] Based on the real-time composite sensing data, a real-time composite electromechanical contact parameter matrix is ​​constructed, which includes piezoelectric channel characteristics, terminal voltage channel characteristics, and contact conductivity channel characteristics.

[0008] Based on the pre-calibrated spring stress-displacement nonlinear curve, the mechanical common mode subspace in the real-time composite electromechanical contact parameter matrix is ​​extracted, and the pure arc initiation precursor critical residual characterizing the arc initiation precursor of the moving and stationary contacts is separated by orthogonal projection decoupling.

[0009] The pre-arc threshold residual of the pure arc initiation is judged by advance trend compensation. When the preset micro-bounce arc initiation triggering condition is met, a conduction command is sent to the solid-state bypass arc extinguishing circuit connected in parallel at both ends of the moving and stationary contacts.

[0010] Before the moving and stationary contacts physically separate, the control circuit current flowing through the moving and stationary contacts is linearly transferred to the solid-state bypass arc extinguishing circuit, so that the moving and stationary contacts physically separate in a zero-current state.

[0011] Once the physical gap between the moving and stationary contacts is calculated and determined to reach the preset safe insulation distance, the solid-state bypass arc extinguishing circuit is controlled to perform a flexible soft shutdown according to the preset shutdown slope.

[0012] To achieve the above objectives, a second aspect of the present invention provides a contact arc suppression control system for a micro switch of a charging gun, comprising:

[0013] The data acquisition module is used to acquire real-time composite sensing data of the charging gun micro switch. The real-time composite sensing data includes piezoelectric voltage signal collected by the micro-nano piezoelectric sensing unit, contact terminal voltage signal collected by the electrode terminal voltage sensor, and loop current signal collected by the loop current sampling unit.

[0014] The parameter matrix construction module is used to construct a real-time composite electromechanical contact parameter matrix containing piezoelectric channel features, terminal voltage channel features and contact conductivity channel features based on the real-time composite sensing data.

[0015] The orthogonal projection decoupling module is used to extract the mechanical common mode subspace in the real-time composite electromechanical contact parameter matrix based on the pre-calibrated spring stress-displacement nonlinear curve, and to separate the pure arc initiation precursor critical residual characterizing the arc initiation precursor of the moving and stationary contacts through orthogonal projection decoupling.

[0016] The arc initiation precursor determination and triggering module is used to make an advance trend compensation judgment on the critical residual of the pure arc initiation precursor. When the preset micro-bounce arc initiation triggering condition is met, a conduction command is sent to the solid-state bypass arc extinguishing circuit connected in parallel at both ends of the moving and stationary contacts.

[0017] The bypass current transfer module is used to linearly transfer the control circuit current flowing through the moving and stationary contacts to the solid-state bypass arc extinguishing circuit before the moving and stationary contacts physically separate, so that the moving and stationary contacts physically separate in a zero-current state.

[0018] The flexible soft shutdown module is used to control the solid-state bypass arc extinguishing circuit to perform flexible soft shutdown according to a preset shutdown slope after calculating and determining that the physical gap between the moving and stationary contacts reaches a preset safe insulation distance.

[0019] To achieve the above objectives, a third aspect of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory. When the computer program is executed by the processor, it implements the above-described method for controlling the contact arc suppression of a micro switch for a charging gun.

[0020] This invention reduces the risk of mechanical bounce-initiated arcing caused by improper pressing actions by identifying pre-arc initiation signs before the physical separation of moving and stationary contacts and transferring the circuit current. During actual plugging and unplugging operations of the charging gun, signals collected by multi-dimensional sensing devices are decoupled through orthogonal projection, enabling the separation of pre-arc initiation features from composite data including hand tremors and spring vibrations. Based on this, the system conducts a parallel solid-state bypass before the physical gap forms between the contacts, linearly transferring the current flowing through the contacts, allowing the moving and stationary contacts to complete physical separation in a zero-current state. This process limits the physical conditions for air breakdown between contacts, reduces high-temperature ablation and adhesion on the contact surface, and the flexible soft-shutdown control of the bypass current smooths out electromagnetic oscillations generated when the circuit is cut off, improving the electrical lifespan of the switch and the operational reliability of the charging control circuit. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the contact arc suppression control method for the micro switch of the charging gun provided by the present invention.

[0022] Figure 2 This is a schematic diagram of the state tracking waveforms before and after normalization of the multi-channel sensing signal in the contact arc suppression control method of the micro switch for charging gun provided by the present invention.

[0023] Figure 3 This is a nonlinear response surface diagram of the mechanical common mode direction vector in the three-dimensional sensing feature space in the contact arc suppression control method of the micro switch for charging gun provided by the present invention.

[0024] Figure 4 This is a comparison of signal denoising and power spectral density before and after orthogonal projection decoupling of real-time sensing data in the contact arc suppression control method of the micro switch for charging gun provided by the present invention.

[0025] Figure 5 This is a transient timing waveform diagram of arc precursor triggering and solid-state bypass current linear transfer in the contact arc suppression control method of the micro switch for charging gun provided by the present invention.

[0026] Figure 6 This is a diagram showing the stiffness decay and contact wear distribution of the microswitch during long-cycle aging in the contact arc suppression control method for the microswitch of the charging gun provided by this invention.

[0027] Figure 7This is a diagram of the adaptive convergence process of the reference parameters based on the first-order exponential moving average algorithm in the contact arc suppression control method of the micro switch for charging gun provided by the present invention.

[0028] Figure 8 This is a schematic diagram illustrating the implementation of the contact arc suppression control system for the micro switch of the charging gun provided by the present invention;

[0029] Figure 9 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0030] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0031] The following description, with reference to the accompanying drawings, details a contact arc suppression control system, method, and electronic device for a charging gun micro switch according to embodiments of the present invention.

[0032] Example 1:

[0033] This embodiment provides a method for controlling the contact arc suppression of a micro switch in a charging gun. This method is mainly applied in DC fast charging or AC high-power charging scenarios for new energy vehicles.

[0034] The microswitch inside the charging gun handle is the core electrical actuator for starting and stopping the charging handshake protocol and for high-voltage safety interlocking. Its reliability directly determines the charging safety of the entire vehicle. In actual use, operators often perform non-standard operations such as slow pressing, hesitant half-pressing, or hand tremors, which prevents the mechanical transmission mechanism inside the microswitch from establishing the designed critical breaking speed.

[0035] At this point, the moving and stationary contacts will experience high-frequency mechanical bouncing within the critical arc initiation distance. The method disclosed in this embodiment solves the arc erosion problem caused by the aforementioned mechanical easing through the coordinated control of the underlying edge computing platform and the parallel solid-state actuator.

[0036] To ensure the real-time performance of the entire control strategy, this embodiment uses a Field-Programmable Gate Array (FPGA) chip as the underlying edge computing platform. Compared to traditional microcontrollers, FPGA chips can provide nanosecond-level parallel hardware pipeline processing capabilities, thereby avoiding uncontrollable delays caused by interrupt scheduling in the software operating system for complex control algorithms, and ensuring that theoretical parameters and actual data throughput do not conflict in terms of resources.

[0037] For example, such as Figure 1 As shown, this embodiment provides a method for controlling the contact arc suppression of a micro switch in a charging gun, including: acquiring real-time composite sensing data of the micro switch in the charging gun, wherein the real-time composite sensing data includes piezoelectric voltage signals collected by a micro-nano level piezoelectric sensing unit, contact terminal voltage signals collected by a terminal voltage sensor, and loop current signals collected by a loop current sampling unit.

[0038] Specifically, the aforementioned micro / nano-scale piezoelectric sensing unit is manufactured using aluminum nitride thin-film technology and is attached to the strain concentration area at the root of the metal spring inside the microswitch. This location can capture the composite mechanical stress deformation generated by the spring during bending under force to the greatest extent and output the corresponding piezoelectric voltage signal. The terminal voltage sensor uses a differential operational amplifier circuit with high input impedance, connected in parallel across the moving and stationary contact terminals of the microswitch, for real-time monitoring of the potential difference across the contacts. The loop current sampling unit uses a precision shunt resistor with low parasitic inductance connected in series in the control loop.

[0039] The data from the three types of sensing units mentioned above are sampled by a synchronously triggered analog-to-digital converter (ADC), and the fixed sampling time step of the ADC is defined as follows: In this embodiment, the fixed sampling time step is determined based on the hardware clock reference of the field-programmable gate array. Set to 0.5 microseconds.

[0040] Optionally, the method provided in this embodiment includes: constructing a real-time composite electromechanical contact parameter matrix that includes piezoelectric channel features, terminal voltage channel features and contact conductivity channel features based on the real-time composite sensing data.

[0041] Specifically, the construction of the real-time composite electromechanical contact parameter matrix includes: normalizing the piezoelectric voltage signal, the contact terminal voltage signal, and the loop current signal according to their respective full-scale reference values.

[0042] Since piezoelectric voltage signals characterize mechanical stress, contact terminal voltage signals characterize electric potential, and loop current signals characterize current flow energy, their physical dimensions are fundamentally different. If matrix operations are performed directly, the characteristics of smaller numerical channels will be masked by larger numerical channels. Therefore, it is necessary to divide each of them by their respective full-scale reference values ​​used in system design and calibration, mapping them to a dimensionless numerical space to ensure the objective accuracy of the geometric meaning in subsequent multidimensional feature space calculations.

[0043] It should be noted that the method further includes: when the absolute value of the contact terminal voltage signal is greater than or equal to a preset effective measurement lower limit, calculating the contact microconductivity in real time using the loop current signal and the contact terminal voltage signal according to Ohm's law.

[0044] In engineering implementation, when the moving and stationary contacts are in a well-closed state, the voltage signal at the contact end will be very weak. Due to the noise floor of the analog-to-digital converter, directly performing a division operation at this time will cause the conductance result to diverge. Therefore, a preset effective measurement lower limit is introduced, and the hardware divider only starts working when the voltage signal at this end is higher than the noise floor threshold.

[0045] Based on the above processing results, in this embodiment, the normalized piezoelectric voltage signal is used as the piezoelectric channel feature, the normalized contact terminal voltage signal is used as the terminal voltage channel feature, and the normalized contact microconductivity is used as the contact conductivity channel feature. The three-dimensional real-time composite electromechanical contact parameter matrix is ​​constructed by arranging them in columns.

[0046] To accurately describe this matrix structure, the real-time composite electromechanical contact parameter matrix is ​​expressed using the following formula:

[0047] ;

[0048] In the formula, Represents the real-time composite electromechanical contact parameter matrix; Represents the current discrete-time sampling sequence number; This represents the normalized piezoelectric channel characteristics; This represents the normalized terminal voltage path characteristics. This represents the normalized contact conductivity channel characteristics.

[0049] The three dimensions of this matrix objectively reflect the coupling relationship between the internal mechanical stress and electrical contact state of the microswitch at the current moment.

[0050] like Figure 2 The diagram illustrates the state tracking waveforms of the multi-channel sensing signal before and after normalization. In the upper half of the diagram, the horizontal axis represents time in milliseconds, and the vertical axis represents the original amplitude in volts or amperes.

[0051] Figure 2 The blue piezoelectric voltage curve shows a non-linear upward trend within a time window of 0 to 2 milliseconds, accompanied by normal mechanical resonance fluctuations, objectively reflecting the gradual increase in the pressure applied by the microswitch; the green loop current curve maintains a small, normal fluctuation of around 5 amperes; the red contact voltage curve initially maintains a stable low contact voltage drop level of around 0.2 volts, while around 1.5 milliseconds, a significant rise and high-frequency oscillation waveform change begin to appear, accurately reflecting the physical micro-bounce state when the contacts are about to separate.

[0052] Because these three physical sensing signals have large differences in dimensions and numerical ranges, their direct participation in subsequent matrix operations can easily cause data truncation or numerical overflow in the hardware multiplier-accumulator of the underlying computing platform, thereby triggering uncontrollable conflicts in computing resources.

[0053] Figure 2 The lower half of the diagram shows the state tracking graph after hardware processing, with the vertical axis changed to a dimensionless normalized amplitude. The system scales the blue curve proportionally to a piezoelectric channel characteristic curve within a safe range at a fixed sampling period of 0.5 microseconds. Simultaneously, after confirming that the voltage at the red contact terminal is greater than the effective measurement lower limit, a division logic is initiated to convert the real-time current-to-voltage ratio into a green contact conductivity channel characteristic curve.

[0054] It can be clearly observed that when the terminal voltage fluctuates and rises after 1.5 milliseconds, the green curve representing the contact conductivity immediately responds in accordance with physical laws, rapidly and smoothly dropping from a high level of 0.8 to below 0.2. Through this series of waveform transformations, the piezoelectric channel characteristics, terminal voltage channel characteristics, and contact conductivity channel characteristics, which originally had huge differences in amplitude, are unified and converged within the dimensionless numerical range of 0 to 1.2.

[0055] This preprocessing process, while fully preserving the true physical fluctuation trends of each signal, removes numerical obstacles for constructing a real-time composite electromechanical contact parameter matrix, truly demonstrating a high degree of matching between theoretical parameter settings and underlying hardware processing capabilities.

[0056] For example, since the piezoelectric voltage signal is mixed with low-frequency stress caused by external hand pressure and vibration stress caused by the inherent spring standing wave of the metal reed, the step of obtaining the current reed displacement estimate is included before the pure arc-starting critical residual is separated by orthogonal projection decoupling.

[0057] Specifically, the above steps are as follows: the piezoelectric voltage signal is processed by a preset sliding average window to extract the low-frequency mechanical component.

[0058] In a field-programmable gate array (FPGA), the moving average filtering algorithm is implemented using shift registers and accumulators, with a moving average window length of 64 sampling points. This process filters out high-frequency spring resonance noise, retaining the low-frequency mechanical components that reflect the actual pressing displacement.

[0059] Furthermore, by combining the preset temperature compensation coefficient function, the low-frequency mechanical components are mapped and calculated using the reed stress-displacement nonlinear curve to obtain the corresponding estimated value of the current reed displacement. The estimated value of the current reed displacement serves as the mechanical state reference and the scheduling variable for subsequent online interpolation.

[0060] The metal spring inside a microswitch is a typical nonlinear mechanical energy storage element, and there is a significant hysteresis loop characteristic between its stress and displacement; moreover, the elastic modulus of the metal material will drift with drastic changes in ambient temperature.

[0061] Therefore, this embodiment embeds a one-dimensional lookup table in the hardware storage area, which is constructed based on a calibrated reed stress-displacement nonlinear curve. Simultaneously, the reading of the negative temperature coefficient thermistor near the microswitch is input into the temperature compensation coefficient function, and the gain of the lookup table mapping result is adjusted. This lookup table mapping design avoids complex floating-point equation solving, maintains a fixed computational delay, and meets the stringent real-time requirements of safety-critical scenarios.

[0062] To quantify this mapping process, the reed displacement estimation calculation is represented by the following formula:

[0063] ;

[0064] In the formula, This represents the current estimated reed displacement. This represents the inverse function mapping relationship of the spring stress-displacement nonlinear curve, which is embedded in the hardware lookup table. This represents the extracted low-frequency mechanical components; This represents the preset temperature compensation coefficient function; This represents the current ambient temperature value.

[0065] To make the orthogonal projection decoupling algorithm feasible for engineering applications, this embodiment provides a factory calibration mechanism.

[0066] Specifically, the preset multiple calibration node displacements and their corresponding mechanical common mode direction vectors are obtained in the following way: during the calibration stage, standard sweep frequency mechanical disturbances are applied at the preset multiple calibration node displacements respectively.

[0067] On the actual production line calibration platform, a high-precision voice coil motor is used to simulate the mechanical pressing action applied externally. The stroke of the micro switch is monitored by a laser displacement sensor, and standard sweep frequency mechanical disturbances with a frequency continuously varying from 10 Hz to 500 Hz are injected at multiple pre-set stroke positions (i.e. multiple pre-set calibration node displacements) to fully simulate the hand tremors and shaking conditions that may occur during human operation.

[0068] During this process, the response ratios of the piezoelectric channel characteristics, the terminal voltage channel characteristics, and the contact conductivity channel characteristics under the standard frequency sweep mechanical disturbance are measured simultaneously. The mechanical common mode direction vectors corresponding to each calibration node are generated and stored in the feedforward calibration library of the underlying edge computing platform.

[0069] Because the sensitivity of a microswitch to changes in the micro-contact area caused by mechanical vibration varies significantly depending on the depth of displacement, even a tiny mechanical disturbance at the critical point where the contact is about to disengage can cause drastic fluctuations in the contact conductivity characteristics; whereas in the initial pressing stage, the same mechanical disturbance will hardly cause any change in conductivity.

[0070] Therefore, it is necessary to record this coupling response ratio at multiple discrete calibration nodes to form a mechanical common mode direction vector with three-dimensional characteristics, thereby providing a real physical coordinate reference for subsequent elimination of electrical contact misjudgment caused by mechanical vibration.

[0071] It should be noted that in this calibration and subsequent decoupling process, the orthogonal projection decoupling process is limited to a fixed-scale matrix pipeline operation to ensure that the time required for computing power overhead is less than or equal to the single sampling period of the underlying edge computing platform.

[0072] This embodiment avoids using complex algorithms with slow convergence because any uncontrollable fluctuations in computation time within the microsecond-level time window when the contact is about to experience arc breakdown will lead to arc extinguishing failure. Fixed-size matrix operations have a definite and unique clock cycle in the field-programmable gate array, providing stable and reliable hardware performance.

[0073] like Figure 3 The nonlinear response surface plot of the mechanical common-mode direction vector in the three-dimensional sensing feature space is shown. The three coordinate axes of the plot correspond to the piezoelectric channel feature, the terminal voltage channel feature, and the contact conductance channel feature, respectively, and all units are dimensionless.

[0074] In three-dimensional space, the mechanical common-mode response surface, composed of gradient colors, represents the reference mapping relationship of the feedforward calibration library established during the factory calibration phase. The surface color changes gradually with height, intuitively mapping the spatial distribution trend of the contact conductivity channel characteristic value from 0 to 1. Multiple red calibration node perturbation scatter points are distributed on the surface. These scatter points objectively record the real measurement data of the multi-channel response when standard frequency sweep mechanical perturbation is applied at different pressing depth nodes.

[0075] It can be observed that in the initial pressing stage when the piezoelectric channel characteristics are between 0.2 and 0.4, the red disturbance points are mainly concentrated in the high-level region where the contact conductivity channel characteristics are between 0.5 and 0.7. At this time, the spatial angle change of the curved surface is relatively gentle, indicating that the coupling effect of normal mechanical vibration on the electrical contact state is relatively stable.

[0076] When the terminal voltage channel characteristic rises to the critical region of 0.4 to 0.6, the response surface undergoes a significant spatial twist and a sharp decline, and the corresponding red scatter points also exhibit a large-scale discrete distribution. The value of the contact conductivity channel characteristic drops rapidly to around 0.3. This spatial twist confirms that in the critical disconnection region of the microswitch, there is a highly nonlinear coupling physical law between its mechanical displacement and electrical state.

[0077] Through this nonlinear response surface, the system can perform linear interpolation based on the current displacement estimate during actual operation, accurately tracking the angular change of the mechanical common-mode direction vector in space. This feature space mapping mechanism based on finite discrete calibration nodes uses basic coordinate lookup tables and discrete point interpolation to replace complex floating-point equation solving, and its computational scale matches the hardware resource throughput of the underlying field-programmable gate array. In execution environments with real sensor measurement fluctuations and noise interference, this mechanism can still provide reliable geometric reference support for orthogonal projection decoupling, ensuring that the control strategy of suppressing mechanical artifacts has objective engineering feasibility and good beneficial effects.

[0078] After obtaining the real-time status and preset benchmark, the method of this embodiment extracts the mechanical common mode subspace in the real-time composite electromechanical contact parameter matrix based on the pre-calibrated spring stress-displacement nonlinear curve, and decouples and separates the pure arc initiation precursor critical residual that characterizes the arc initiation precursor of the moving and stationary contacts through orthogonal projection.

[0079] Specifically, the step of decoupling and separating the pure arc-initiating precursor critical residual characterizing the arc-initiating precursor of the moving and stationary contacts through orthogonal projection includes: calculating the slow-varying trend vector of the real-time composite electromechanical contact parameter matrix within the preset sliding mean window. This slow-varying trend vector reflects the background state caused by extremely slow physical processes such as the gradual peeling of the oxide film and the gradual increase of the pressing pressure.

[0080] Subsequently, based on the current estimated reed displacement, linear piecewise interpolation is performed between multiple preset calibration node displacements to calculate the mechanical common-mode direction vector that adaptively changes with the reed displacement. This is because the actual reed displacement is a continuous variable and cannot remain precisely at discrete calibration nodes.

[0081] By using linear piecewise interpolation, the system can smoothly transition and estimate the precise mechanical vibration coupling direction based on the interval in which the current reed displacement estimate is located.

[0082] Next, the mechanical common mode direction vector is combined with the slowly varying trend vector to construct the mechanical common mode subspace basis matrix. This subspace basis matrix spans a special plane in the three-dimensional perceptual data space. All data changes falling within this plane are physically caused by mechanical actions (such as normal hand pressing, abnormal hand shaking, and spring standing wave friction), and do not include electrical precursors of impending electrical breakdown.

[0083] Furthermore, based on a least-squares projection algorithm with preset regularization coefficients, the orthogonal projection of the real-time composite electromechanical contact parameter matrix onto the mechanical common-mode subspace basis matrix is ​​calculated; then, the orthogonal projection is subtracted from the real-time composite electromechanical contact parameter matrix to remove the common-mode artifacts caused by mechanical vibration and friction, and the pure arc-starting precursor critical residual orthogonal to the mechanical common-mode subspace basis matrix is ​​output.

[0084] To accurately illustrate the joint solution process of orthogonal projection and residual extraction, the following formula for calculating the critical residual vector of the pure arc initiation precursor is given:

[0085] ;

[0086] In the formula, The pure arc-starting precursor critical residual vector represents the calculated output. Represents the real-time composite electromechanical contact parameter matrix; Represents the basis matrix of the mechanical common mode subspace; Represents the transpose operation of a matrix; This represents the preset regularization coefficient; Represents the identity matrix.

[0087] The physical significance of introducing a preset regularization coefficient is to prevent numerical calculation overflow during the inversion process when the slowly varying trend vector and the mechanical common mode direction vector are nearly parallel and collinear due to extreme working conditions, thus ensuring that the algorithm behavior does not violate the rigorous logic of mathematics and physics.

[0088] The subtraction operation in this formula is geometrically equivalent to extracting the normal vector component of the sensing data perpendicular to the mechanical common-mode subspace. This component objectively filters out all mechanical operational disturbances, leaving behind a characteristic signal that accurately characterizes the field-induced emission current phenomenon at the microscopic protrusions on the surfaces of the moving and stationary contacts, i.e., the actual precursor to arc initiation.

[0089] like Figure 4 The diagram shows a comparison of signal denoising and power spectral density before and after orthogonal projection decoupling of real-time sensed data. In the upper half of the diagram, the horizontal axis represents time in milliseconds, and the vertical axis represents signal amplitude in dimensionless units.

[0090] Figure 4The blue pre-decoupling composite sensing signal curve exhibits obvious low-frequency fluctuations in the time interval of 0 to 1.5 milliseconds, mixed with dense mechanical vibration noise. This part of the data corresponds to the operator's hand tremor and the spring standing wave of the micro switch metal spring itself.

[0091] The red decoupled pure arc-starting critical residual curve was significantly suppressed to near the baseline with near-zero amplitude in the same time interval, demonstrating the effective suppression of mechanical common-mode artifacts by the orthogonal projection matrix.

[0092] When the time reaches 1.5 milliseconds, both curves show synchronous step rise and high-frequency characteristic fluctuations, which objectively reflects that the electrical precursors generated when the micro-contacts are about to disengage are effectively preserved.

[0093] exist Figure 4 The lower half of the graph shows the frequency on the horizontal axis (set to Hertz) and the power spectral density on the vertical axis (set to dB per Hertz).

[0094] The blue power spectral density curve before decoupling shows two significant energy spikes in the low-frequency range and around 500 Hz, reaching a maximum of about -40 dB per Hz. This physically confirms the concentrated distribution of energy from external pressure and spring mechanical vibration.

[0095] After processing by the underlying hardware matrix, the red decoupled power spectral density curve achieved a significant attenuation of about 20 to 30 dB in the energy peak values ​​of the two mechanical characteristic frequency bands mentioned above. In the high-frequency broadband band above 1000 Hz, the energy distribution characteristics of the red curve and the blue curve remained highly consistent.

[0096] The comparison between the time-domain waveform transformation and the frequency-domain energy curve shows that the orthogonal projection decoupling algorithm in the control strategy can separate truly dangerous electrical contact anomalies from complex electromechanical coupling data, while meeting the computational resource constraints of the underlying computing platform. This figure verifies that the system possesses the objective engineering capability to distinguish between conventional mechanical vibrations and actual electrical arcing hazards. The residual background noise fluctuations it exhibits conform to the actual physical denoising execution rules, providing reliable data support for the subsequent accurate calculation of the arcing redline criterion scalar.

[0097] For example, the method performs an advance trend compensation judgment on the pure arc initiation precursor critical residual, and when the preset micro-bounce arc initiation trigger condition is met, it sends a conduction command to the solid-state bypass arc extinguishing circuit connected in parallel at both ends of the moving and stationary contacts.

[0098] Specifically, the aforementioned leading trend compensation judgment process includes: calculating the vector L2 norm of the critical residual of the pure arc initiation precursor to obtain the scalar characteristics of the arc initiation precursor. In the underlying hardware, the solution of the vector L2 norm can be efficiently implemented by the Coordinate Rotating Digital Computer (CORDIC) algorithm module within more than ten clock cycles to perform the square and square root operations, avoiding the occupation of general-purpose multiplier resources.

[0099] Subsequently, a first-order difference calculation is performed on the scalar features of the arc initiation precursor to obtain the rate of change of the precursor scalar features. This first-order difference operation is obtained by subtracting the scalar features of the previous sampling period from the scalar features at the current moment.

[0100] Next, the preset arc-starting precursor amplitude threshold is subtracted from the arc-starting precursor scalar feature, and the product of the rate of change of the precursor scalar feature and the preset precursor rise time constant is added to obtain the arc-starting red line criterion scalar.

[0101] For example, the calculation logic of this criterion is illustrated using the following formula:

[0102] ;

[0103] In the formula, The scalar representing the starting point of the red line criterion; Represents the scalar characteristics of precursors to arc initiation; This represents the preset threshold value for the pre-arc initiation amplitude obtained based on historical extreme values. This represents the preset precursor rise time constant; This represents the rate of change of the precursor scalar characteristic.

[0104] When the scalar value of the arc initiation red line criterion is greater than or equal to zero, the preset micro-bounce arc initiation trigger condition is determined to be met. This design logic does not simply wait for the scalar characteristic to cross the amplitude threshold. This is because in the very early stages of arc breakdown, although the precursor amplitude is still small, its rate of increase (rate of change) is extremely high.

[0105] By introducing compensation calculations that superimpose the rate of change of the precursor scalar characteristics, the system can induce the criterion scalar to cross the zero point several sampling cycles before the physical arc initiation occurs, thus achieving actual advanced trend prediction and reserving a critical margin of several microseconds for subsequent mechanical execution modules.

[0106] To avoid invalid solid-state bypass activation due to noise when the contacts have separated and formed an insulating state, in this embodiment, a transfer feasibility constraint determination is required before sending the activation command.

[0107] Specifically, the constraint determination steps include: calculating the differential rate of change of the current reed displacement estimate; subtracting the current reed displacement estimate from a preset reed displacement separation threshold and dividing by the differential rate of change to predict and calculate the remaining separation time.

[0108] For example, this prediction is calculated using the following formula:

[0109] ;

[0110] In the formula, Represents the remaining separation time; This represents the preset reed displacement separation threshold, which is locked by the internal mechanical structure dimensions of the microswitch. This represents the current estimated reed displacement. This represents the differential rate of change of the current estimated reed displacement.

[0111] Subsequently, it is determined whether the remaining separation time is greater than the sum of the preset transfer conduction time and the preset safety margin time. If so, the linear transfer is performed within the time window before physical separation to ensure that the current transfer completion time is earlier than the physical separation time of the moving and stationary contacts.

[0112] The preset transfer conduction time is limited by the turn-on delay of the solid-state device and the resistance of the current change rate by the parasitic inductance of the circuit; the preset safety margin time is to accommodate the fixed overhead of system communication and logic gate toggling. This judgment logic plays an important role in extremely rapid pressing conditions: if the operator presses very quickly, making the remaining separation time extremely short or even insufficient to transfer the current, the fast-acting mechanism of the microswitch has obtained sufficient critical spring force to safely disconnect the circuit through purely mechanical means. The system then abandons solid-state intervention to avoid the algorithm logic from violating physical laws.

[0113] For example, when the system determines that the transfer conditions are met, before the moving and stationary contacts physically separate, the control circuit current flowing through the moving and stationary contacts is linearly transferred to the solid-state bypass arc extinguishing circuit, so that the moving and stationary contacts physically separate in a zero-current state.

[0114] Specifically, the process of linearly transferring the control circuit current to the solid-state bypass arc-extinguishing circuit includes: triggering the power solid-state switch in the solid-state bypass arc-extinguishing circuit to turn on. The solid-state bypass arc-extinguishing circuit is arranged in parallel with the physical contacts of the micro switch.

[0115] In the hardware modification scheme of this embodiment, the power solid-state switch adopts a silicon carbide metal-oxide-semiconductor field-effect transistor (SiC MOSFET) with low on-resistance. It has the characteristics of high voltage resistance and extremely fast switching speed, and is matched with an isolated gate driver chip with desaturation protection function. This parallel modification scheme does not change the original series breaking mechanical and electrical safety attributes of the microswitch at all, retains the bottom line of defense against physical isolation failure, and the technical solution has objective engineering feasibility.

[0116] After the activation signal is sent, within the preset transfer conduction time, the shunting ratio of the physical moving and stationary contacts decreases at a gradual rate until the current of the physical moving and stationary contacts returns to zero, and the voltage across the moving and stationary contacts is clamped below the preset minimum arc-starting voltage of the contact material.

[0117] According to circuit theory, opening a parallel, extremely low-impedance path forces a redistribution of current. As the gate-source voltage of a silicon carbide device crosses the Miller plateau, its drain-source path conducts, and the physical moving and stationary contacts are clamped by the device's extremely low saturation voltage drop. Taking silver-based alloy contacts as an example, the minimum potential difference threshold required to maintain plasma arc combustion is typically around 12 volts. However, the voltage drop after a silicon carbide device is turned on is typically less than 1 volt, which physically eliminates the potential conditions required for arc generation.

[0118] More importantly, from the moment the conduction command is sent until the flexible soft shutdown ends, the solid-state bypass arc-extinguishing circuit remains continuously conductive, ensuring that the entire conduction action of a single trigger covers the mechanical high-frequency bounce range caused by the separation of the physical moving and stationary contacts. Due to the operator's hand tremors, the contacts of the microswitch often do not separate all at once at the moment of separation, but rather experience dozens of microscopic high-frequency impacts and rebounds within a micrometer-level gap. By maintaining the continuous conduction of the solid-state bypass, all electrical contact bounces occur in a state of "zero current and zero potential difference," thus achieving physical suppression of contact arcs.

[0119] like Figure 5 The transient timing waveforms of the arc initiation precursor triggering and the linear transfer of the solid-state bypass current are shown. The attached figure is divided into three parts, with the horizontal axis representing time in microseconds.

[0120] exist Figure 5 At the top, the vertical axis represents the dimensionless criterion scalar. The blue arc-starting red line criterion curve is in the negative safe zone and accompanied by slight background noise fluctuations 4 microseconds before the time reaches 4.5 microseconds. When the time reaches 4.5 microseconds, the blue curve shows a steep upward trend and successfully crosses the red zero-point trigger threshold dotted line, which objectively reflects the effective identification of arc-starting precursors by the advanced trend compensation judgment logic.

[0121] Figure 5The middle vertical axis represents the driving voltage in volts. After the criterion scalar crosses 0, due to the fixed hardware processing delay of the underlying computing platform, the black bypass driving signal curve begins to flip at 5.0 microseconds and rises to a high level of 15 volts according to the actual physical resistance and capacitance charging and discharging slope.

[0122] Figure 5 The bottom vertical axis represents the current amplitude in amperes. During the transition conduction time from 5.5 microseconds to 7.5 microseconds, the red moving and stationary contact current curve slopes down from the steady-state operating level of 10 amperes and returns to zero after 7.5 microseconds. At the same time, the green solid-state bypass current curve rises synchronously to the total current of the main circuit at 10 amperes with a similar slope. During this transition, the two current curves exhibit normal high-frequency fluctuations in power device operation.

[0123] The above waveform transformation process intuitively demonstrates that the control system can perform linear current transfer within an extremely short time window, so that the physical moving and stationary contacts are in a zero-current state when mechanical separation occurs subsequently. This transient timing waveform verifies the reasonable matching between the technical solution parameter settings and the underlying hardware capabilities.

[0124] For example, after calculating and determining that the physical gap between the moving and stationary contacts reaches the preset safe insulation distance, the solid-state bypass arc extinguishing circuit is controlled to perform a flexible soft shutdown according to the preset shutdown slope.

[0125] Specifically, the control of the solid-state bypass arc-extinguishing circuit to perform flexible soft shutdown includes: utilizing the initial gap of a preset contact closure state. The difference between the current estimated reed displacement and the preset nominal displacement of the closed state is superimposed to calculate the gap between the moving and stationary contacts in real time. This calculation process is derived using purely mechanical linkage relationships.

[0126] When the gap between the moving and stationary contacts is greater than or equal to the preset safety insulation distance, and this state continues to meet the preset holding step number, it is determined that the safety insulation distance has been crossed. The preset safety insulation distance is determined by multiplying the air breakdown critical distance corresponding to the control circuit bus voltage by a preset safety margin coefficient.

[0127] According to Paschen's Law, under standard atmospheric pressure, the breakdown voltage of an air gap is definitely related to the gap distance. In this embodiment, based on the maximum voltage of the current charging system circuit bus, Paschen's curve is consulted to derive the theoretical critical distance for avalanche breakdown, which is then multiplied by, for example, a preset safety margin factor as the final judgment criterion. The requirement that the state continuously meet a preset holding step count is to prevent a single extreme mechanical rebound from causing a sudden increase in gap size and leading to misjudgment.

[0128] After confirming safe insulation, a flexible soft shutdown command is triggered, causing the bypass current in the solid-state bypass arc extinguishing circuit to drop to zero according to a controlled shutdown ramp. The controlled shutdown ramp is configured with a limit based on a preset upper limit of the current change rate.

[0129] For example, the shut-off control ramp uses the following formula for amplitude limiting constraint:

[0130] ;

[0131] In the formula, The time rate of change of the bypass current; This is the real-time bypass current; This represents the preset upper limit of the current change rate, which is derived from the requirements of automotive electromagnetic compatibility (EMC) regulations.

[0132] If a method of rapidly disconnecting solid-state devices is used, the parasitic inductance that inevitably exists in the circuit will generate an extremely high reverse induced electromotive force, which may not only break down the solid-state devices, but also radiate strong electromagnetic interference to the outside.

[0133] By using a gate drive resistor array or active gate voltage clamping technology, the ramp of current drop is actively lengthened, allowing the magnetic field energy stored in the parasitic inductor to be released smoothly, thus achieving a soft landing without electromagnetic overshoot during the entire power-off process.

[0134] For example, to further enhance the system's functional safety level, after the solid-state bypass arc extinguishing circuit performs a flexible soft shutdown to reduce the current to zero, a closed-loop disconnection self-test step is also included.

[0135] Specifically, the above steps include: detecting whether the contact terminal voltage signal is greater than or equal to a preset disconnection confirmation voltage threshold, which is set based on the control circuit bus voltage. When the circuit is completely disconnected and a physical break is formed, the two ends of the contact will bear the full voltage of the bus system. Therefore, a steady-state value is collected using a terminal voltage sensor to compare whether the expected circuit-breaking resting voltage level has been reached.

[0136] Verify that the loop current signal remains zero. Determine if leakage current exists by reading the minute noise level of the loop current sampling unit in the resting state.

[0137] If the contact terminal voltage signal is greater than or equal to the preset disconnection confirmation voltage threshold, and the loop current signal remains zero, it indicates that the physical mechanical contacts and solid-state bypass devices have returned to the disconnection blocking state as instructed. In this case, the arc-free disconnection is confirmed to be successful, and the state machine parameters of the underlying hardware are reset, waiting for the next charging plugging / unplugging action. If the contact terminal voltage signal is less than the preset disconnection confirmation voltage threshold, or the loop current signal does not remain zero, for example, the terminal voltage fails to rise to the bus voltage or there is still trickle current, it is determined that a physical contact welding adhesion fault or solid-state device breakdown failure has occurred. The system immediately triggers the protection mechanism, blocks the charging enable command of the new energy vehicle battery management system, and outputs the highest priority adhesion alarm signal through the controller local area network (CAN bus).

[0138] In existing technology, the micro switch inside the charging gun mainly relies on a mechanical fast-acting spring to generate the breaking displacement. Its breaking speed is completely controlled by the uncontrollable pressing technique of the operator. Under slow pressing conditions, it is difficult to avoid strong arcing caused by the slow separation of moving and stationary contacts, which leads to a sharp reduction in contact life and brings fire hazards.

[0139] This embodiment provides a novel hardware and software collaborative arc suppression mechanism. By introducing an extremely low-latency orthogonal projection algorithm into the underlying hardware to separate mechanical artifacts, it successfully transforms physical perception information into proactive intervention commands for electrical control. At the microscopic scale, it utilizes parallel solid-state devices to actively receive energy transfer, forcibly eliminating the physical conditions required for arc initiation during the mechanical separation phase. Without compromising the original forced redundancy of the physical mechanical breakpoints, this solution reduces the probability of micro-contacts ablation and welding adhesion due to arcing, effectively suppressing electromagnetic interference radiation during the separation process, and significantly improving the electrical reliability and system safety of the charging gun under abnormal operating conditions throughout its entire lifecycle.

[0140] Example 2:

[0141] Based on Example 1, this example further expands upon its solution.

[0142] During its actual physical lifespan, a charging gun typically undergoes tens of thousands of high-frequency insertions and removals, as well as alternating shocks from high and low temperatures. During this long-term service, the metal mechanical spring inside the microswitch will fatigue and yield due to continuous stress cycles, leading to irreversible physical attenuation of its mechanical stiffness. Simultaneously, the surfaces of the moving and stationary contacts will wear down due to prolonged micro-mechanical friction and thermal stress from high current flow, causing a gradual increase in the initial physical gap and a shift in the baseline of the steady-state contact resistance.

[0143] If the system uses the static mapping parameters set at the factory throughout its entire lifecycle, the aforementioned long-term physical aging drift (i.e., low-frequency disturbance) will be misjudged by the orthogonal decoupling engine as abnormal slow-pressing behavior of the operator or a precursor to arcing, which will lead to the failure of the mechanical common mode direction vector of the feedforward calibration library.

[0144] To address the issue of control benchmark inaccuracy caused by the long-term aging of physical equipment, this embodiment, building upon Embodiment 1, further provides a benchmark adaptive evolution method based on the aging characteristics throughout the entire lifecycle. This method can keep the mapping benchmark of the underlying edge computing platform synchronized with the actual aging state of the physical microswitch without consuming computing resources during the critical time window of rapid arc extinguishing. Specifically:

[0145] Specifically, the method provided in this embodiment also includes a baseline adaptive evolution step based on the full life cycle aging characteristics, which specifically includes: during the system idle period after confirming that the arc-free disconnection of the moving and stationary contacts is successful, reading the piezoelectric voltage history sequence of this disconnection action in the pure physical pre-travel stage and the contact microconductivity history sequence in the steady-state conduction stage cached in memory, wherein the pure physical pre-travel stage is defined as the absolute closing time window from the piezoelectric voltage signal response to the initial rise and sudden change of the contact end voltage signal.

[0146] In the above steps, due to limitations in the underlying hardware processing capabilities, the multipliers and logic operation units within the Field Programmable Gate Array (FPGA) or Microcontroller Unit (MCU) operate under extremely high-load pipeline concurrency when performing microsecond-level high-frequency arc extinguishing control. To avoid resource conflicts between the adaptive calculation of theoretical parameters and the actual arc extinguishing data throughput, this embodiment employs strict timing-based peak-shifting scheduling logic. The control strategy stipulates that the system is only allowed to enter a "system idle period" with absolutely abundant computing power after the flexible soft shutdown described in Embodiment 1 has completely ended and the closed-loop disconnection self-test step confirms that the moving and stationary contacts of the microswitch have formed a stable physical breakpoint (i.e., the terminal voltage rises to the bus voltage and the loop current returns to zero).

[0147] During the system's idle period after the charging gun is unplugged, the controller reverse-engineers and retrieves historical sampling data from its internal dual-port random access memory (DPRAM) or circular buffer (FIFO) for the just-completed breaking process. This buffer is designed with a circular overwrite mode, and its storage depth is strictly limited to the number of sampling points matching the longest single press, thus rejecting performance commitments that exceed the limits of the hardware's physical memory.

[0148] It should also be noted that, in order to ensure that the extracted data can objectively reflect the mechanical state without being affected by electrical interference, the system has defined strict physical boundaries for the extracted data window.

[0149] The purely physical pre-travel phase corresponds, at the physical level, to the moment when the operator's finger touches the microswitch button and applies pressure, causing the reed to begin bending deformation, but the moving and stationary contacts still maintain a tight fit. Within this time window, since no microscopic separation has occurred between the contacts, the terminal voltage signal monitored by the terminal voltage sensor remains at an extremely low contact voltage drop level. Once the slope of the terminal voltage signal shows its first abrupt increase, indicating that the microscopic protrusions on the contact surface have begun to peel off, the system records this abrupt change as the end point of the time window.

[0150] By this definition, the piezoelectric voltage history sequence extracted by the system contains pure mechanical stress change information, without the noise introduced by subsequent changes in electrical state. Similarly, the steady-state conduction stage is defined as the stable low voltage drop stage before the terminal voltage signal experiences a sudden rise. The contact microconductivity history sequence extracted here can accurately characterize the effective contact conduction capability of the moving and stationary contacts under rated mechanical pressure.

[0151] Optionally, the system performs discrete integration on the piezoelectric voltage history sequence to obtain the current integral feature value, and compares the current integral feature value with the reference integral feature value recorded in the feedforward calibration library to obtain the stiffness offset coefficient used to characterize the fatigue decay of the micro switch spring.

[0152] In the underlying engineering implementation of this step, the discrete integration operation utilizes a hardware accumulator module to accumulate all discrete voltage sample values ​​within a specified time window with equal weights. According to the physical laws of the piezoelectric effect, the voltage output by the piezoelectric thin film sensor is proportional to the transient deformation stress experienced by the metal spring inside the microswitch; the time-domain integration of this voltage signal within the purely physical pre-stroke stage is physically equivalent to evaluating the mechanical strain energy accumulated on the spring by the external pressing force.

[0153] When a metal spring undergoes fatigue yielding with increasing insertion and extraction cycles, its elastic modulus decreases, resulting in a softening of stiffness. This means that to achieve the same trigger displacement, the mechanical strain energy applied by the operator will decrease, and the corresponding discrete integral result will also decrease proportionally.

[0154] To transform this abstract attenuation law into calculable, specific engineering parameters, the relevant algorithm formula is set as follows:

[0155] ;

[0156] In the formula, This represents the stiffness offset coefficient obtained after comparison and calculation; This is the cumulative index variable used in the discrete integral operation process; This represents the total number of sampling points within the time window corresponding to the pure physical pre-run stage. This represents the instantaneous voltage value in the piezoelectric history sequence extracted from the memory cache; This represents the fixed sampling time step corresponding to the system clock of the field-programmable gate array (FPGA). This represents the reference integral characteristic value that is fixed in the feedforward calibration library at the time of manufacture.

[0157] The output of this formula defines a dimensionless ratio coefficient, which approaches one when the microswitch is brand new; as the device ages, this coefficient gradually decreases according to physical laws. The algorithm's behavior design under extreme conditions needs to consider the scenario where the operator strikes the switch abnormally quickly, in which case the total number of sampling points in the time window will decrease drastically. Evaluating using an integral form rather than a single extreme value form can mitigate measurement noise caused by differences in the speed of single force application, ensuring the rigor of the control strategy logic.

[0158] Specifically, the system then calculates the arithmetic mean of the contact microconductivity history sequence during the steady-state conduction phase to obtain the contact state offset used to characterize the wear state of the contact surface.

[0159] During each closing and opening of the moving and stationary contacts, the silver-based alloy plating on the surface undergoes micron-level mechanical wear, and oxygen in the environment generates an oxide film on the friction surface. The most direct manifestation of these physical changes is an increase in the contact resistance of the contacts, i.e., a decrease in microscopic conductivity.

[0160] To extract objective conductivity from measured data containing electromagnetic noise, the system, during the steady-state conduction phase, sums the historical microconductivity sequences of all contacts using a hardware addition tree and divides the sum by the number of sampling points to obtain the arithmetic mean. The arithmetic mean is used instead of the median or filtered value because addition and shift division can be performed using low-level programmable logic resources in very few clock cycles, resulting in low algorithm complexity that aligns with the real-world computing power environment of edge computing nodes. By subtracting this arithmetic mean from the initial conductivity value calibrated at the factory, the physical parameter characterizing wear thickness and oxidation degree, i.e., the contact state offset, can be obtained.

[0161] like Figure 6 This figure shows the stiffness decay and contact wear distribution of a microswitch during long-term aging. The horizontal axis represents the stiffness offset coefficient (dimensionless), and the vertical axis represents the contact state offset (dimensionless). Figure 6 The pseudo-colored bar on the right indicates the number of insertion / removal cycles, in units of times.

[0162] Figure 6The test data points, which show a large number of aging states that gradually change from dark blue to dark red, objectively record the true physical state characteristics of the device extracted during each idle period of the system within 50,000 insertion and removal test cycles.

[0163] In the initial stage of testing, the dark blue scatter points representing the device's brand-new state were concentrated in the initial area with an x-coordinate of 1.0 and a y-coordinate of 0.0. As the number of insertion and removal cycles increased, the color of the scatter points gradually transitioned to light blue, then yellow, and finally became dark red, while the scatter point group as a whole shifted to the upper left.

[0164] The stiffness offset coefficient on the horizontal axis gradually decreases nonlinearly to around 0.7, reflecting the fatigue decay and stiffness softening of the metal spring inside the microswitch due to long-term stress cycle accumulation. Meanwhile, the contact state offset on the vertical axis climbs to around 0.4, characterizing the decrease in micro conductivity of the moving and stationary contact surfaces due to long-term mechanical friction and oxidation accumulation.

[0165] Throughout the aging evolution path, the measured scatter points exhibit a certain degree of discrete fluctuation, confirming the unavoidable system noise and random microscopic changes in the contact surface that are inherent in engineering measurements. The solid black line running through the scatter point cluster represents the degradation trend front curve, clearly fitting the long-term drift direction of the two physical parameters mentioned above.

[0166] Should Figure 6 This intuitively demonstrates that microswitches exhibit irreversible reference offsets during long-term service, and that using factory static parameters would result in control errors. This provides a physical basis for introducing an adaptive evolution update step into the system, indicating that by extracting and weighting these measured offsets, the control device can smoothly track aging physical trends, ensuring that various judgment thresholds match the actual data throughput and mechanical degradation status, thus maintaining the long-term effectiveness of the contact arc suppression control strategy.

[0167] For example, after obtaining the attenuation and wear characteristics representing the current single insertion and removal action, the system uses a first-order exponential moving average algorithm with a preset forgetting factor to weight and fuse the stiffness offset coefficient and the contact state offset with the historical reference parameters of the previous insertion and removal cycle, respectively, to calculate and generate updated spring stress-displacement nonlinear curve parameters and updated initial gap of the contact closure state.

[0168] For safety-critical underlying electrical control systems, it is strictly prohibited to use complex algorithms that have not been validated through long-term engineering and are essentially black-boxed for parameter updates. If slow-converging, computationally intensive Kalman filters or multi-layer neural network models are used, they are not only likely to exceed the capacity of the microcontroller's static random access memory (SRAM), but also, when encountering an abnormal operation with a severe mechanical impact (i.e., a poisoned sample), the complex iterative algorithm may produce unpredictable state jumps that violate physical laws.

[0169] Therefore, this embodiment uses the first-order exponential moving average (EMA) algorithm. This algorithm is a lightweight autoregressive moving average model, and its network structure is a single-node first-order differential feedback. It does not require storing a large number of historical state matrices, but only needs to retain the previous fusion result. The training data source is the real-time sensor features extracted after a single plug-in / plug-out.

[0170] For example, to clearly illustrate the integration of the algorithm with specific physical scenarios, a weighted fusion formula for updating the initial gap of the contact closure state is given here:

[0171] ;

[0172] In the formula, This represents the initial gap used to calculate and generate the updated contact closure state for this cycle. Represents a preset forgetting factor with limited weights; This represents the historical baseline initial gap left over from the previous insertion / removal cycle, i.e., the initial gap of the preset contact closure state currently used by the system in the real-time control phase in Example 1. This represents the initial clearance at the origin set when the equipment leaves the factory; This represents a pre-specified electrical wear loss conversion factor, used to linearly map changes in electrical conductance to changes in physical spatial distance. This represents the contact state offset obtained in step three of this cycle.

[0173] In the actual hardware circuit implementation, in order to make the algorithm objectively feasible for engineering applications, the preset forgetting factor is usually set to a negative integer power of two (for example, a value of 0.015625, i.e., ...). ).

[0174] The significant advantage of this hyperparameter setting is that the field-programmable gate array (FPGA) does not need to call up a large floating-point multiplier core; it can complete the multiplication operation simply by performing a hardware shift operation to the right on the data bus (such as shifting it six bits to the right). This setting not only speeds up the execution of the algorithm, but also allows the algorithm to perform extremely strong low-pass filtering attenuation on drastic data jumps that occur during a single insertion or removal.

[0175] Only when the microswitch has undergone hundreds of wear cycles, resulting in a long-term, stable, monotonic change in the contact state offset, will this trend slowly accumulate and be reflected in the final updated parameters. Similarly, the system uses the same first-order exponential moving average structure to incorporate the stiffness offset coefficient of the current cycle into the parameters of the reed stress-displacement nonlinear curve (such as the slope and offset term of the curve).

[0176] like Figure 7The graph illustrates the adaptive convergence process of the baseline parameters based on the first-order exponential moving average algorithm. The horizontal axis represents the number of insertion / removal cycles (in seconds), and the vertical axis represents the initial gap offset (in micrometers).

[0177] Figure 7 The curve includes two comparison curves: the light blue curve represents the measured gap offset during a single insertion / removal, and the red curve represents the updated initial gap offset. The light blue curve shows a slow upward trend, reflecting the physical degradation process where wear on the microswitch contact surface leads to a gradual increase in the gap.

[0178] Throughout the entire 2,000 test cycle, the light blue curve exhibits high-frequency fluctuations of approximately 3 to 5 micrometers around the main trend line. Furthermore, dramatic data jumps of up to 75 micrometers and nearly -20 micrometers occur around the 500th and 1200th tests, respectively, objectively simulating the sensor measurement noise floor and abnormal interference samples caused by extreme mechanical impacts during actual service.

[0179] The red curve performs a weighted fusion process on the light blue curve. This calculation utilizes a preset minimum forgetting factor, which is equivalent to the following at the hardware level: Shift operation. In terms of waveform transformation, the red curve maintains a smooth and monotonous convergence trend, gradually tracking from the initial zero micrometers to around 35 micrometers at the end of the test.

[0180] Faced with the high-frequency noise of the light blue curve, the red curve demonstrates a good smoothing effect; especially at the points where the light blue curve experiences sharp jumps, the red curve does not exhibit a follow-up abrupt change, but effectively isolates the interference of abnormal samples and maintains the original physical convergence direction. This graphical comparison verifies that the selected exponential moving average algorithm can accurately filter out random errors while avoiding resource conflicts with the actual data throughput of the underlying hardware, ensuring that the benchmark parameters copied to the feedforward calibration library conform to the actual physical development laws.

[0181] It should be noted that after the calculation is completed, the system copies the updated spring stress-displacement nonlinear curve parameters and the updated initial gap of the contact closure state into the feedforward calibration library as a reference for the next insertion and removal cycle.

[0182] In the underlying hardware implementation, the feedforward calibration library is mapped into data pages of the electrically erasable read-only memory (EEPROM) or built-in flash memory on the control board. Considering the physical limits of the storage medium's erase / write lifespan, the system does not forcibly initiate erase / write commands during every idle period. Instead, it only performs a rewrite operation by issuing page programming instructions via the Serial Peripheral Interface (SPI) when the difference between the updated parameters and the parameters recorded in the library exceeds a set register dead-time threshold. This protection mechanism conforms to the manufacturing constraints of real printed circuit board (PCB) memory, preventing memory block damage due to frequent rewriting.

[0183] Meanwhile, to ensure the control strategy logic is rigorous and sound, the system performs a safety boundary limit check before executing the overwrite operation. If the calculated initial gap of the updated contact closure state exceeds the physical limit allowed by the geometry of the microswitch housing, or if the updated spring stress-displacement nonlinear curve parameters exhibit logical contradictions such as negative stiffness that violate the laws of physical material mechanics, the system will determine that the currently acquired historical sequence is invalid data due to interference. In this case, the system will discard the update request, maintain the original benchmark in the feedforward calibration library, and may selectively report sensor fault codes.

[0184] Based on existing technology analysis, traditional microswitch control systems for charging guns typically employ fixed physical thresholds or unchanging electrical compensation parameters. As the equipment ages, the softening of the reeds and contact erosion cause the actual physical boundaries to gradually deviate from the factory reference, leading to false alarms or failure to operate, thus reducing the switch's overall lifecycle reliability.

[0185] The method provided in this embodiment and the orthogonal decoupling control method in Embodiment 1 support each other functionally and produce significant beneficial effects. By reading historical sensing data from the purely physical stage during the absolute system idle period without arc interference, the system can objectively quantify the mechanical stiffness decay and contact conductivity wear of the microswitch caused by long-term service. Utilizing a highly hardware-adaptive first-order exponential moving average algorithm, the system successfully isolates these long-cycle aging trends and smoothly integrates and updates them into the underlying calibration reference. This adaptive evolution scheme can continuously calibrate the mechanical common-mode matrix required by the orthogonal projection engine throughout the entire service life of the equipment, significantly reducing the risk of sensing inaccuracies caused by equipment aging drift. This allows the high-frequency arc extinguishing control logic to maintain high-precision precursor extraction capabilities even after tens of thousands of wear cycles, demonstrating high engineering practicality.

[0186] Example 3:

[0187] Based on the technical problem mentioned in the background art, where the micro switch of the charging gun, under complex or non-standard pressing conditions, suffers from mechanical bouncing of the moving and stationary contacts within a critical distance due to the difficulty in establishing the designed breaking speed, leading to arc erosion, this embodiment provides a specific system architecture solution. This system, in the form of physical hardware modules and embedded logic circuits, implements a control device capable of intervening and transferring current within a microsecond-level time window.

[0188] For example, this embodiment discloses a contact arc suppression control system for a charging gun microswitch. The main control computing platform of the system adopts an industrial-grade field-programmable gate array (FPGA) chip. Since the FPGA has a pure hardware parallel processing architecture, it avoids the time uncertainty of software interrupt response in traditional microcontrollers, thereby providing a strictly nanosecond-level execution cycle and ensuring that the control strategy logic does not violate the strict requirements of high-speed physical arc extinguishing.

[0189] Specifically, this embodiment provides a contact arc suppression control system for a charging gun micro switch, comprising: a data acquisition module for acquiring real-time composite sensing data of the charging gun micro switch, wherein the real-time composite sensing data includes piezoelectric voltage signals collected by a micro-nano level piezoelectric sensing unit, contact terminal voltage signals collected by a pole terminal voltage sensor, and loop current signals collected by a loop current sampling unit.

[0190] In actual engineering manufacturing and equipment layout, the front-end physical layer of the data acquisition module includes a multi-channel synchronously triggered high-speed analog-to-digital converter (ADC). The micro-nano-level piezoelectric sensing unit uses an aluminum nitride piezoelectric film deposited on the deformation region of the metal spring root of the microswitch using microelectromechanical systems (MEMS) technology to convert mechanical stress into an analog voltage quantity in real time; the terminal voltage sensor uses an isolated differential sampling circuit with a high common-mode rejection ratio, and its two probe terminals are directly soldered to the two ends of the stationary contact pin and the moving contact pin of the microswitch; the loop current sampling unit uses a precision alloy shunt resistor with extremely low temperature coefficient and extremely low parasitic inductance, connected in series in the charging control loop.

[0191] Driven by a fixed clock source provided by the FPGA, the analog-to-digital converter samples the three sensors in parallel with a fixed sampling time step of 0.5 microseconds. This ensures that the acquired physical signals are strictly synchronized on the time axis, avoiding timing deviations that could lead to logical inconsistencies in the algorithm.

[0192] Optionally, the control system further includes a parameter matrix construction module, used to construct a real-time composite electromechanical contact parameter matrix containing piezoelectric channel characteristics, terminal voltage channel characteristics and contact conductivity channel characteristics based on the real-time composite sensing data.

[0193] At the underlying hardware logic implementation level, the parametric matrix construction module calls the shift register and fixed-point divider IP cores inside the FPGA. Due to the significant differences in the physical dimensions and numerical ranges of the piezoelectric signal, terminal voltage signal, and current signal, this module first uses the full-scale reference value preset in the read-only memory (ROM) to perform normalized division operations on the three signals.

[0194] Meanwhile, to prevent division-by-zero overflow errors when the terminal voltage is close to zero, a hardware comparator is configured outside the divider, enabling the Ohm's law conductance calculation path only when the terminal voltage is higher than the quantization noise floor. Subsequently, the module arranges the three normalized eigenvalues ​​column-wise and writes them into the designated on-chip block random access memory (BRAM) address space, forming a three-dimensional state matrix for subsequent algorithm calls. This structured processing of input data establishes a rigorous correspondence between physical signals and abstract algorithms.

[0195] Furthermore, the control system includes an orthogonal projection decoupling module, which is used to extract the mechanical common mode subspace in the real-time composite electromechanical contact parameter matrix based on the pre-calibrated spring stress-displacement nonlinear curve, and to separate the pure arc initiation precursor critical residual characterizing the arc initiation precursor of the moving and stationary contacts through orthogonal projection decoupling.

[0196] This module is the core hardware accelerator for solving the problem of overlapping mechanical jitter interference and electrical precursors. In the hardware mapping scheme, the pre-calibrated reed stress-displacement nonlinear curve is stored in a non-volatile memory area in the form of a one-dimensional lookup table (LUT). The addressing control logic within the module takes a low-frequency mechanical voltage component as input and outputs the corresponding estimated value of the reed's physical displacement within one clock cycle.

[0197] Subsequently, the module invokes the FPGA's internal digital signal processing (DSP) logic chip to perform matrix multiplication and addition operations. After obtaining the mechanical common-mode direction vector corresponding to the current displacement through a linear interpolator, the DSP logic chip completes the construction of the orthogonal projection matrix and vector subtraction operations within three system clock cycles. This operation physically filters out the signal common-mode components caused by external hand vibrations or the standing waves of the reed itself, and the output residual data purely reflects the resistivity abrupt changes and field emission effects at the microscopic protrusions on the surfaces of the moving and stationary contacts. The hardware overhead of this processing is strictly limited to a single sampling period, which is consistent with the actual computing power limit of the selected computing platform.

[0198] It should be noted that the control system also includes an arc initiation precursor determination and triggering module, which is used to make an advance trend compensation judgment on the critical residual of the pure arc initiation precursor. When the preset micro-bounce arc initiation triggering condition is met, a conduction command is sent to the solid-state bypass arc extinguishing circuit connected in parallel at both ends of the moving and stationary contacts.

[0199] In the specific design of this module, the algorithm features are closely integrated with the actual execution mechanism. For the calculation of the L2 norm of the critical residual before the pure arc initiation, the module adopts the coordinate rotating digital computer (CORDIC) hardware architecture, which completes the square and square root operations through basic shift and addition / subtraction operations, significantly reducing hardware resource consumption.

[0200] Subsequently, the rate of change is obtained by first-order differential calculation using historical values ​​stored in registers, and a leading trend is determined by combining this with a preset threshold for the amplitude of the arc precursor. When the criterion scalar obtained from the weighted calculation crosses zero, the hardware comparator immediately undergoes a level flip, generating a high-priority hardware interrupt signal or directly outputting a transistor-to-transistor logic (TTL) level turn-on instruction to the subsequent driver circuit.

[0201] For example, the control system includes a bypass current transfer module for linearly transferring the control circuit current flowing through the moving and stationary contacts to the solid-state bypass arc extinguishing circuit before the moving and stationary contacts physically separate, so that the moving and stationary contacts physically separate in a zero-current state.

[0202] In the objective engineering implementation and modification scheme, the solid-state bypass arc extinguishing circuit is welded in parallel to the two electrical terminals of the microswitch. The core actuator of the solid-state bypass uses a silicon carbide metal oxide semiconductor field-effect transistor (SiC MOSFET) with extremely low on-resistance and microsecond-level switching characteristics.

[0203] When the arc initiation precursor detection and triggering module issues a turn-on command, the isolated gate driver with desaturation short-circuit protection injects drive current into the gate of the SiC MOSFET. As the device enters the turn-on state, the potential difference between the moving and stationary contacts of the microswitch is instantaneously clamped to the on-state voltage drop of the SiC device (typically less than one volt).

[0204] Because the clamping voltage is much lower than the ionization voltage required to sustain the arc, the current in the control circuit is forced to rapidly transfer to the solid-state bypass at a linear slope. At this time, the transmission mechanism inside the microswitch continues to perform its mechanical breaking action, but the moving and stationary contacts open the gap in a physical environment of zero current and extremely low potential difference, effectively blocking the physical conditions for arc formation.

[0205] Finally, the control system includes a flexible soft shutdown module, which controls the solid-state bypass arc extinguishing circuit to perform flexible soft shutdown at a preset shutdown slope after calculating and determining that the physical gap between the moving and stationary contacts reaches a preset safe insulation distance.

[0206] To avoid the generation of a destructive back electromotive force (i.e., due to the instantaneous interruption of current) on the parasitic inductance (Effect) After confirming that the physical gap between the contacts has crossed the safety insulation boundary for air breakdown, this module does not adopt a hard shutdown strategy.

[0207] In practice, the flexible soft-shutdown module controls the pull-in / sink-out current output capability of the isolated gate driver (e.g., by switching multiple gate resistor levels or outputting a pulse-width modulation signal with a gradually changing duty cycle to achieve active gate voltage clamping), allowing the drain-source current of the SiC MOSFET to decrease smoothly according to the automotive electromagnetic compatibility (EMC) limits. This design objectively mitigates the electromagnetic overshoot phenomenon at the moment of power failure, preventing radiated interference from affecting the charging handshake protocol communication bus.

[0208] In terms of existing technology, traditional charging gun microswitches based on purely physical transmission disconnection mechanisms are prone to a vicious cycle of frequent friction between moving and stationary contacts and repeated arc ignition when slowly pressed by an operator, leading to rapid contact ablation or even welding. The contact arc suppression control system provided in this embodiment, by introducing a hardware computing platform with nanosecond-level computing power and an orthogonal decoupling model, accurately identifies the pre-arc ignition precursor within an extremely short physical gap; and combined with a parallel solid-state bypass circuit, it completes the active transfer of current before the micro-motion contacts actually separate.

[0209] The entire solution effectively suppresses arc erosion caused by improper operation without compromising the original mechanical breakpoint safety redundancy design of the microswitch. At the same time, it limits electromagnetic interference through a flexible shutdown strategy, effectively improving the electrical life and system reliability of the underlying safety interlocking execution unit of the charging gun.

[0210] Example 4:

[0211] Corresponding to the above embodiments, the present invention also proposes an electronic device.

[0212] like Figure 9 The diagram shows a structural schematic of an electronic device according to the present invention. The electronic device 100 includes a processor 101 and a memory 103. The processor 101 and the memory 103 are connected, for example, via a bus 102. Optionally, the electronic device 100 may further include a transceiver 104. It should be noted that in practical applications, the transceiver 104 is not limited to one unit, and the structure of this electronic device 100 does not constitute a limitation on the embodiments of the present invention.

[0213] Processor 101 may be a CPU, a general-purpose processor, a DSP, an ASIC, an FPGA, or other programmable logic device, transistor logic device, hardware component, or any combination thereof. It may implement or execute the various exemplary logic blocks, modules, and circuits described in connection with this disclosure. Processor 101 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0214] Bus 102 may include a path for transmitting information between the aforementioned components. Bus 102 may be a PCI bus or an EISA bus, etc. Bus 102 may be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 9 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0215] The memory 103 stores a computer program corresponding to a contact arc suppression control method for a micro switch of a charging gun according to the above embodiments of the present invention. This computer program is executed by the processor 101. The processor 101 executes the computer program stored in the memory 103 to implement the content shown in the aforementioned method embodiments.

[0216] Among them, electronic devices 100 include, but are not limited to: mobile terminals such as laptops and PADs (tablet computers) and fixed terminals such as desktop computers. Figure 9 The electronic device 100 shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of the present invention.

[0217] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for controlling the contact arc suppression of a micro switch in a charging gun, characterized in that, include: The real-time composite sensing data of the charging gun micro switch is acquired. The real-time composite sensing data includes the piezoelectric voltage signal collected by the micro-nano piezoelectric sensing unit, the contact terminal voltage signal collected by the electrode terminal voltage sensor, and the loop current signal collected by the loop current sampling unit. Based on the real-time composite sensing data, a real-time composite electromechanical contact parameter matrix is ​​constructed, which includes piezoelectric channel characteristics, terminal voltage channel characteristics, and contact conductivity channel characteristics. Based on the pre-calibrated spring stress-displacement nonlinear curve, the mechanical common mode subspace in the real-time composite electromechanical contact parameter matrix is ​​extracted, and the pure arc initiation precursor critical residual characterizing the arc initiation precursor of the moving and stationary contacts is separated by orthogonal projection decoupling. The pre-arc threshold residual of the pure arc initiation is judged by advance trend compensation. When the preset micro-bounce arc initiation triggering condition is met, a conduction command is sent to the solid-state bypass arc extinguishing circuit connected in parallel at both ends of the moving and stationary contacts. Before the moving and stationary contacts physically separate, the control circuit current flowing through the moving and stationary contacts is linearly transferred to the solid-state bypass arc extinguishing circuit, so that the moving and stationary contacts physically separate in a zero-current state. Once the physical gap between the moving and stationary contacts is calculated and determined to reach the preset safe insulation distance, the solid-state bypass arc extinguishing circuit is controlled to perform a flexible soft shutdown according to the preset shutdown slope.

2. The method according to claim 1, characterized in that, The construction of the real-time composite electromechanical contact parameter matrix includes: The piezoelectric voltage signal, the contact terminal voltage signal, and the loop current signal are each normalized according to their respective full-scale reference values. When the absolute value of the contact terminal voltage signal is greater than or equal to the preset effective measurement lower limit, the contact microconductivity is calculated in real time using the loop current signal and the contact terminal voltage signal according to Ohm's law. The normalized piezoelectric voltage signal is used as the piezoelectric channel feature, the normalized contact terminal voltage signal is used as the terminal voltage channel feature, and the normalized contact microconductivity is used as the contact conductivity channel feature. These are arranged in columns to construct a three-dimensional real-time composite electromechanical contact parameter matrix.

3. The method according to claim 1, characterized in that, Before decoupling and separating the pure arc-starting precursor critical residual through the orthogonal projection, the method also includes a step of obtaining the current reed displacement estimate, specifically: The piezoelectric voltage signal is processed by a preset sliding average window to extract the low-frequency mechanical component; By combining the preset temperature compensation coefficient function, the low-frequency mechanical components are mapped and calculated using the reed stress-displacement nonlinear curve to obtain the corresponding estimated value of the current reed displacement. The estimated value of the current reed displacement is used as the mechanical state reference and the scheduling variable for subsequent online interpolation.

4. The method according to claim 3, characterized in that, The process of decoupling and separating the pure arc-starting precursor critical residual characterizing the arc-starting precursor of the moving and stationary contacts through orthogonal projection includes: Calculate the slow-varying trend vector of the real-time composite electromechanical contact parameter matrix within the preset sliding mean window; Based on the current estimated reed displacement, linear piecewise interpolation is performed between multiple preset calibration node displacements to calculate the mechanical common mode direction vector that adapts to the reed displacement. The mechanical common mode direction vector is combined with the slowly varying trend vector to construct the mechanical common mode subspace basis matrix. ,in This is the current discrete-time sampling sequence number; Based on preset regularization coefficients The least squares projection algorithm utilizes the identity matrix. The basis matrix of the subspace co-mode with the mechanical system is calculated and output according to the formula. The orthogonal pure arc-starting precursor critical residual vector To remove common-mode artifacts caused by mechanical vibration and friction, the formula is: ； in, This represents the transpose operation of a matrix. The operation represents the inversion of a matrix. This is the real-time composite electromechanical contact parameter matrix.

5. The method according to claim 1, characterized in that, The step of performing advanced trend compensation judgment on the pure arc initiation precursor critical residual, when the preset micro-bounce arc initiation trigger condition is met, includes: Calculate the vector L2 norm of the critical residual of the pure arc initiation precursor to obtain the scalar characteristic of the arc initiation precursor. ,in This is the current discrete-time sampling sequence number; For the arc initiation precursor scalar characteristics Perform first-order difference calculations to obtain the rate of change of the precursor scalar characteristics. ; Combined with the preset arc-starting precursor amplitude threshold and the preset precursor rise time constant The scalar of the arc-starting red line criterion is calculated using the following formula. : ； When the arc-starting red line criterion scalar When the value is greater than or equal to zero, it is determined that the preset micro-bounce arc triggering condition is met.

6. The method according to claim 3, characterized in that, Before sending the activation command, a transfer feasibility constraint determination must be performed, including: Calculate the estimated value of the current reed displacement. Differential rate of change ,in The current discrete-time sampling sequence number; using a preset reed displacement separation threshold. The remaining separation time can be predicted and calculated using the following formula. : ； Determine the remaining separation time If the current transfer time is greater than the sum of the preset transfer conduction time and the preset safety margin time, then the linear transfer is performed within the time window before physical separation to ensure that the current transfer is completed before the physical separation time of the moving and stationary contacts.

7. The method according to claim 3, characterized in that, The control of the solid-state bypass arc extinguishing circuit to perform flexible soft shutdown includes: Utilizing the initial gap of the preset contact closure state The difference between the current estimated reed displacement and the preset nominal displacement of the closed state is superimposed to calculate the gap between the moving and stationary contacts in real time. When the gap between the moving and stationary contacts is greater than or equal to the preset safety insulation distance, and the state continues to meet the preset holding step number, it is determined that the safety insulation distance has been crossed. The safety insulation distance is determined by multiplying the air breakdown critical distance corresponding to the control circuit bus voltage by a preset safety margin coefficient. Triggering a flexible soft-shutdown command causes the real-time bypass current in the solid-state bypass arc extinguishing circuit to... The controlled turn-off ramp descends to zero, and the controlled turn-off ramp is subject to a preset upper limit for the rate of change of current. The limiting constraint satisfies the following limiting configuration formula: ； in, For the real-time bypass current The time rate of change, with the absolute value sign indicating a limit on the magnitude of the rate of change; After the solid-state bypass arc extinguishing circuit performs a flexible soft shutdown to reduce the current to zero, a closed-loop disconnection self-test step is also included: The system detects whether the voltage signal at the contact point is greater than or equal to a preset disconnection confirmation voltage threshold, which is set based on the bus voltage of the control circuit. Verify that the loop current signal remains zero; If the contact voltage signal is greater than or equal to the preset disconnection confirmation voltage threshold, and the circuit current signal remains zero, then the arc-free disconnection is confirmed to be successful and the system state is reset to wait for the next plugging and unplugging action. If the contact terminal voltage signal is less than the preset disconnection confirmation voltage threshold, or the circuit current signal is not kept at zero, a contact welding adhesion fault is determined to have occurred, the charging enable command is blocked, and an adhesion alarm signal is output.

8. The method according to claim 4, characterized in that, The preset displacements of multiple calibration nodes and their corresponding mechanical common-mode direction vectors are obtained in the following way: During the calibration phase, standard frequency sweep mechanical disturbances are applied at multiple calibration node displacements. The response ratios of the piezoelectric channel characteristics, the terminal voltage channel characteristics, and the contact conductivity channel characteristics under the standard frequency sweep mechanical disturbance are simultaneously measured, and the mechanical common mode direction vectors corresponding to each calibration node are generated and stored in the feedforward calibration library of the underlying edge computing platform. The orthogonal projection decoupling process is limited to a fixed-scale matrix pipeline operation to ensure that the time required for computing power overhead is less than or equal to the single sampling period of the underlying edge computing platform.

9. The method according to claim 1, characterized in that, The process of linearly transferring the control loop current to the solid-state bypass arc-extinguishing loop includes: Trigger the power solid-state switch in the solid-state bypass arc extinguishing circuit to turn on; Within the preset transfer conduction time, the shunting ratio of the physical moving and stationary contacts decreases at a slope until the current of the physical moving and stationary contacts returns to zero, and the voltage across the moving and stationary contacts is clamped below the preset minimum arc-starting voltage of the contact material. From the time the conduction command is sent until the flexible soft shutdown ends, the solid-state bypass arc extinguishing circuit remains in a continuous conducting state, so that the conduction action triggered once covers the entire mechanical high-frequency bounce range caused by the separation of physical moving and stationary contacts.

10. A contact arc suppression control system for a micro switch of a charging gun, characterized in that, include: The data acquisition module is used to acquire real-time composite sensing data of the charging gun micro switch. The real-time composite sensing data includes piezoelectric voltage signal collected by the micro-nano piezoelectric sensing unit, contact terminal voltage signal collected by the electrode terminal voltage sensor, and loop current signal collected by the loop current sampling unit. The parameter matrix construction module is used to construct a real-time composite electromechanical contact parameter matrix containing piezoelectric channel features, terminal voltage channel features and contact conductivity channel features based on the real-time composite sensing data. The orthogonal projection decoupling module is used to extract the mechanical common mode subspace in the real-time composite electromechanical contact parameter matrix based on the pre-calibrated spring stress-displacement nonlinear curve, and to separate the pure arc initiation precursor critical residual characterizing the arc initiation precursor of the moving and stationary contacts through orthogonal projection decoupling. The arc initiation precursor determination and triggering module is used to make an advance trend compensation judgment on the critical residual of the pure arc initiation precursor. When the preset micro-bounce arc initiation triggering condition is met, a conduction command is sent to the solid-state bypass arc extinguishing circuit connected in parallel at both ends of the moving and stationary contacts. The bypass current transfer module is used to linearly transfer the control circuit current flowing through the moving and stationary contacts to the solid-state bypass arc extinguishing circuit before the moving and stationary contacts physically separate, so that the moving and stationary contacts physically separate in a zero-current state. The flexible soft shutdown module is used to control the solid-state bypass arc extinguishing circuit to perform flexible soft shutdown according to a preset shutdown slope after calculating and determining that the physical gap between the moving and stationary contacts reaches a preset safe insulation distance.

Images