Zero differential blind mate connector system

Abstract

The application relates to the field of high-power electrical connection and provides a zero-voltage-difference blind insertion connector system, which comprises a first connector, a second connector, a controller and a pre-charge circuit. The first connector is provided with a first insulating end face, at least one power supply pin exposed from the first insulating end face and an ultrasonic phased array sensor embedded in the first insulating end face. The second connector is provided with a second insulating end face and a power supply socket corresponding to the power supply pin and exposed from the second insulating end face. The second insulating end face is provided with an acoustic reflection array opposite to the ultrasonic phased array sensor. The controller is electrically connected with the ultrasonic phased array sensor to receive a sensing signal representing the relative pose of the first connector and the second connector and generate a trigger instruction based on the sensing signal. The pre-charge circuit is configured to adjust the voltage difference between the power supply pin and the power supply socket to below a preset safety threshold before the power supply pin and the power supply socket are physically contacted in response to the trigger instruction. In this way, the zero-voltage-difference blind insertion of the connector is realized without spark under strong electromagnetic interference.

Description

Technical Field

[0001] This invention relates to the technical field of high-power electrical connections, and more specifically to a zero-differential-voltage blind mating connector system. Background Technology

[0002] In energy storage and power scenarios involving parallel multi-level battery modules, operators frequently need to perform blind mating operations on high-power connectors on-site. Blind mating involves holding one connector and attempting to mate it with another connector already fixed to the equipment, without seeing the specific positions of the pins and sockets, relying entirely on feel. Under this method, if there is a voltage difference between the two ends of the pins before contact, an electrical spark will be generated at the moment of contact. The greater the voltage difference and the higher the system power, the more severe the sparking, which can not only burn the surface of the pins but also cause safety accidents.

[0003] To suppress arcing, a common practice is to incorporate a long and short pin structure on the connector, allowing the ground pin to contact first, thus bringing the potentials of the two ends closer together, before the power pin contacts. The problem with this approach is that the long and short pins only provide a fixed time difference, which may not be sufficient to eliminate the voltage difference below the safe range. This is especially true when the battery's state of charge varies significantly, or when the system voltage platform is high, the protection provided by the fixed time difference becomes inadequate.

[0004] Later, the pre-charging concept emerged, which involves using an auxiliary circuit to pass a small current before the pins actually make contact, gradually leveling out the voltage across the terminals. This requires knowing when the two connectors approach a sufficiently close distance to initiate pre-charging in a timely manner. Common distance sensing methods include capacitive proximity sensing and photoelectric sensors. While capacitive sensing works well in the laboratory, in actual battery stacks, the strong electromagnetic field generated by the high-current charging and discharging completely overwhelms the signal from the capacitive sensor, resulting in inconsistent distance readings and making it unreliable to trigger pre-charging. Photoelectric sensors are highly sensitive to dust and oil, conditions often present in the operating environment of the connectors.

[0005] Another thorny issue is that the hand cannot be held perfectly straight during blind insertion. The connector will pitch, roll, and shift laterally as it approaches; if it is forced in at an angle, the pins will wear off or hit the socket face. Even if the voltage is balanced in this off-center state, the contact resistance will be higher due to insufficient contact area, causing heat during operation and eventually leading to problems.

[0006] Therefore, how to obtain the precise distance and attitude deviation between the connectors simultaneously without relying on electrical quantity sensing in the environment of strong electromagnetic interference, and how to complete the voltage equalization before the pins actually make contact, under the premise that both of these information are reliable, has become a key problem that needs to be solved in this field. Summary of the Invention

[0007] Based on the above-mentioned defects, one of the objectives of this invention is to achieve zero-differential-voltage spark-free blind mating of connectors under strong electromagnetic interference.

[0008] To address the aforementioned objectives, in one embodiment of the present invention, a zero-difference-voltage blind-mating connector system is proposed, comprising: a first connector having a first insulating end face, at least one power pin exposed from the first insulating end face, and an ultrasonic phased array sensor embedded in the first insulating end face; a second connector having a second insulating end face, a power socket exposed from the second insulating end face corresponding to the power pin, wherein an acoustic reflection array opposite to the ultrasonic phased array sensor is disposed on the second insulating end face; a controller electrically connected to the ultrasonic phased array sensor to receive a sensing signal characterizing the relative pose of the first connector and the second connector, acquired by the ultrasonic phased array sensor by emitting ultrasonic waves to the acoustic reflection array and receiving the echoes, and generating a trigger command based on the sensing signal; and a pre-charging circuit having an equalization path electrically connected to the power pin and the power socket, the pre-charging circuit being electrically connected to the controller and configured to, in response to the trigger command, adjust the voltage difference between the power pin and the power socket to below a preset safety threshold before the power pin and the power socket physically contact.

[0009] In this solution, the ultrasonic phased array sensor acquires the relative pose between the first and second connectors in a non-contact state by emitting ultrasonic waves into the acoustic reflection array and receiving the echoes. This acoustic sensing method fundamentally eliminates the influence of strong electromagnetic interference from the battery stack on the ranging signal. The controller generates the trigger command based on the sensing signal before the pins make contact. The pre-charging circuit then adjusts the voltage difference between the power pins and the power socket to below the preset safety threshold before physical contact, ensuring that there is no potential difference sufficient to trigger an arc at the moment of electrical contact. The equalization path provides a controlled path for the pre-charging current, allowing for smooth energy transfer. Therefore, the system can reliably achieve zero-difference spark-free blind mating even in environments with strong electromagnetic interference.

[0010] Optionally, in some embodiments of the system, the ultrasonic phased array sensor includes: a ring piezoelectric transducer array, embedded in the periphery of the first insulating end face, and composed of multiple independent array elements evenly distributed along the circumferential direction; a multi-channel transceiver control chip, electrically connected to the ring piezoelectric transducer array, configured to apply pulse excitation with a predetermined phase delay to each of the independent array elements; and an acoustic matching layer covering the front end face of the ring piezoelectric transducer array; wherein the acoustic reflection array includes multiple miniature corner reflectors; the acoustic reflection array further includes a ring-shaped metal target disk, embedded in the second insulating end face, and positioned corresponding to the ring piezoelectric transducer array in the axial direction.

[0011] Optionally, in some embodiments of the system, the pre-charge circuit includes: a pre-contact electrode disposed on the first connector, the front end of which protrudes axially from the front end of the power pin; a programmable slope constant current source electrically connected to the pre-contact electrode and the controller, configured to inject or draw current into the equalization path at a controlled current slope in response to the trigger command; the programmable slope constant current source includes: a digital-to-analog converter electrically connected to the controller; an operational amplifier, the non-inverting input of which is electrically connected to the output of the digital-to-analog converter; a power metal-oxide-semiconductor field-effect transistor array, the gate of which is electrically connected to the output of the operational amplifier, the drain of which is electrically connected to the equalization path, and the source of which is grounded through a sampling resistor, the voltage across the sampling resistor being fed back to the inverting input of the operational amplifier to form a closed-loop constant current control; and a second type of compensation network connected between the output and the inverting input of the operational amplifier to make the phase margin of the closed-loop constant current control greater than a predetermined angle.

[0012] Optionally, in some embodiments of the system, the programmable slope constant current source further includes: a ramp generator electrically connected to the digital-to-analog converter and the operational amplifier, including an analog switch connected in series between the constant current source and the integrating capacitor, the control terminal of the analog switch being electrically connected to the controller; a ramp reset switch connected in parallel with the integrating capacitor, the control terminal of the ramp reset switch being electrically connected to the controller; the controller is configured to close the analog switch when the trigger command is generated, the constant current source charges the integrating capacitor with a constant current, causing a linearly rising ramp voltage to be generated across the integrating capacitor, the ramp voltage being buffered by the operational amplifier and driving the power metal-oxide-semiconductor field-effect transistor array; and, when the trigger command is released, closing the ramp reset switch to discharge the integrating capacitor.

[0013] Here, the ramp generator closes the analog switch when the trigger command is generated. The constant current source charges the integrating capacitor with a constant current, generating a linearly rising ramp voltage. This ramp voltage, after being buffered by the operational amplifier, drives the power metal-oxide-semiconductor field-effect transistor array, causing the pre-charge current to increase at a controllable constant slope, avoiding damage to the power pins and the power socket from step current surges. The ramp reset switch discharges the integrating capacitor when the trigger command is released, providing the same initial conditions for the next pre-charge operation and ensuring consistency of repeated operations. The controller's timing control of the analog switch and the ramp reset switch ensures complete control over the current rise and fall processes, thereby minimizing electromagnetic interference and thermal stress during the pre-charge process and ensuring the stability of zero-difference voltage access.

[0014] Optionally, in some embodiments of the system, the front end face of the pre-contact electrode is provided with a plurality of micro-protrusion structures. The plurality of micro-protrusion structures are evenly distributed along the circumference of the pre-contact electrode. Each micro-protrusion structure has a conical tip with a radius of curvature of less than a preset radius value. When the pre-contact electrode and the corresponding electrode are in physical contact, the oxide film layer on the surface of the corresponding electrode is broken with concentrated pressure, thereby reducing the initial contact resistance.

[0015] Optionally, in some embodiments of the system, the pre-charge circuit further includes: a fast discharge branch connected in parallel between the power pin and the reference ground, including a discharge resistor and a discharge switch connected in series; a constant current discharge circuit connected in series between the discharge resistor and the discharge switch, consisting of a depletion-type metal-oxide-semiconductor field-effect transistor and a gate bias resistor; the control terminal of the discharge switch is electrically connected to the controller, the controller being configured to briefly close the discharge switch before the trigger command is generated, so as to discharge the residual electrostatic charge on the power pin in the form of a constant current.

[0016] Optionally, in some embodiments, the system further includes: an insulation resistance monitoring circuit, comprising a voltage divider resistor network connected in series between the power pin and a reference ground, and an isolated analog-to-digital converter connected across a sampling resistor in the voltage divider resistor network; the isolated analog-to-digital converter is electrically connected to the controller and is used to provide the controller with DC leakage current data characterizing the insulation resistance between the first connector and the second connector before the power pin makes physical contact with the power socket; the controller is further operable to: suppress the generation of the trigger command when the DC leakage current data indicates that the insulation resistance is lower than a preset insulation threshold.

[0017] Optionally, in some embodiments of the system, the controller is operable to: calculate the six-degree-of-freedom relative pose between the first connector and the second connector based on the sensing signal, the six-degree-of-freedom relative pose including axial distance, horizontal offset, vertical offset, pitch angle, roll angle, and yaw angle; before generating the trigger command, construct a state vector based on the time series of the axial distance, using the axial distance, relative approach velocity, and relative approach acceleration, recursively derive the state vector with a constant rate of change of acceleration constraint, and iteratively estimate the relative approach velocity using the time series of the axial distance as the observation; generate the trigger command in advance of the axial distance reaching a preset trigger threshold based on the relative approach velocity and the response delay of the pre-charge circuit; when the horizontal offset, vertical offset, pitch angle, or roll angle exceeds a preset alignment tolerance range, suppress the generation of the trigger command and generate an alignment failure indication signal, which is output to a visual indicator on the first connector.

[0018] In this embodiment, the controller calculates the six-degree-of-freedom relative pose, extending pose perception from a single distance dimension to six degrees of freedom in space, enabling the system to know not only "how far" but also "how much off-center." Before generating the trigger command, the state vector is recursively derived using the constant rate of change of acceleration constraint, and motion constraints are used to smooth observation noise. Even if the sensing signal experiences momentary fluctuations under strong interference, the iteratively estimated relative approach speed remains continuous and stable. Combined with the response delay of the pre-charging circuit, the trigger command is generated in advance, compensating for the time lag from detection to execution, ensuring that the voltage is balanced at the instant the power pin makes physical contact with the power socket. When the pose deviation exceeds the preset alignment tolerance range, the trigger command is suppressed, preventing charged contact in the off-center state at the decision level, thus completely eliminating the risk of arcing caused by misalignment.

[0019] Optionally, in some embodiments of the system, the controller is further operable to: perform synthetic aperture focusing processing on the sensing signal to reconstruct a three-dimensional topographic point cloud of the second insulating end face; pair each point in the three-dimensional topographic point cloud with the nearest corresponding point in a pre-stored reference model of the second insulating end face, and iteratively solve the rigid transformation matrix from the three-dimensional topographic point cloud to the reference model with the goal of minimizing the sum of squared Euclidean distances between paired point pairs, so as to solve the six-degree-of-freedom relative pose.

[0020] Here, the controller performs synthetic aperture focusing processing on the sensing signal, coherently superimposing the echoes from each element of the ultrasonic phased array to reconstruct the three-dimensional topographic point cloud of the second insulating end face, transforming the originally discrete echo signal into topographic information with spatial continuity. The iterative solution aims to minimize the sum of squared Euclidean distances between paired point pairs, continuously optimizing the rigid transformation matrix from the three-dimensional topographic point cloud to the reference model, gradually approximating the true six-degree-of-freedom pose. This pose calculation method based on global point cloud matching does not rely on the accuracy of a single ranging point, but utilizes the statistical information of the entire end face topography to significantly reduce the impact of local interference on pose estimation. Therefore, even under complex working conditions such as dust and oil contamination, it can still output highly reliable six-degree-of-freedom relative poses, providing a solid data foundation for the accurate generation of trigger commands.

[0021] Optionally, in some embodiments of the system, the controller is further operable to: extract echo amplitudes of multiple frequency components from the sensing signal, the multiple frequency components including a first frequency and a second frequency, the echo of the first frequency having a first penetration depth into the surface oxide layer of the second insulating end face, and the echo of the second frequency having a second penetration depth into the surface oxide layer; identify and suppress false echo points introduced by the surface oxide layer according to the echo amplitude ratio of the first frequency and the second frequency; assign a reduced confidence weight to points in the three-dimensional topography point cloud affected by the surface oxide layer; and perform matching with the confidence-weighted point cloud during the iterative solution process.

[0022] In this scheme, the first frequency and the second frequency generate differentiated echo amplitudes due to their different penetration depths when the surface oxide layer is present. The controller uses this amplitude ratio to identify false echo points, effectively distinguishing the reflection from the surface oxide layer from the true echo from the acoustic reflection array. Points affected by the oxide layer are assigned reduced confidence weights, ensuring that the optimization objective during the iterative solution is primarily driven by the true echo points, suppressing the contribution of outliers introduced by the oxide layer to the solution of the rigid transformation matrix. This multi-frequency joint denoising strategy utilizes the differences in the penetration characteristics of different frequency sound waves into the medium at the physical level, overcoming the pose calculation deviation caused by the change in oxide layer thickness at a single frequency. Therefore, the system can maintain high-precision alignment even after long-term service, delaying the degradation of anti-sparking performance caused by end-face oxidation.

[0023] Optionally, in some embodiments of the system, the controller is further operable to: perform a differential comparison between the three-dimensional topography point cloud reconstructed in the current insertion / removal operation and the three-dimensional topography point cloud of the previous insertion / removal operation stored in non-volatile memory; segment the regions in the differential point cloud that exceed a preset wear detection threshold into connected components, and extract the center coordinates, area, and depth of each connected component; generate an end-face abnormal wear indication signal when the depth of any of the connected components exceeds a preset micro-wear warning depth; map the center coordinates of each connected component to the projection position of the power pin on the second insulating end face; correlate the area and depth of each connected component with the resistance change rate of the corresponding power pin's shrinkage resistor to construct a joint degradation map between end-face topography degradation and electrical contact degradation, wherein the resistance change rate is recursively estimated; and identify end-face wear hotspot regions that cause electrical contact degradation based on the joint degradation map.

[0024] This embodiment extends the three-dimensional topographic point cloud from a real-time pose perception tool to a long-term degradation monitoring method. The differential comparison extracts the microscale topographic changes of the end face between two insertion / removal cycles, and the connected component segmentation aggregates discrete wear points into physically meaningful wear regions. By mapping the geometric features of the connected components to the projected positions of the corresponding power pins and correlating them with the resistance change rate of the shrinkage resistor, the joint degradation map establishes a quantitative mapping relationship between end face topographic degradation and electrical contact performance deterioration. This cross-physical domain correlation allows arcing hazards, which were previously only detectable retrospectively, to be identified in the early stages when abnormal wear occurs on the end face but before electrical faults occur. This elevates the anti-arcing strategy from passive response to proactive prediction, significantly extending the safe service life of the connector system.

[0025] Optionally, in some embodiments of the system, the controller is further operable to: dynamically adjust the observation noise covariance matrix according to the real-time signal-to-noise ratio of the sensing signal in the iterative update step; when the real-time signal-to-noise ratio is lower than a preset signal-to-noise ratio threshold, increase the diagonal elements of the observation noise covariance matrix to reduce the correction weight of the current observation on the state estimation, so that the estimated value of the relative proximity velocity depends more on the recursive result of the motion constraint during periods of strong interference.

[0026] Here, the controller monitors the real-time signal-to-noise ratio (SNR) of the sensing signal in the update step of the iterative recursion and dynamically adjusts the observation noise covariance matrix. When the real-time SNR decreases due to strong electromagnetic interference, the diagonal elements of the observation noise covariance matrix increase, which is equivalent to informing the recursive algorithm that the unreliability of the current observation has increased. The algorithm automatically reduces the correction weight of this observation for the state estimation. Under extreme interference, the estimation of the relative approach velocity is almost entirely driven by the recursive result of the constant rate of change of acceleration constraint, avoiding the direct injection of interference pulses into the state estimation channel. This adaptive confidence adjustment mechanism enables the system to maintain a smooth, non-jumping estimate of the relative approach velocity in the actual operating condition where the interference intensity fluctuates drastically, thereby ensuring that the generation time of the trigger command is stable and reliable, and will not be falsely triggered or missed due to instantaneous interference.

[0027] Optionally, in some embodiments of the system, the controller is further operable to: calculate the estimated contact time between the power pin and the power socket based on the solution results of the six-degree-of-freedom relative pose; monitor the rate of change of current in the equalization path within a preset verification time window after the estimated contact time; when the rate of change of current jumps from zero and exceeds a preset turn-on threshold, determine that the equalization path has established an electrical connection, confirm that the plugging and unplugging operation is successful, and switch the pre-charging circuit from the current-limiting mode to the full-power conduction state.

[0028] In this embodiment, the controller uses the calculation results of the six-degree-of-freedom relative pose to estimate the contact time and monitors the current change rate in the equalization path within the preset verification time window after that time. When the pre-contact electrode makes contact first, the pre-charging current begins to appear in the equalization path, and the current change rate jumps from zero; once the jump exceeds the preset connection threshold, it indicates that the electrical connection has been reliably established. At this time, the pre-charging circuit is switched from the current-limiting mode to the full-power conduction state, so that the main power circuit immediately starts working. This connection confirmation strategy based on the real-time current change rate avoids false connection judgments caused by mechanical rebound or poor contact, and ensures that the contact resistance has dropped to the safe level before full-power conduction, thereby eliminating the risk of local overheating or secondary arcing caused by a large current passing through the high-resistance contact point.

[0029] Optionally, in some embodiments of the system, the controller is operable to: calculate in real time the axial distance and pose deviation angle between the first connector and the second connector from the sensing signal; generate the trigger command when the axial distance enters a preset trigger range and the pose deviation angle is less than a preset alignment tolerance; suppress the trigger command and generate an alignment correction request signal when the axial distance enters the preset trigger range but the pose deviation angle exceeds the preset alignment tolerance; and prompt the operator with the alignment correction request signal through a visual indicator or audio indicator on the first connector to complete the pose calibration before physical pin contact.

[0030] Here, the controller uses the axial distance and the pose deviation angle as the joint decision conditions. The trigger command is generated only when both conditions simultaneously meet their respective thresholds; if the axial distance has entered the preset trigger range and the pose deviation angle exceeds the limit, triggering is actively suppressed, and instead, the alignment correction request signal is output. This dual-condition interlocking mechanism prevents false triggering when the distance condition is met but the pose condition is not, intercepting potential safety hazards before they reach the execution layer. The alignment correction request signal is fed back to the operator in real time via the visual indicator or the audio indicator, forming a closed-loop calibration process in human-machine collaboration. This allows even non-professional operators to complete high-precision blind insertion under guidance, thus significantly reducing the risk of misalignment and arcing accidents caused by arbitrary operation.

[0031] Optionally, in some embodiments of the system, the controller is operable to: perform time-frequency analysis on the sensing signal by multiplying the sensing signal segment by segment with a window function sliding along the time axis, and then solving for the complex amplitude distribution of each windowed signal segment at different frequency components to extract multipath reflection features in the gap between the first connector and the second connector; distinguish between the real echo of the acoustic reflection array and the parasitic reflection echo from the inner wall of the first connector housing based on the multipath reflection features; calculate the relative pose based only on the real echo and generate the trigger command; wherein, peak search is performed on the time spectrum obtained from the time-frequency analysis, and spectral peaks whose peak frequencies deviate from the center frequency of the ultrasonic phased array sensor's emitted pulse by more than a preset deviation range are identified as the parasitic reflection echoes.

[0032] This embodiment expands the sensing signal onto the time-frequency two-dimensional plane through time-frequency analysis, causing echoes from different propagation paths to occupy different energy distribution regions in the time spectrum. The true echo from the acoustic reflection array, due to reflection by the precisely designed corner reflectors, has a peak frequency highly consistent with the center frequency of the transmitted pulse; while the parasitic reflected echo, due to multiple scattering and absorption on the inner wall of the housing, has a shifted arrival time and frequency. The peak search automatically identifies and eliminates the parasitic reflected echo by detecting the deviation between the spectral peak frequency and the center frequency, ensuring that the subsequent relative pose calculation relies only on the uncontaminated true echo. This echo filtering method based on differences in physical propagation characteristics eliminates structural noise interference to the ranging from the signal source, thus enabling the system to maintain high-precision non-contact sensing even in the compact, complexly internally reflected connector housing.

[0033] Optionally, in some embodiments of the system, the controller is further operable to: continuously acquire multiple frames of the sensing signal at the pulse repetition frequency of the ultrasonic phased array sensor during the relative approach of the first connector and the second connector, and reconstruct the corresponding multiple frames of the three-dimensional topography point cloud; perform inter-frame rigid body motion compensation on the multiple frames of the three-dimensional topography point cloud to align the multiple frames of the three-dimensional topography point cloud to the same reference coordinate system; perform weighted average fusion of the aligned multiple frames of the three-dimensional topography point cloud within a voxel mesh to generate a super-resolution point cloud with a spatial resolution higher than that of a single frame of the three-dimensional topography point cloud; and replace the single frame of the three-dimensional topography point cloud with the super-resolution point cloud to perform the iterative solution.

[0034] Optionally, in some embodiments of the system, the controller is further operable to: calculate the relative deviation angle of the normal vector of each point in the neighborhood of the point for each point in the three-dimensional topographic point cloud, and encode the frequency distribution of the relative deviation angle of the normal vector in each angle interval as a multi-dimensional feature vector; calculate the corresponding multi-dimensional feature vector for each point in the pre-stored reference model of the second insulating end face; and in the iterative solution process, use the feature space distance between the multi-dimensional feature vector of each point in the three-dimensional topographic point cloud and the multi-dimensional feature vector of the corresponding nearest point in the reference model as the distance metric between the paired point pairs, replacing the squared Euclidean distance. Attached Figure Description

[0035] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1A schematic diagram of the overall architecture of a zero-differential-pressure blind mating connector system according to an embodiment of the present invention is shown; Figure 2 A schematic diagram of a multi-cell parallel capacitive coupling detection circuit structure for a zero-difference blind mating connector system according to an embodiment of the present invention is shown.

[0036] In the description of the accompanying drawings, the same, similar or corresponding reference numerals represent the same, similar or corresponding units, elements or functions. Detailed Implementation

[0037]

Example 1

[0038] The system includes a first connector 20 and a second connector 40. The first connector 20 is the end held by the operator, and the second connector 40 is typically fixed to the battery module or the device panel. The first connector 20 has a first insulating end face 21, which is made of zirconium oxide ceramic and is approximately 3 mm thick, and is embedded in the front opening of the housing. At least one power pin 22, made of copper alloy and silver-plated, is exposed from the first insulating end face 21 to carry the large current of the main circuit. The ultrasonic phased array sensor 30 is embedded in the peripheral area of ​​the first insulating end face 21.

[0039] The core of the ultrasonic phased array sensor 30 is a ring-shaped piezoelectric transducer array, which consists of 64 independent elements evenly distributed along the circumference. Each independent element is a rectangular piezoelectric ceramic sheet with dimensions of 0.8 mm × 0.6 mm, and the resonant frequency is set at 200 kHz. Sound-absorbing material is filled between the independent elements to suppress acoustic crosstalk between them. The front end face of the ring-shaped piezoelectric transducer array is covered with an acoustic matching layer, which is a composite material of epoxy resin and alumina powder. The acoustic impedance is approximately 6 MRayl, between that of the piezoelectric ceramic and air, to improve the transmission efficiency of sound waves. A silicone rubber sealing ring forms a seal between the acoustic matching layer and the first insulating end face 21. The back side of the ring piezoelectric transceiver array is connected to the multi-channel transceiver control chip via the flexible circuit board. The chip is integrated in the internal cavity of the first connector 20 and includes 64 independent transmit waveform generators and receive amplifiers. It can apply the nanosecond-level phase-delayed pulse excitation to each of the independent array elements to synthesize the transmit beam that can be programmably scanned in the space.

[0040] In one specific implementation of the ultrasonic phased array sensor 30, the annular piezoelectric transducer array is embedded in the periphery of the first insulating end face 21. An annular groove is machined on the periphery of the first insulating end face 21. The depth of the annular groove is 2 mm, and its width is 8 mm. Both the inner and outer walls of the annular groove are coated with a metal shielding layer. The metal shielding layer is electrically connected to the grounding terminal of the housing via conductive adhesive, forming the electromagnetic shielding relationship. The annular piezoelectric transducer array has 64 independent array elements. Each independent array element is a rectangular lead zirconate titanate piezoelectric ceramic sheet with a radial length of 8 mm, a width of 1.0 mm, and a thickness of 0.5 mm. The circumferential spacing between adjacent independent array elements is 1.7 mm, which is equal to half the wavelength corresponding to the ultrasonic phased array sensor 30's operating frequency of 200 kHz in air. The polarization direction of each independent array element is along the thickness direction. The positive electrode is disposed on the front end face of the independent array element, and the negative electrode is disposed on the back side of the independent array element. Both the positive and negative electrodes are sintered silver electrodes with a thickness of 5 μm. The back side of each independent array element is bonded to the bottom surface of the annular groove with epoxy resin to form a support relationship. The gap between adjacent independent array elements is filled with sound-absorbing material, which is tungsten-doped epoxy resin with a tungsten powder mass fraction of 85%. The acoustic impedance after curing is 15 MRayl, close to the piezoelectric ceramic acoustic impedance of 20 MRayl of the independent array element, to suppress acoustic crosstalk propagating between the independent array elements through the first insulating end face 21. The multi-channel transceiver control chip is soldered onto the flexible printed circuit board (PCB). The PCB, after being bent, is positioned between the back side of the annular piezoelectric transceiver array and the rear wall of the housing. The 64 transmit pins and 64 receive pins of the multi-channel transceiver control chip are respectively connected to the positive and negative electrodes of each independent array element via 64 pairs of differential traces on the PCB, forming the electrical connection. The length difference between each pair of differential traces is controlled within 0.5 mm to ensure consistent signal transmission delay across all channels.The multi-channel transceiver control chip integrates 64 transmit waveform generators. Each transmit waveform generator includes a 10-bit phase accumulator and a 10-bit digital-to-analog converter. The clock frequency of the phase accumulator is 40 MHz, and the phase accumulation step size is determined by the phase control word written by the controller 60 through the serial peripheral interface. The phase control word is a 16-bit integer, and the phase resolution is 360 degrees divided by 2 to the power of 16 equals 0.0055 degrees. The digital-to-analog converter converts the output of the phase accumulator into an analog sine wave, which is amplified by the power amplifier and outputs a pulse with a positive peak value and a negative peak value difference of 100 V to excite the corresponding independent array element. The phase control word of each transmit waveform generator is calculated according to the delay rule that makes the synthesized beam point to the preset scanning direction in space, and the phase delay between adjacent independent array elements is [value missing]. Multiply by the circumferential spacing of the independent array elements, divide by the operating wavelength, and then multiply by the sine of the scanning angle.

[0041] The second connector 40 has a second insulating end face 41, made of the same material as the first insulating end face 21. A power socket 42 corresponding to the power pin 22 is exposed on the second insulating end face 41. The power socket 42 contains the elastic slotted contact made of beryllium copper alloy. An acoustic reflection array 50, opposite to the ultrasonic phased array sensor 30, is disposed on the second insulating end face 41. The acoustic reflection array 50 includes multiple micro corner reflectors, each consisting of three mutually perpendicular reflective surfaces. The reflective surfaces are micro-pyramidal structures coated with a metal reflective layer, formed on the exposed surface of an annular metal target disk using a micro-nano imprinting process. The annular metal target disk is embedded in the second insulating end face 41, corresponding to the annular piezoelectric transducer array in the axial direction. The micro corner reflectors are approximately 0.3 mm in size, and the array spacing is 0.5 mm, enabling efficient reflection of incident ultrasonic waves back to the ultrasonic phased array sensor 30 along the original path.

[0042] Regarding the further construction of the acoustic reflection array 50, the number of its micro corner reflectors is 256. Each micro corner reflector is a square micro-pyramid with a base side length of 0.5 mm and a height of 0.25 mm. It consists of three mutually perpendicular reflective surfaces, each with an angle of 45 degrees to the base. Each reflective surface is coated with an aluminum reflective layer with a thickness of 200 nm and a reflectivity of greater than 95% for 200 kHz ultrasound. The 256 micro corner reflectors are evenly arranged in four concentric rings along the circumference of the annular metal target disk, with 64 reflectors in each ring. The radial distance between two adjacent rings is 1.2 mm, and the central angle between two adjacent micro corner reflectors within the same ring is 5.625 degrees. The annular metal target disk is made of stainless steel, with an outer diameter of 50 mm, an inner diameter of 30 mm, and a thickness of 1.5 mm. It is embedded in the corresponding annular groove of the second insulating end face 41, flush with the second insulating end face 41. The exposed surface of the annular metal target disk is sandblasted to form a diffuse scattering background surface with a roughness of 2 μm. The miniature corner reflector protrudes above the diffuse scattering background surface, such that the echo intensity of the miniature corner reflector is more than 18 dB higher than the echo intensity of the diffuse scattering background surface.

[0043] The controller 60 employs a heterogeneous computing architecture, comprising a field-programmable gate array (FPGA) and a microcontroller unit, which communicate via a serial peripheral interface. The FPGA is responsible for performing parallel beamforming and synthetic aperture focusing on the multi-channel echo data from the ultrasonic phased array sensor 30. The microcontroller unit is responsible for pose calculation, the generation logic of trigger commands, and the timing control of the pre-charge circuit 70. The controller 60 is connected to the multi-channel transceiver control chip of the ultrasonic phased array sensor 30 via a low-voltage differential signal interface to receive the sensing signals.

[0044] The core of the pre-charge circuit 70 is a programmable slope constant current source, whose input terminal is connected to the DC bus, and whose output terminal is connected to the power pin 22 through the equalization path 71. The pre-charge circuit 70 also includes a pre-contact electrode, which is disposed on the first connector 20. Its front end face protrudes approximately 2 mm axially from the front end face of the power pin 22, and makes contact with the corresponding electrode of the second connector 40 before the power pin 22 makes physical contact, thus providing the pilot path for the pre-charge current. After receiving the sensing signal from the ultrasonic phased array sensor 30, the controller 60 calculates the relative pose between the first connector 20 and the second connector 40. When the axial distance enters the preset trigger range and the pose deviation angle is within the tolerance, the controller generates the trigger command and sends it to the pre-charging circuit 70. The pre-charging circuit 70 then adjusts the voltage difference between the two ends to below the preset safety threshold through the pre-contact electrode and the equalization path 71. The entire process is completed within a few millimeters of the gap before the power pin 22 and the power socket 42 make physical contact.

[0045] Another implementation method can be used in the process of the controller 60 calculating the pose and generating trigger commands. The sensing signal output from the multi-channel transceiver control chip of the ultrasonic phased array sensor 30 is acquired. The sensing signal is 64-channel parallel digital echo data, with each channel containing 1024 sampling points at a sampling rate of 2 MS / s. Each sampling point represents the echo amplitude as a 16-bit integer. Inter-channel amplitude calibration is performed on the sensing signal by dividing the echo amplitude of each channel by the sensitivity coefficient recorded during factory calibration to obtain the normalized echo amplitude. Bandpass filtering is applied to the normalized echo amplitude, with a filter center frequency of 200 kHz and a bandwidth of 40 kHz to filter out out-of-band noise. For the filtered channel data, a segment with a flight time between 0.2 ms and 2.0 ms is extracted, corresponding to the working distance range of 5 mm to 500 mm between the first connector 20 and the second connector 40. The extracted channel data are beamformed and accumulated according to the spatial coordinates of the independent array elements in the annular piezoelectric transducer array to synthesize 128 directional echo signals in the scanning directions. Envelope detection is performed on the directional echo signal in each scanning direction, and the flight time corresponding to the envelope peak value is extracted. The flight time is multiplied by the speed of sound in air, 340 m / s, to obtain the slant range measurement value for that scanning direction. The 128 slant range measurements are projected onto the Cartesian coordinate system according to the spatial angle of their respective scanning directions. Triangulation and interpolation are performed on the projection points to generate the three-dimensional topographic point cloud of the second insulating end face 41. Each point in the three-dimensional topographic point cloud contains four attributes: x-coordinate, y-coordinate, z-coordinate, and echo intensity. The x-coordinate, y-coordinate, and z-coordinate of each point in the 3D topographic point cloud are paired with the coordinates of the nearest corresponding point in the pre-stored reference model. The squared Euclidean distance of each pair of paired points is calculated, and the squared Euclidean distances of all paired points are summed to obtain the total registration error. With the goal of minimizing the total registration error, the gradient descent method is used to iteratively update the six parameters of the rigid transformation matrix. These six parameters are the translation along the x-axis, the translation along the y-axis, the translation along the z-axis, the rotation angle around the x-axis, the rotation angle around the y-axis, and the rotation angle around the z-axis. The step size for each iteration is 0.01 mm or 0.01 degrees. The iteration terminates when the relative change in the total registration error is less than a preset convergence threshold. The translation amount is extracted from the converged rigid transformation matrix as the horizontal offset and the vertical offset, the rotation angle is extracted as the pitch angle and the roll angle, and the translation amount along the z-axis is extracted as the axial distance to form the six-degree-of-freedom relative pose.From the time series of the axial distance, the values ​​of the axial distance at the five most recent sampling moments are read, with an interval of 1 ms between adjacent sampling moments, forming an observation sequence of length 5. Using the observation sequence as input and the constant rate of change of acceleration as the motion constraint, the predicted value of the state vector is recursively calculated, and the posterior estimate of the state vector is obtained to obtain the estimated value of the relative approach velocity at the current moment. The hardware link delay is read from the circuit parameters of the pre-charge circuit 70, where the hardware link delay is the time interval from the generation of the trigger command to the actual start of current rise in the equalization path 71. The estimated value of the relative approach velocity is multiplied by the hardware link delay to obtain the distance compensation amount. The distance compensation amount is subtracted from the preset trigger threshold to obtain the compensated trigger distance threshold. When the axial distance decreases to the compensated trigger distance threshold, the trigger command is generated and written into the trigger register of the pre-charge circuit 70. Check whether the horizontal offset, vertical offset, pitch angle and roll angle are all within the preset alignment tolerance range. If any one of them exceeds the tolerance, clear the trigger instruction in the trigger register, generate the alignment failure indication signal, and write the alignment failure indication signal into the display buffer of the visual indicator.

[0046] Assuming that during a certain insertion / removal operation, the echo flight time measured in the scanning direction directly in front of the ultrasonic phased array sensor 30 is 0.0588 ms, then the current measured value of the axial distance is... mm. The axial distances at the previous four sampling times were 10.5 mm, 10.3 mm, 10.2 mm, and 10.1 mm, respectively, forming the observation sequence [10.5, 10.3, 10.2, 10.1, 10.0], in mm. The rate of change of acceleration was set to be constant at 0, i.e., the relative approach velocity was uniform. The difference in the axial distance between adjacent times was divided by the sampling interval of 1 ms to obtain the observed values ​​of the relative approach velocity at each time, which were (10.3-10.5) / 0.001=-200 mm / s, (10.2-10.3) / 0.001=-100 mm / s, (10.1-10.2) / 0.001=-100 mm / s, and (10.0-10.1) / 0.001=-100 mm / s. The average estimated relative approach speed is -125 mm / s, with the negative sign indicating proximity. The measured hardware link delay is 0.5 ms, therefore the distance compensation is... The preset trigger threshold is set to 5.0 mm, and the compensated trigger distance threshold is 5.0625 mm. When the axial distance decreases from 10.0 mm to approximately 5.06 mm, the controller 60 generates the trigger command. If the horizontal offset is 0.3 mm, the vertical offset is 0.2 mm, the pitch angle is 1.5 degrees, and the roll angle is 1.0 degree, all within the preset alignment tolerance range (horizontal offset threshold 0.5 mm, vertical offset threshold 0.5 mm, pitch threshold 3 degrees, roll threshold 3 degrees), then the trigger command takes effect normally.

[0047]

Example 2

[0048] The multi-channel transceiver control chip of the ultrasonic phased array sensor 30 controls the annular piezoelectric transducer array to perform a spiral conical scan in space at a pulse repetition frequency of 1 kHz, transmitting 64 pulses per scan to cover the entire area of ​​the second insulating end face 41. During the scan, each independent array element receives the echo from the acoustic reflection array 50. The echo signal is amplified by the low-noise amplifier inside the multi-channel transceiver control chip and then transmitted to the controller 60 through the low-voltage differential signal interface.

[0049] The field-programmable gate array (FPGA) unit of the controller 60 first performs synthetic aperture focusing (SAR) processing on the sensing signal to reconstruct the three-dimensional topographic point cloud of the second insulating end face 41. The specific process of SAR processing is as follows: the echo data in each scanning direction is matched and filtered with a copy of the transmitted pulse to extract the time of flight; then, based on the spatial geometric relationship of each independent array element, the echo amplitude is back-projected onto the spatial grid; and the complex amplitudes of the overlapping regions are coherently superimposed to form the three-dimensional topographic point cloud. The spatial resolution of each point in the point cloud is approximately 0.5 mm, containing approximately 200,000 valid points.

[0050] Synthetic aperture focusing can also be implemented in more detail. The transmit-receive echo data of all 64 independent elements of the annular piezoelectric transducer array are extracted from the sensing signal. The echo data is stored in a three-dimensional array, where the first dimension is the transmit element number, the second dimension is the receive element number, and the third dimension is the time sampling point index. For each transmit-receive channel pair, the acoustic wave propagation distance corresponding to the time sampling point index is calculated. The propagation distance is equal to the time sampling point index multiplied by the sampling interval, multiplied by the speed of sound, and divided by 2, to obtain the spatial detection depth corresponding to that channel at that sampling moment. The amplitude of the echo data is multiplied by a gain coefficient that increases with the propagation distance. The gain coefficient is equal to 1 plus the square of the propagation distance divided by the square of a preset reference distance to compensate for the attenuation of the long-distance echo. For each transmitting element, all 64 corresponding receiving channels are traversed. The gain-compensated echo amplitude is allocated to the spatial grid of the plane containing the second insulating end face 41 according to the geometric midpoint coordinates between the transmitting and receiving elements. The grid spacing of the spatial grid is 0.5 mm. When the echo amplitudes of multiple channels are allocated to the same grid point, the echo amplitudes are coherently superimposed in complex form. The phase of the coherent superposition is determined by dividing the total length of the transmitting and receiving paths of the channel by the wavelength and taking the remainder. After completing the grid allocation and coherent superposition of all 4096 transmitting-receiving channel pairs, a two-dimensional complex amplitude distribution map of the second insulating end face 41 is obtained. The modulus of the complex value of each grid point in the two-dimensional complex amplitude distribution map represents the scattering intensity at that grid point. The scattering intensity value is obtained by taking the modulus of the complex value of each grid point in the two-dimensional complex amplitude distribution map. Grid points with scattering intensity values ​​lower than the preset noise threshold are filtered out, and the remaining grid points constitute the three-dimensional topography point cloud. The x-coordinate and y-coordinate of each point are the grid point index multiplied by the grid point spacing, and the z-coordinate is the height value obtained after the scattering intensity value is mapped by the preset linear mapping. The reference model is obtained as follows: during system factory calibration, the first connector 20 and the second connector 40 are subjected to a synthetic aperture focusing process in a clean and dry standard docking state. The generated three-dimensional topography point cloud is stored as the reference model in the non-volatile memory of the controller 60. In each insertion / removal operation, the real-time reconstructed three-dimensional topography point cloud is compared with the reference model, and the total registration error of the iterative nearest point registration between the two sets of point clouds is calculated. When the total registration error exceeds the preset registration quality threshold, the point cloud reconstruction quality is deemed unqualified, and the synthetic aperture focusing process is re-executed.

[0051] Assuming the preset reference distance is 10 mm and the propagation distance corresponding to a certain grid point is 20 mm, then the gain coefficient is... The original echo amplitude of this channel at this grid point is 120, and after gain compensation, it is 600. The original echo amplitude of the other channel at the same grid point is 100, and after gain compensation, it is 500. The total lengths of the transmission path and the receiving path for both are 40.2 mm and 39.8 mm, respectively. The operating wavelength of the ultrasonic phased array sensor 30 is... m The total length divided by 1.7 mm and taking the remainder yields 0.65 mm and 0.40 mm respectively, which are converted into the radian phase difference. rad. The complex number of the first channel is expressed as The complex number of the second channel is represented as The sum of complex numbers is The modulus of the complex number is The scattering intensity value of this grid point is 986. If the preset noise threshold is 200, then this grid point is retained. Dividing the scattering intensity value of 986 of this grid point by the preset linear mapping coefficient 50 yields the z-coordinate of 19.72 mm.

[0052] To improve the pose calculation accuracy, the controller 60 extracts the echo amplitudes of two frequency components from the sensing signal. The first frequency is 200 kHz, corresponding to the center transmission frequency of the ultrasonic phased array sensor 30; the second frequency is 380 kHz, which is the harmonic frequency. The echo of the first frequency has a relatively deep first penetration depth through the surface oxide layer of the second insulating end face 41, approximately penetrating the 0.1 mm thick oxide layer; the echo of the second frequency has a shallower penetration depth, mainly reflected from the surface of the oxide layer. When the surface oxide layer is present, the ratio of the echo amplitude of the first frequency to the echo amplitude of the second frequency will exhibit a characteristic shift. The controller 60 identifies false echo points introduced by the surface oxide layer based on this amplitude ratio, and assigns a reduced confidence weight to points in the three-dimensional topography point cloud affected by the surface oxide layer. The weight value decreases linearly according to the degree to which the amplitude ratio deviates from the normal range.

[0053] Next, the controller 60 performs iterative nearest-point registration between each point in the weighted 3D topographic point cloud and the pre-stored reference model of the second insulating end face 41. The reference model is the point cloud data stored after a high-precision scan of the second insulating end face 41 in a clean, oxidation-free state during the system's factory calibration. The registration process aims to minimize the sum of squared Euclidean distances between the paired point pairs. Each iteration calculates a rigid transformation matrix, which includes three translational components and three rotational components. After the iteration converges, the rigid transformation matrix represents the six-degree-of-freedom relative pose of the first connector 20 relative to the second connector 40.

[0054] When the controller 60 detects that the axial distance has entered the preset trigger threshold range, it makes a comprehensive judgment based on the calculated horizontal offset, vertical offset, pitch angle, and roll angle. If all the deviations are within the preset alignment tolerance range, a trigger command is generated and sent to the pre-charge circuit 70. If any deviation exceeds the tolerance, the generation of the trigger command is suppressed, and an alignment failure indication signal is output through the visual indicator on the first connector 20 to prompt the operator to correct the insertion posture.

[0055] Regarding parasitic reflected echoes that may be mixed in the sensing signal, the following identification and removal measures can be taken. All echo sampling data from a single scan is read from the multi-channel transceiver control chip. The echo sampling data is a one-dimensional array of length 2048, where each element is a 16-bit integer, and the sampling rate is 2 MS / s. A Hanning window with a width of 256 sampling points is defined. The starting position of the Hanning window is from the first sampling point of the echo sampling data, sliding to the right in steps of 64 sampling points. At each stop position of the Hanning window, the 256 sampling points within the window are multiplied point by point by the window function to obtain the windowed signal segment. The windowed signal segment is subjected to a complex form of the Discrete Fourier Transform to convert the time-domain windowed signal segment into a complex amplitude distribution in the frequency domain. The complex amplitude distribution is a complex array of 128 frequency components, where each complex number represents the amplitude and phase of the corresponding frequency component. Arrange the complex amplitude distributions at all dwell positions in chronological order to obtain a two-dimensional time-frequency matrix. The rows of the time-frequency matrix correspond to the frequency components, the columns correspond to the time windows, and the matrix elements are the magnitudes of the complex amplitudes. In the time-frequency matrix, find the peak frequency along the frequency axis. The peak frequency is the frequency value corresponding to the frequency component with the largest complex amplitude magnitude in that time window. Compare the deviation between the peak frequency and the center frequency (200 kHz) of the emitted pulse from the ultrasonic phased array sensor 30, and calculate the absolute value of the deviation in kHz. If the absolute value of the deviation exceeds the preset deviation range of 20 kHz, then the complex amplitude magnitudes of all frequency components in that time window are marked as candidate components of the parasitic reflected echo. For frequency-time regions where multiple consecutive time windows are marked as candidate components, merge the connected components to form a complete time-frequency region of the parasitic reflected echo. The complete time-frequency region of the parasitic reflected echo is removed from the time-frequency spectrum matrix, and the complex amplitude modulus values ​​of the remaining region in the time-frequency spectrum matrix are used as the time-frequency distribution of the true echo. Based on the time-frequency distribution of the true echo, the flight time corresponding to the frequency component with the largest complex amplitude modulus value in each time window is found, and the flight time is converted into the distance value to obtain the measured value of the axial distance corresponding to the true echo.

[0056] Assuming the Hanning window is at its 10th dwell position, after the discrete Fourier transform of the 256 sampling points within the window, the maximum complex amplitude modulus of the 128 frequency components appears at the 32nd frequency component, corresponding to the frequency of... The center frequency of the transmitted pulse is 200 kHz, with a deviation of 300 kHz, exceeding the preset deviation range by 20 kHz. This time window is marked as a candidate. The peak frequencies at the 11th and 12th stop positions are 490 kHz and 480 kHz, respectively, both exceeding the preset deviation range. These three consecutive time windows are marked and merged into a complete time-frequency region of the parasitic reflected echo, corresponding to a frequency range of 480 kHz to 500 kHz and a time range of the 10th to 12th time windows. The corresponding flight time is from the center time of the 10th window (0.640 ms) to the center time of the 12th window (0.768 ms). This region is removed from the time-spectrum matrix. At the 15th stop position, the peak frequency is 205 kHz, with a deviation of 5 kHz, within the preset deviation range. The echo in this time window is retained as the true echo, and the flight time corresponding to its peak frequency is 0.960 ms, which is converted to the distance value. mm.

[0057]

Example 3

[0058] The programmable slope constant current source employs the closed-loop constant current topology. The digital-to-analog converter receives the digital command representing the desired current slope from the controller 60 and converts it into an analog reference voltage of 0-3.3 V. The non-inverting input of the operational amplifier receives the reference voltage, and its output drives the gate of the power MOSFET array. The power MOSFET array consists of four N-channel MOSFETs connected in parallel, with their drains connected to the equalization path 71 and their sources grounded through the sampling resistor. The sampling resistor is a 0.1 Ω precision power resistor, and the voltage across it is fed back to the inverting input of the operational amplifier via a differential line, forming the closed-loop constant current control. When the reference voltage is 1 V, the current in the equalization path after the closed loop stabilizes is 10 A.

[0059] The second type of compensation network is connected between the output terminal and the inverting input terminal of the operational amplifier, and consists of a feedback capacitor and a feedback resistor connected in series. The feedback capacitor has a capacitance of 47 nF, and the feedback resistor has a resistance of 1 kΩ. The values ​​of both are determined based on the total gate charge characteristics of the power metal-oxide-semiconductor field-effect transistor array, so that the phase margin of the closed-loop constant current control is maintained at approximately 55 degrees, ensuring that no oscillation occurs under different load conditions.

[0060] The programmable slope constant current source also integrates the ramp generator. The ramp generator includes a constant current source, an integrating capacitor, an analog switch, and a ramp reset switch. The constant current source outputs a constant current of 100 μA, and the integrating capacitor has a capacitance of 10 nF. The analog switch is a single-pole single-throw analog switch, and its control terminal is connected to the controller 60. The ramp reset switch is connected in parallel with the integrating capacitor and is an N-channel MOSFET; its control terminal is also connected to the controller 60.

[0061] When the controller 60 generates the trigger command, it simultaneously closes the analog switch and opens the ramp reset switch, and the constant current source charges the integrating capacitor with a constant current of 100 μA. Because... The voltage across the integrating capacitor rises linearly at a rate of 10 V / ms, generating a ramp voltage with a constant slope. This ramp voltage, buffered by the operational amplifier, drives the power MOSFET array, causing the pre-charge current in the equalization path 71 to increase steadily at a slope of 1 A / ms. When the trigger command is released, the controller 60 disconnects the analog switch and closes the ramp reset switch. The integrating capacitor discharges to zero within approximately 2 μs, preparing for the next trigger.

[0062] Before generating the trigger command, the controller 60 needs to accurately estimate the relative approach speed between the first connector 20 and the second connector 40. The controller 60 constructs the state vector using the axial distance, the relative approach speed, and the relative approach acceleration, with the recursive process using a constant rate of change of acceleration as the motion constraint. Specifically, within each recursive cycle (approximately 1 ms), the controller 60 recursively calculates the prior state prediction for the current cycle based on the posterior state estimate from the previous cycle and the assumption of a constant rate of change of acceleration. Then, using the time series of the axial distance as the observation, it calculates the posterior state estimate for the current cycle. During the recursive process, when the real-time signal-to-noise ratio of the sensing signal is lower than the preset signal-to-noise ratio threshold, the controller 60 increases the diagonal elements of the observation noise covariance matrix and decreases the correction weight of the current observation on the state estimate, making the estimated relative approach speed more dependent on the recursive result of the motion constraint, thereby maintaining the continuous stability of the speed estimate during periods of strong electromagnetic interference.

[0063] Based on the estimated relative approach speed and the hardware response delay of the pre-charging circuit 70, the controller 60 generates the trigger command at a certain moment before the axial distance reaches the preset trigger threshold. The compensation amount is the product of the relative approach speed and the response delay, so that when the pre-charging circuit 70 actually completes the voltage equalization, the power pin 22 is in physical contact with the power socket 42.

[0064] Furthermore, the assessment of end-face wear during insertion / removal can be tracked using differential point clouds and joint degradation maps. The three-dimensional topography point cloud stored in the previous insertion / removal operation is read from the non-volatile memory as the historical point cloud, and the three-dimensional topography point cloud reconstructed in the current insertion / removal operation is used as the current point cloud. The current point cloud and the historical point cloud are precisely registered. Using the coordinate system of the historical point cloud as a reference, the current point cloud is rigidly transformed to be completely aligned with the historical point cloud. Point-by-point differencing is performed on the aligned current point cloud and historical point cloud. The closest corresponding point in the historical point cloud is found for each point in the current point cloud, and the difference in the z-coordinates of the two corresponding points is calculated. This difference is used as the difference depth value for that point. Points whose difference depth value exceeds a preset wear detection threshold (0.01 mm) are marked as wear candidate points. The wear candidate points are segmented into connected components using the eight-neighbor connectivity criterion, grouping spatially adjacent wear candidate points into the same connected component. For each connected component, the average of the x-coordinate and the average of the y-coordinate of all points within that component are calculated as the center coordinate of that component, in mm. The average of the differential depth values ​​of all points within that component is calculated as the depth of that component, in mm. The number of points contained in each connected component is counted, and this number is multiplied by the projected area of ​​each point on the second insulating end face 41, 0.0025 mm², to obtain the area of ​​the connected component, in mm². The depth of each connected component is compared with a preset micro-wear warning depth of 0.05 mm. When the depth of any connected component exceeds the preset micro-wear warning depth, an abnormal wear indication signal for the end face is generated and written into the status register of the vision indicator. For each power pin 22, the projection area of ​​the power pin 22 on the second insulating end face 41 is determined by querying the mechanical design parameters of the first connector 20. The center coordinates of each connected region are compared with the projection area of ​​each power pin 22. If the center coordinates are located within a certain projection area, the connected region is mapped to the power pin 22. The resistance change rate of the shrinkage resistor of the power pin 22 is synchronized from the cloud database. The resistance change rate is a scalar value obtained by recursion estimation, with units of mΩ / thousandth insertions and removals. The sum of the areas of all connected regions mapped by each power pin 22 is taken as the cumulative wear area of ​​the power pin 22. The cumulative wear area and the resistance change rate of the power pin 22 are filled into the corresponding entries of the joint degradation map. The joint degradation map is a two-dimensional table, with the row index being the power pin number and the columns including the cumulative wear area and the resistance change rate.Traverse all rows of the joint degradation map, find the row where the cumulative wear area exceeds the preset wear area threshold, mark the power pin number corresponding to the row as the end face wear hot spot area, and write the list of end face wear hot spot area numbers into the alarm information buffer of the visual indicator.

[0065] Assuming that in a certain plug-in / plug-out operation, the connected region contains 120 points, and multiplying the number of points by the projected area of ​​0.0025 mm² yields an area of ​​0.3 mm² for the connected region. The average differential depth of all points within this connected region is 0.08 mm, exceeding the preset micro-wear warning depth of 0.05 mm, triggering the abnormal end-face wear indication. The center coordinates of this connected region are (3.2 mm, 1.5 mm), located within the projected area of ​​the power pin 22. The resistance change rate of the power pin 22 is synchronized from the cloud as 0.15 mΩ / thousandth plug-in / plug-out cycles. The cumulative wear area of ​​the power pin 22 in the joint degradation map is 0.3 mm², and the resistance change rate is 0.15. If the preset wear area threshold is 0.5 mm², the power pin 22 has not yet exceeded the limit and is not marked as a hotspot area. If the cumulative wear area of ​​the power pin 22 is 0.8 mm², exceeding the preset wear area threshold, then the power pin 22 is marked as the end face wear hotspot area and written into the alarm information buffer.

[0066]

Example 4

[0067] The pre-contact electrode in the pre-charging circuit 70 is located at the front end of the first connector 20 and is a circular structure made of brass. Its inner diameter matches the outer diameter of the power pin 22, and it is fitted onto the outer periphery of the power pin 22. The front end face of the pre-contact electrode protrudes approximately 2 mm axially from the front end face of the power pin 22. A plurality of micro-protrusion structures are provided on the front end face, evenly distributed circumferentially along the front end face, totaling eight structures. Each micro-protrusion structure has a conical tip with a radius of curvature of less than 0.05 mm. The conical tip is formed by electrical discharge machining, and its surface hardness reaches HRC55 or higher after quenching treatment.

[0068] When the first connector 20 approaches the second connector 40, the front end face of the pre-contact electrode first makes physical contact with the corresponding electrode of the second connector 40. The conical tip of the micro-protrusion structure acts with concentrated pressure on the surface of the corresponding electrode, and the local pressure can reach hundreds of MPa, which is sufficient to mechanically break the oxide film layer on the surface of the corresponding electrode, directly exposing the metal substrate, thereby reducing the initial contact resistance from tens of mΩ to several mΩ, providing the low-impedance path for the subsequent pre-charging current.

[0069] The programmable slope constant current source then responds to the trigger command of the controller 60, injecting current into the equalization path 71 at a controlled current slope. The pre-charge current passes through the pre-contact electrode, the corresponding electrode, the internal lead of the second connector 40, and the power socket 42, ultimately adjusting the voltage difference between the two ends to below the preset safety threshold before the power pin 22 and the power socket 42 make physical contact.

[0070] The pre-charge circuit 70 also includes a fast discharge branch connected in parallel between the power pin 22 and the reference ground. This branch includes a discharge resistor and a discharge switch connected in series. The discharge resistor has a resistance of 1 kΩ, and the discharge switch is an N-channel MOSFET, whose control terminal is connected to the controller 60. A constant current discharge circuit, consisting of a depletion-type MOSFET and a gate bias resistor, is connected in series between the discharge switch and the discharge resistor. The drain of the depletion-type MOSFET is connected to the discharge resistor, the source is connected to the drain of the discharge switch, and the gate is grounded through the gate bias resistor. This topology is in the on-state without an external bias voltage, and the on-current is set to approximately 5 mA by the gate bias resistor.

[0071] Before each trigger command is generated, the controller 60 briefly closes the discharge switch for approximately 2 ms. Any residual electrostatic charge on the power pin 22 is discharged to the reference ground via the constant current discharge circuit with a constant current of 5 mA. The purpose of using constant current discharge instead of direct resistance discharge is to suppress current spikes during the discharge process and avoid electromagnetic radiation interference with the normal operation of the ultrasonic phased array sensor 30 at the moment the discharge switch closes. After the discharge is completed, the discharge switch opens, and the pre-charge process starts normally.

[0072] As an optional method for connection confirmation, the estimated contact time can also be calculated and verified according to the following process: From the solution results of the six-degree-of-freedom relative pose, read the estimated values ​​of the axial distance and the relative approach velocity at the current moment. Divide the axial distance by the estimated value of the relative approach velocity to obtain the time offset of the estimated contact time relative to the current moment, in seconds. Add the time offset to the current system clock reading to obtain the absolute time value of the estimated contact time, and store the absolute time value in the timing comparison register of the controller 60. Within the preset verification time window after the estimated contact time, continuously collect the current value in the equalization path 71 at a sampling rate of 10 kHz, with a preset verification time window width of 5 ms. Perform differential processing on the collected current value sequence, calculate the difference between the current values ​​of two adjacent sampling points, and obtain the current change rate sequence. In the current change rate sequence, search whether the current change rate of three consecutive sampling points exceeds the preset connection threshold, which is 0.5 A / ms. If three consecutive sampling points satisfying the conditions are found, the position where the current change rate jumps from zero is determined as the electrical connection moment, a connection confirmation flag is generated, and the connection confirmation flag is set to 1. If no three consecutive sampling points satisfying the conditions are found within the preset verification time window, the connection confirmation flag is kept at 0, a insertion / removal failure indication signal is generated, and the insertion / removal failure indication signal is output to the visual indicator. When the connection confirmation flag is set to 1, a mode switching command is sent to the pre-charging circuit 70 to switch the pre-charging circuit 70 from the current limiting mode to the full-power conduction state. The current limiting mode corresponds to the current upper limit of the programmable slope constant current source being 10% of the rated current, and the full-power conduction state corresponds to the removal of the current upper limit restriction, allowing the current in the equalization path 71 to reach the rated value.

[0073] Assuming the current axial distance is 2.0 mm and the estimated relative approach velocity is 100 mm / s, then the time offset is... s The current system clock is 1000 ms, and the absolute time value of the estimated contact time is 1020 ms. The preset verification window is 5 ms, that is, the current of the equalization path 71 is monitored between 1020 ms and 1025 ms. 50 current sampling points are collected within this window at a sampling rate of 10 kHz. The current values ​​of the first 5 sampling points are [0.1, 0.2, 0.3, 0.8, 1.5] A, and the adjacent differences are [0.1, 0.1, 0.5, 0.7] A, corresponding to a current change rate of [100, 100, 500, 700] A / s, which is converted to [0.1, 0.1, 0.5, 0.7] A / ms. The third difference is 0.5 A / ms, reaching the preset connection threshold. The fourth difference is 0.7 A / ms, exceeding the preset connection threshold. Two consecutive sampling points meet the condition but do not meet the requirement of three consecutive differences. If the current value after the fifth sampling point is 2.4 A, the difference is 0.9 A, the rate of change is 0.9 A / ms, and three consecutive sampling points (the third, fourth, and fifth differences) all exceed 0.5 A / ms, then the connection confirmation flag is determined to be at position 1, and the pre-charging circuit 70 switches from the current limiting mode to the full-power conduction state.

Claims

1. A zero-differential-pressure blind mating connector system, characterized in that, include: The first connector has a first insulating end face, at least one power pin exposed from the first insulating end face, and an ultrasonic phased array sensor embedded in the first insulating end face. The second connector has a second insulating end face and a power socket exposed from the second insulating end face corresponding to the power pin. An acoustic reflection array opposite to the ultrasonic phased array sensor is provided on the second insulating end face. The controller is electrically connected to the ultrasonic phased array sensor to receive the sensing signal characterizing the relative pose of the first connector and the second connector, which is obtained by the ultrasonic phased array sensor by emitting ultrasonic waves into the acoustic reflection array and receiving the echoes, and to generate a trigger command based on the sensing signal. A pre-charging circuit having an equalization path electrically connected to the power pin and the power socket, the pre-charging circuit being electrically connected to the controller and configured to, in response to the trigger command, adjust the voltage difference between the power pin and the power socket to below a preset safety threshold before the power pin and the power socket make physical contact. The ultrasonic phased array sensor includes: A ring-shaped piezoelectric transducer array is embedded in the periphery of the first insulating end face and is composed of multiple independent array elements evenly distributed along the circumferential direction. A multi-channel transceiver control chip is electrically connected to the ring piezoelectric transducer array and is configured to apply pulse excitation with a predetermined phase delay to each of the independent array elements; And an acoustic matching layer covering the front end face of the annular piezoelectric transducer array; the acoustic reflection array includes multiple miniature corner reflectors; The acoustic reflection array also includes an annular metal target disk, which is embedded in the second insulating end face and corresponds to the annular piezoelectric transducer array in the axial direction.

2. The zero-differential-pressure blind mating connector system according to claim 1, characterized in that, The controller can also be operated to: Based on the sensing signals, the six-degree-of-freedom relative pose between the first connector and the second connector is calculated. The six-degree-of-freedom relative pose includes axial distance, horizontal offset, vertical offset, pitch angle, roll angle and yaw angle. Before generating the trigger command, a state vector is constructed based on the time series of the axial distance, the relative approach velocity, and the relative approach acceleration. The state vector is recursively derived with the constant rate of change of acceleration as a constraint. The relative approach velocity is iteratively estimated using the time series of the axial distance as an observation. Based on the relative approach speed and the response delay of the pre-charging circuit, the trigger command is generated before the axial distance reaches the preset trigger threshold; When the horizontal offset, vertical offset, pitch angle, or roll angle exceeds the preset alignment tolerance range, the generation of the trigger command is suppressed, and an alignment failure indication signal is generated and output to the visual indicator on the first connector.

3. The zero-differential pressure blind mating connector system according to claim 1, characterized in that, The pre-charging circuit includes: A pre-contact electrode is disposed on the first connector, and its front end face protrudes axially from the front end face of the power pin. A programmable slope constant current source, electrically connected to the pre-contact electrode and the controller, is configured to inject or draw current into the equalization path at a controlled current slope in response to the trigger command. The programmable slope constant current source includes: A digital-to-analog converter, electrically connected to the controller; An operational amplifier, the non-inverting input of which is electrically connected to the output of the digital-to-analog converter; A power metal-oxide-semiconductor field-effect transistor array, the gate of which is electrically connected to the output terminal of the operational amplifier, the drain of which is electrically connected to the equalization path, and the source of which is grounded through a sampling resistor. The voltage across the sampling resistor is fed back to the inverting input terminal of the operational amplifier to form a closed-loop constant current control. The second type of compensation network is connected between the output terminal and the inverting input terminal of the operational amplifier to make the phase margin of the closed-loop constant current control greater than a predetermined angle.

4. The zero-differential pressure blind mating connector system according to claim 3, characterized in that, The programmable slope constant current source also includes: A ramp generator, electrically connected to the digital-to-analog converter and the operational amplifier, includes an analog switch connected in series between a constant current source and an integrating capacitor, the control terminal of which is electrically connected to the controller. A ramp reset switch is connected in parallel with the integrating capacitor, and the control terminal of the ramp reset switch is electrically connected to the controller; The controller is configured to close the analog switch when the trigger command is generated, and the constant current source charges the integrating capacitor with a constant current to generate a linearly rising ramp voltage across the integrating capacitor. The ramp voltage is buffered by the operational amplifier and then drives the power metal-oxide-semiconductor field-effect transistor array. Furthermore, when the trigger command is released, the ramp reset switch is closed to discharge the integrating capacitor.

5. The zero-differential-pressure blind mating connector system according to claim 3, characterized in that, The front end face of the pre-contact electrode is provided with a plurality of micro-protrusion structures, which are evenly distributed along the circumference of the pre-contact electrode. Each micro-protrusion structure has a conical tip with a radius of curvature smaller than a preset radius value. When the pre-contact electrode and the corresponding electrode are in physical contact, the oxide film layer on the surface of the corresponding electrode is broken with concentrated pressure, thereby reducing the initial contact resistance.

6. The zero-differential-pressure blind mating connector system according to claim 3, characterized in that, The pre-charging circuit also includes: A fast discharge branch, connected in parallel between the power supply pin and the reference ground, includes a discharge resistor and a discharge switch connected in series. A constant current discharge circuit, connected in series between the discharge resistor and the discharge switch, is composed of a depletion-type metal-oxide-semiconductor field-effect transistor and a gate bias resistor. The control terminal of the discharge switch is electrically connected to the controller, which is configured to briefly close the discharge switch before the trigger command is generated, so as to discharge the residual electrostatic charge on the power pin in the form of a constant current.

7. The zero-differential-pressure blind mating connector system according to claim 1, characterized in that, Also includes: An insulation resistance monitoring circuit includes a voltage divider resistor network connected in series between the power supply pin and a reference ground, and an isolated analog-to-digital converter connected across a sampling resistor in the voltage divider resistor network. The isolated analog-to-digital converter is electrically connected to the controller and is used to provide the controller with DC leakage current data characterizing the insulation resistance between the first connector and the second connector before the power pins make physical contact with the power socket. The controller is further operable to suppress the generation of the trigger command when the DC leakage current data indicates that the insulation resistance is lower than a preset insulation threshold.

Images