A lidar transmitter and receiver prototype

Abstract

This application belongs to the field of optoelectronic detection and perception technologies, and specifically discloses a lidar transmitting and receiving prototype, including: The receiving prototype includes: a transmitting module, a receiving module, and a signal processing module. Among them, the transmitting module is used to emit optical signals to the target to be measured; the receiving module is used to receive the optical signals reflected from the target to be measured and convert them into electrical signals; the signal processing module is used to digitally process the electrical signals to obtain information about the target to be measured. This application can achieve high-precision light speed dynamic regulation within an ultra-large field of view.

Description

Technical Field

[0001] This application belongs to the technical field of optoelectronic detection and perception, and particularly relates to a lidar transmitter and receiver prototype. Background Art

[0002] As a core sensor for high-precision environmental perception, lidar has important applications in fields such as autonomous driving, robot navigation, and 3D mapping. However, traditional lidar systems have significant deficiencies in dynamic beam control and signal processing: on the one hand, the mechanical scanning structure relies on rotating components to adjust the beam direction, resulting in a large volume, high response latency, and limited reliability; on the other hand, although the emission scheme based on a static phased array can achieve electronic scanning, it is limited by insufficient phase modulation accuracy and crosstalk between channels, making it difficult to achieve high-resolution detection in a large field of view. In addition, the receiving end generally faces problems such as low capture efficiency of reflected light signals and weak noise suppression ability. Especially in complex environments, multi-path interference and phase misalignment are prominent, severely restricting the target recognition accuracy and detection distance. With the increasingly stringent requirements for real-time performance, miniaturization, and anti-interference ability in application scenarios, there is an urgent need for a new lidar architecture to break through the existing technical bottlenecks by optimizing the dynamic regulation of optical signals and intelligent signal processing mechanisms. Summary of the Invention

[0003] Aiming at the deficiencies in the prior art, the main purpose of this application is to provide a lidar transmitter and receiver prototype. This application aims to achieve high-precision dynamic regulation of the light speed within an ultra-large field of view to support the adaptive compensation mechanism in complex environments and provide a high-performance solid-state lidar solution for intelligent driving and intelligent robots.

[0004] To achieve the above objectives, this application provides the following technical solutions:

[0005] A lidar transmitter and receiver prototype, wherein the receiver prototype includes: a transmitting module, a receiving module, and a signal processing module. Among them, the transmitting module is used to transmit optical signals to the target to be measured; the receiving module is used to receive the optical signals reflected from the target to be measured and convert them into electrical signals; the signal processing module is used to digitally process the electrical signals to obtain information about the target to be measured.

[0006] Optionally, the transmitting module includes: a laser, an optical fiber coupler, a 3D transmitting chip, and an edge emitting unit connected in sequence. Among them, the laser is used to output optical signals of a certain wavelength; the optical fiber coupler is used to couple the optical signals into the optical fiber and transmit them to the 3D transmitting chip; the 3D transmitting chip is used to split the coupled optical signals and perform phase modulation on the split optical signals; the edge emitting unit is used to adjust the phase-modulated optical signals to a preset angle and transmit them to the target to be measured.

[0007] Optionally, the receiving module includes: a grating receiving chip, a dynamically tunable grating, a beam combiner, a photodetector array, a preamplifier, and a phase adjustment network connected in sequence. Among them, the grating receiving chip is used to capture the optical signal reflected by the target to be measured and focus it on a specific channel for directional transmission; the dynamically tunable grating is used to dynamically optimize the capture of the directionally transmitted optical signal; the beam combiner is used to integrate multiple optical signals optimized and captured by the dynamically tunable grating into a single optical signal; the photodetector array is used to convert the single optical signal into an electrical signal; the preamplifier is used to amplify the electrical signal; the phase adjustment network is used to adjust the phase of the amplified electrical signal and input the phase-adjusted electrical signal into the signal processing module.

[0008] Optionally, the phase adjustment network includes: an input unit, a phase modulation unit, and a feedback unit. Among them, the input unit is used to convert the voltage signal output by the preamplifier into a current signal; the phase modulation unit is used to modulate the current signal; the feedback unit is used to perform real-time phase monitoring on the modulated current signal and feedback the monitoring result to the phase modulation unit to dynamically adjust the phase modulation parameters.

[0009] Optionally, the input unit includes: a fourth resistor, a fifth resistor, a sixth resistor, a transconductance amplifier, and an eleventh capacitor. Among them, the non-inverting input terminal of the transconductance amplifier is connected to the input signal through the fourth resistor, the inverting input terminal of the transconductance amplifier is connected to the third ground terminal through the eleventh capacitor, the output terminal of the transconductance amplifier is connected to the input terminal of the phase modulation unit through the fifth resistor, the positive pin of the transconductance amplifier is connected to the +5V power supply, the negative pin is connected to the -5V power supply, the sixth resistor is a variable resistor, including a first fixed terminal, a second fixed terminal, and a sliding terminal. Among them, the first fixed terminal is connected to the +5V power supply, the sliding terminal is connected to the bias current control pin of the transconductance amplifier, and the second fixed terminal is short-circuited to the sliding terminal.

[0010] Optionally, the feedback unit includes: a phase detector, a buffer operational amplifier, and a second microcontroller. Among them, the non-inverting input terminal of the buffer operational amplifier is connected to the output terminal of the phase modulation unit, the inverting input terminal of the buffer operational amplifier is connected to the first output terminal of the phase detector, the first input terminal of the phase detector is connected to the output terminal of the phase modulation unit, the second input terminal of the phase detector is used to receive a reference clock signal, and the second output terminal of the phase detector is connected to the analog input terminal of the second microcontroller.

[0011] Optionally, the signal processing module includes a microprocessor.

[0012] Optionally, the microprocessor is built-in with a pre-trained target information recognition model, and the target information recognition model includes: an input layer, a backbone network, and a multi-task output layer.

[0013] The present application can bring the following beneficial effects:

[0014] Through the innovative optical phased array architecture and intelligent closed-loop control system, the present application has achieved significant optimization of lidar technology: adopting all-electronic phase modulation technology, realizing high-resolution beam dynamic control within an ultra-large field of view, significantly reducing the volume compared with traditional mechanical scanning schemes and eliminating the need for moving parts, thus significantly improving the system reliability; through the coordinated regulation of the distributed capacitance array and digital potentiometer, combined with the real-time error compensation mechanism, realizing high-precision phase shift control and excellent temperature stability; the unique differential noise suppression design effectively reduces environmental interference and supports high-precision ranging capabilities in complex scenarios; at the same time, the modular optoelectronic integration solution optimizes the mass production cost and power consumption, providing a high-performance, high-reliability, and low-cost solid-state lidar solution for fields such as autonomous driving and intelligent robots. Description of the Drawings

[0015] Figure 1 is the overall structural schematic diagram of a lidar transmitter and receiver prototype provided by an embodiment of the present application;

[0016] Figure 2 is the structural schematic diagram of the transmitting module;

[0017] Figure 3 is the structural schematic diagram of the edge transmitting unit;

[0018] Figure 4 is the structural schematic diagram of the receiving module;

[0019] Figure 5 is the structural schematic diagram of the input unit;

[0020] Figure 6 is the structural schematic diagram of the phase modulation unit;

[0021] Figure 7 is the structural schematic diagram of the feedback unit. Detailed Embodiments

[0022] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0023] It should be noted that all directional indications (such as up, down, left, right, front, back...) in the embodiments of this application are only used to explain the relative positional relationship and movement conditions between components in a specific posture (as shown in the attached drawings). If the specific posture changes, the directional indications will also change accordingly.

[0024] In this application, unless otherwise clearly specified and limited, the terms "connection", "fixation", etc. should be understood in a broad sense. For example, "fixation" can be a fixed connection, a detachable connection, or integrated; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two components or the interaction relationship between two components, unless otherwise clearly limited. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0025] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of this application, the descriptions of "first", "second", etc. are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one such feature. In addition, the meaning of "and / or" appearing throughout the text includes three parallel solutions. Taking "A and / or B" as an example, it includes solution A, solution B, or the solution where A and B are satisfied simultaneously. In addition, the technical solutions between various embodiments can be combined with each other, but it must be based on the fact that those of ordinary skill in the art can implement it. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0026] Figure 1 This is a kind provided by an exemplary embodiment of this application, such as Figure 1 As described above, the prediction method includes the following steps:

[0027] Figure 1 This is a schematic structural diagram of a lidar transmitting and receiving prototype provided by an exemplary embodiment of this application, as Figure 1 As shown, the receiving prototype includes: a transmitting module, a receiving module, and a signal processing module. Among them, the transmitting module is used to transmit optical signals to the target to be measured; the receiving module is used to receive the optical signals reflected from the target to be measured and convert them into electrical signals; the signal processing module is used to digitally process the electrical signals to obtain information about the target to be measured.

[0028] In another exemplary embodiment, such as Figure 2As shown, the transmitting module includes, connected in sequence: a laser, an optical fiber coupler, a 3D transmitting chip, and an edge emitting unit. Among them, the laser is used to output an optical signal of a certain wavelength; the optical fiber coupler is used to couple the optical signal into an optical fiber and transmit it to the 3D transmitting chip; the 3D transmitting chip is used to split the coupled optical signal and perform phase modulation on the split optical signal to control the deflection degree of the optical signal; the edge emitting unit is used to adjust the phase-modulated optical signal to a preset angle and emit it to the target to be measured.

[0029] In this embodiment, the transmitting module generates a laser signal with a specific wavelength through a laser, and the optical fiber coupler efficiently transmits the optical signal to the 3D transmitting chip; the 3D transmitting chip divides a single beam of light into multiple paths through an internal splitter, and dynamically adjusts the phase difference of each channel by using an internal phase modulator, and synthesizes a wavefront in a specific direction through an interference effect to achieve the deflection of the optical signal; finally, the edge emitting unit emits the phase-modulated optical signal to the target to be measured at a preset angle, thus completing the high-precision and dynamically controllable directional output of the optical signal.

[0030] It should be noted that the laser, the optical fiber coupler, and the 3D transmitting chip in this embodiment are all prior arts, and this embodiment does not involve improving their internal structures. For example, the laser can adopt the DFB laser of Lumentum, which can output a narrow linewidth optical signal of 1550 nm, has high stability and low noise, and is suitable for long-distance detection. The optical fiber coupler can adopt the FPH-1550 collimator of Thorlabs, which realizes the low-loss (<0.5 dB) coupling of a single-mode optical fiber through an aspherical lens and a precision alignment mechanism. The 3D transmitting chip can adopt the silicon photonics chip of Analog Photonics, and can divide the coupled optical signal into multiple paths (such as 16 channels) through a Y-shaped waveguide.

[0031] In another exemplary embodiment, as Figure 3As shown, the edge emission unit includes: a temperature-compensated crystal oscillator (TCXO), a digital synthesizer (DDS, such as AD9959), a first capacitor C1, a second capacitor C2, a third capacitor C3, a fourth capacitor C4, a fifth capacitor C5, a sixth capacitor C6, a seventh capacitor C7, an eighth capacitor C8, a ninth capacitor C9, a tenth capacitor C10, a first resistor R1, a second resistor R2, a third resistor R3, a first inductor L1, a second inductor L2, a third inductor L3, a fourth inductor L4, a first zener diode D1, a second zener diode D2, a first triode T1, a second triode T2, a third triode T3, a fourth triode T4, and an antenna (ANT). Among them, the first input terminal (power input terminal) of the temperature-compensated crystal oscillator is connected to the 3.3V power supply through the series-connected first capacitor C1 and second capacitor C2, the output terminal of the temperature-compensated crystal oscillator is connected to the clock input terminal (CLK_IN) of the digital synthesizer through the third capacitor C3 and the first inductor L1, and the ground terminal of the temperature-compensated crystal oscillator is connected to the first ground terminal G1; the first output terminal of the digital synthesizer is connected to the base of the first triode T1 through the fourth capacitor C4 and the second inductor L2, and the base of the first triode T1 is also connected to the positive phase control signal output terminal (3D+) of the 3D emission chip. The second output terminal of the digital synthesizer is connected to the second input terminal (temperature compensation input terminal) of the temperature-compensated crystal oscillator through the first resistor R1; the collector of the first triode T1 is connected to the first power supply V CC1 through the series-connected fifth capacitor C5 and third inductor L3. The emitter of the first triode T1 is connected to the base of the second triode T2 through the series-connected second resistor R2 and first zener diode D1. The collector of the second triode T2 is connected to the antenna through the sixth capacitor C6. The emitter of the second triode T2 is connected to the second ground terminal G2 through the eighth capacitor C8; the third triode T3 and the fourth triode T4 are symmetrically arranged with the first triode T1 and the second triode T2 respectively. The base of the third triode T3 is connected to the negative phase control signal output terminal (3D-) of the 3D emission chip. The collector of the third triode T3 is connected to the second power supply V CC2 through the series-connected seventh capacitor C7 and fourth inductor L4. The emitter of the third triode T3 is connected to the base of the fourth triode T4 through the series-connected third resistor R3 and second zener diode D2. The collector of the fourth triode T4 is connected to the antenna through the ninth capacitor C9. The emitter of the fourth triode T4 is connected to the second ground terminal G2 through the tenth capacitor C10.

[0032] In this embodiment, the edge emission unit serves as the core beam control module of the lidar system. Among them, the temperature-compensated crystal oscillator (TCXO) is driven by a 3.3V power supply to output a highly stable clock signal. After filtering out high-frequency noise through the low-pass filter network composed of the third capacitor C3 and the first inductor L1, it provides a pure clock reference for the digital synthesizer (DDS). The digital synthesizer (DDS) generates a programmable high-precision frequency signal based on the input clock. Its first output terminal injects the signal into the base of the first triode T1 through the second inductor L2 and the fourth capacitor C4, and at the same time intersects with the forward phase control signal of the 3D emission chip. The first triode T1 filters out non-target frequency band noise through the resonant circuit composed of the fifth capacitor C5 and the third inductor L3, amplifies the signal and transmits it to the second triode T2. At this time, the bias network composed of the second resistor R2 and the first zener diode D1 provides a stable operating point for the second triode T2 to ensure linear amplification of the signal without distortion. The mirror circuit composed of the third triode T3 and the fourth triode T4 synchronously processes the negative phase control signal. After resonant filtering through the seventh capacitor C7 and the fourth inductor L4, the two signals are respectively coupled to the antenna through the sixth capacitor C6 and the ninth capacitor C9 to form a differential output with complementary phases - this design is like the active noise reduction technology in acoustics, eliminating common-mode interference through the superposition of reverse signals, making the optical signal finally emitted by the antenna as pure as passing through a "noise purifier". The entire circuit realizes "electrical isolation" through the inductor-capacitor decoupling network of the dual-power supply system of the first power supply VCC1 and the second power supply VCC2, and cooperates with the overvoltage protection function of the zener diode to ensure that high-frequency signals can still maintain "zero-distortion transmission" in a complex electromagnetic environment. This design that combines a symmetric architecture, dynamic phase modulation, and multi-stage noise suppression not only achieves a beam deflection accuracy of sub-nanometer level, but also completes the real-time reconstruction of the beam direction with a microsecond-level response speed, enabling the lidar to have "light tentacles" in the complex scenarios of autonomous driving and accurately capture the fine contours of each dynamic target.

[0033] This application has upgraded the lidar system through the symmetric differential architecture and multi-stage dynamic regulation mechanism of the edge emission unit: based on the cooperation of the temperature-compensated crystal oscillator (TCXO) and the digital synthesizer (DDS), the phase modulation accuracy is improved to the sub-nanometer level, reducing the beam deflection angle error to 0.01 °The following; through the nested design of a triode differential amplifier circuit and a capacitor-inductor resonant network, the common-mode noise is suppressed by more than 40 dB in the wide frequency band range of 10 MHz - 6 GHz. Combining the phase complementary output technology at the antenna end, the signal-to-noise ratio is increased to 3.2 times that of the traditional scheme; using a silicon photonics chip and dynamic programmable phase control, the beam direction can be reconstructed within microseconds, supporting high-speed scanning with an angular resolution of 0.1° within an ultra-large field of view of 120°×90°; its innovative power isolation and temperature adaptive compensation mechanism maintains a phase stability of 0.05% in the extreme environment from -40°C to 85°C, providing centimeter-level real-time three-dimensional perception capabilities for scenarios such as autonomous driving and UAV obstacle avoidance. At the same time, the module volume is compressed to 1 / 5 of the traditional mechanical scanning scheme, having the advantages of high precision, anti-interference, and miniaturization of lidar.

[0034] In another exemplary embodiment, as Figure 4 shown, the receiving module includes a grating receiving chip, a dynamically tunable grating, a beam combiner, a photodetector array, a preamplifier, and a phase adjustment network connected in sequence. Among them, the grating receiving chip is used to capture the optical signal reflected by the target to be measured and focus it on a specific channel for directional transmission; the dynamically tunable grating is used to dynamically optimize the capture of the directionally transmitted optical signal; the beam combiner is used to integrate the multiple optical signals optimized and captured by the dynamically tunable grating into a single optical signal; the photodetector array is used to convert the single optical signal into an electrical signal; the preamplifier is used to amplify the electrical signal; the phase adjustment network is used to adjust the phase of the amplified electrical signal and input the phase-adjusted electrical signal into the signal processing module.

[0035] In this embodiment, the receiving module captures the optical signal reflected by the target to be measured through the grating receiving chip and focuses it on a specific channel. The dynamically tunable grating optimizes the receiving angle in real time to enhance the capture efficiency of the reflected optical signal; then the beam combiner combines the multi-channel optical signals into a single channel. After being converted into an electrical signal by the photodetector array, the signal is amplified by the preamplifier, and finally the phase of the electrical signal is accurately corrected by the phase adjustment network to compensate for the optical path difference and the delay between channels, and finally output to the signal processing module to extract the target information.

[0036] It should be noted that the grating receiving chip, dynamically tunable grating, beam combiner, photodetector array, and preamplifier involved in this embodiment are all existing devices, and this embodiment does not involve improvements to the internal structures of any of the above devices. For example, the grating receiving chip can use Intel's silicon photonics device, which is used to couple the optical signal input by the optical fiber to the waveguide, supports a wavelength of 1550 nm, and is applied to the receiving end of optical communication and lidar. The dynamically tunable grating can use the Meadowlark Optics liquid crystal tunable grating, which adjusts the arrangement of liquid crystal molecules through voltage to dynamically adjust the grating period to achieve wavelength selection or angle control, and is suitable for spectral imaging and lidar receiving optimization. The beam combiner can use the Thorlabs PM fiber beam combiner (model PBC-1550), which supports the combination of multiple single-mode fiber inputs into a single output, with a loss <0.3 dB, and is suitable for the synthesis of lidar multi-channel signals. The photodetector array can use the Hamamatsu S13360-3050CS silicon APD array, which is a 32-channel silicon avalanche phototransistor array with a response wavelength of 300-1000 nm and is suitable for high-speed detection in the visible to near-infrared band. The preamplifier can use the Analog Devices AD8000 series ultra-low noise amplifier, which is suitable for amplifying photodetection signals.

[0037] In another exemplary embodiment, the phase adjustment network includes: an input unit, a phase modulation unit, and a feedback unit. Among them, the input unit is used to convert the voltage signal output by the preamplifier into a current signal; the phase modulation unit is used to modulate the current signal; the feedback unit is used to perform real-time phase monitoring on the modulated current signal and feed back the monitoring result to the phase modulation unit to dynamically adjust the phase modulation parameters.

[0038] In another exemplary embodiment, as Figure 5As shown, the input unit includes a fourth resistor R4, a fifth resistor R5, a sixth resistor R6, a transconductance amplifier OTA, and an eleventh capacitor C11. Among them, the non-inverting input terminal (+) of the transconductance amplifier OTA is connected to the input signal (Vin, i.e., the voltage signal amplified by the preamplifier) through the fourth resistor R4. The inverting input terminal (-) of the transconductance amplifier OTA is connected to the third ground terminal G3 through the eleventh capacitor C11. The output terminal of the transconductance amplifier OTA is connected to the input terminal of the phase modulation unit through the fifth resistor R5. The positive pin of the transconductance amplifier OTA is connected to the +5V power supply, and the negative pin is connected to the -5V power supply. The sixth resistor R6 is a variable resistor, including a first fixed terminal, a second fixed terminal, and a sliding terminal. Among them, the first fixed terminal is connected to the +5V power supply, the sliding terminal is connected to the bias current control pin (B) of the transconductance amplifier, and the second fixed terminal is short-circuited with the sliding terminal to avoid forming a voltage division structure and ensure pure resistance adjustment.

[0039] In this embodiment, the input unit serves as the signal preprocessing core of the phase adjustment network and adopts a three-level collaborative mechanism of "transconductance conversion-noise purification-dynamic optimization". Specifically, the input signal (Vin) is first injected into the non-inverting terminal of the transconductance amplifier (OTA) after being current-limited by the fourth resistor R4, and the inverting terminal is grounded through the eleventh capacitor C11 for AC, forcing the OTA to enter the transconductance amplification mode, and linearly converting the input voltage signal into a current signal with a transconductance gain of up to 100mS; when this current signal flows through the fifth resistor R5, it is reconstructed into a low-noise voltage signal, and at the same time, it forms a low-pass filter with the eleventh capacitor C11 with a cut-off frequency of 1.6MHz (calculated by 1 / (2π×R5×C11)), effectively filtering out common high-frequency electromagnetic interference in the lidar environment (such as 5G communication band noise); the sixth resistor R6 innovatively adopts a "pure resistive sliding" design, directly adjusting the bias current of the OTA through the mechanical sliding terminal, realizing stepless adjustment of the transconductance gain in the range of 10μA to 1mA, so as to dynamically adapt to a wide-range input signal from -80dBm to +10dBm, ensuring an ultra-low harmonic distortion of 0.02% in both weak reflection signals and strong echo scenarios; the ±5V symmetric power supply architecture cleverly eliminates the DC offset of the single power supply system, making the output drift of the input unit less than 50μV in the temperature range of -40°C to 125°C. This "signal-noise-power" trinity design enables the input stage to achieve a dynamic range of 104dB in the ultra-wide frequency band from 0.1Hz to 10MHz, providing a pure input base comparable to a laboratory-level signal source for subsequent nanoscale phase modulation.

[0040] In another exemplary embodiment, as Figure 6As shown, the phase modulation unit includes the twelfth capacitor C12, the thirteenth capacitor C13, the fourteenth capacitor C14, the fifteenth capacitor C15, the first MOS transistor Q1, the second MOS transistor Q2, the third MOS transistor Q3, the first microcontroller U1, the seventh resistor R7, the eighth resistor R8, the ninth resistor R9, the tenth resistor R10, and a digital potentiometer DP (such as AD5292). Among them, the first end of the twelfth capacitor C12 is connected to the output end of the transconductance amplifier OTA. The second end of the twelfth capacitor C12 is sequentially connected to the seventh resistor R7 and the digital potentiometer DP, and then connected to the first ends of the thirteenth capacitor C13, the fourteenth capacitor C14, and the fifteenth capacitor C15 which are connected in parallel. The second ends of the thirteenth capacitor C13, the fourteenth capacitor C14, and the fifteenth capacitor C15 are respectively connected to the drains of the first MOS transistor Q1, the second MOS transistor Q2, and the third MOS transistor Q3. The sources of the first MOS transistor Q1, the second MOS transistor Q2, and the third MOS transistor Q3 are grounded (as Figure 6 shown, respectively connected to the fourth ground terminal G4, the fifth ground terminal G5, and the sixth ground terminal G6). The gates of the first MOS transistor Q1, the second MOS transistor Q2, and the third MOS transistor Q3 are respectively connected to the first GPIO pin, the second GPIO pin, and the third GPIO pin of the first microcontroller U1 through the eighth resistor R8, the ninth resistor R9, and the tenth resistor R10.

[0041] In this embodiment, the phase modulation unit realizes high-precision phase regulation through the collaborative mechanism of multi-stage capacitor switching and digital resistor adjustment. Its working principle can be decomposed into the following three core links:

[0042] 1. DC isolation and noise suppression: The twelfth capacitor C12 serves as an AC coupling capacitor to filter out the DC component in the output signal of the transconductance amplifier (OTA) (such as the bias voltage introduced by the ±5V power supply), and only allows AC signals with a frequency range of 0.1Hz - 10MHz to enter the phase modulation unit, avoiding non-linear distortion caused by the static operating point shift in the subsequent circuit. At the same time, the seventh resistor R7 (10Ω) is connected in series between the twelfth capacitor C12 and the distributed capacitor bank (C13 - C15). Using its small resistance characteristic, it suppresses the high-frequency noise (such as GHz-level harmonics) generated by the MOS transistor switching action from being fed back to the OTA output end, ensuring the purity of the input signal.

[0043] 2. Dynamic RC Parameter Regulation: The distributed capacitor bank (C13: 1 pF, C14: 10 pF, C15: 100 pF) and the digital potentiometer DP together form a Req-Ceq adjustable network. The first microcontroller U1 outputs high and low levels (3.3 V / 0 V) through the GPIO pins to control the conduction and cutoff of the first MOS transistor Q1, the second MOS transistor Q2, and the third MOS transistor Q3, realizing the binary combination switching of the capacitance value (for example, when Q1 is conducting, 1 pF is connected, and when Q2 is conducting, 10 pF is superimposed, and the total equivalent capacitance Ceq = 1 + 10 = 11 pF), supporting a capacitance adjustment range from 1 pF to 111 pF (1 + 10 + 100). At the same time, the digital potentiometer DP dynamically adjusts the equivalent resistance R eq (range 50 Ω - 10 kΩ), by changing the RC time constant τ = R eq × C eq , achieving precise control of the phase offset from 10 ns (τ = 50 Ω × 1 pF) to 1.11 ms (τ = 10 kΩ × 111 pF).

[0044] 3. Multi-Level Resolution and Anti-Interference Design:

[0045] Coarse Tuning Mode: Achieve 10 ° -level phase stepping through capacitor combination switching (for example, when Ceq = 100 pF, τ = 10 kΩ × 100 pF = 1 μs, corresponding to 10 ° @1 MHz signal);

[0046] Fine Tuning Mode: Combine the 1 Ω -level resistance fine tuning of the digital potentiometer DP to achieve 1 ° -level phase accuracy (for example, when Req is adjusted from 100 Ω to 101 Ω, the phase offset change is 0.8° @ 1 MHz);

[0047] Noise Suppression: The physical isolation design of the MOS transistors (the sources of Q1 to Q3 are independently grounded to G4 to G6) blocks the crosstalk between channels, and the distributed capacitor layout reduces the parasitic inductance, making the phase offset error less than ±0.5° within the 10 MHz - 1 GHz wide frequency band, and the signal-to-noise ratio is increased to more than 70 dB.

[0048] The phase regulation unit completes the full-range phase calibration from coarse tuning to fine tuning within 1 μs through a hybrid control strategy of "capacitor discrete switching + resistor continuous adjustment", and at the same time, with an equivalent capacitance resolution of <0.1 pF and a 1 Ω resistor stepping accuracy, supports the lidar system to achieve dynamic beam scanning with an angular resolution of 0.1 ° in a 120 ° field of view, providing sub-nanometer optical path difference compensation ability for real-time 3D perception in complex environments.

[0049] It should be noted that in the actual circuit design, the drains of the first MOS transistor Q1, the second MOS transistor Q2, and the third MOS transistor Q3 are short-circuited by wires on the circuit board to form a common node (not shown in the figure), and this common node is the output terminal of the phase modulation unit.

[0050] In another exemplary embodiment, as Figure 7 shown, the feedback unit includes a phase detector PD (such as AD8302), a buffer operational amplifier (such as ADA4817), and a second microcontroller U2 (such as STM32G4). Among them, the non-inverting input terminal of the buffer operational amplifier is connected to the output terminal of the phase modulation unit, the inverting input terminal of the buffer operational amplifier is connected to the first output terminal of the phase detector, the first input terminal of the phase detector is connected to the output terminal of the phase modulation unit, the second input terminal of the phase detector is used to receive a reference clock signal, and the second output terminal of the phase detector is connected to the analog input terminal of the second microcontroller U2.

[0051] In this embodiment, the output signal of the phase modulation unit is divided into a first path signal and a second path signal. Among them, the first path signal is sent to the non-inverting input terminal of the buffer operational amplifier. After being buffer-amplified, the first path signal is divided into two paths again. One path is sent to the signal processing module for subsequent processing, and the other path is sent to the first analog input terminal of the second microcontroller U2 to monitor the signal characteristics in real time; the second path signal is sent to the first input terminal of the phase detector PD, compared with the signal from the reference clock signal source, generates an error voltage reflecting the phase difference between the two, and sends this error voltage to the second microcontroller U2 through the second analog input terminal of the second microcontroller U2. The inverting input terminal of the buffer operational amplifier is connected to the first output terminal of the phase detector to form a local negative feedback, which helps to stabilize the amplification process and improve the signal quality. Based on the error voltage provided by the phase detector, the second microcontroller U2 analyzes the error voltage through the built-in PID control algorithm shown below:

[0052]

[0053] Among them, represents the error voltage; ,, respectively represent the proportional, integral, and differential gains; represents the error voltage from the initial time 0 to the current time t cumulative sum; is the error voltage the first derivative of with respect to time t, representing the instantaneous change rate of the error voltage; represents according to the error voltage The calculated control quantity is used to adjust the phase modulation unit to make its output close to the target value.

[0054] Furthermore, the second microcontroller U2 dynamically adjusts the control parameters of the phase modulation unit (such as the resistance value of the digital potentiometer, the switching combinations of the first to third MOS transistors) through serial communication (such as UART) and the SPI bus to compensate for the phase shift, so as to ensure the phase synchronization accuracy (such as sub-nanosecond phase stability) between the second path signal in the output signal of the phase modulation unit and the reference clock signal. The feedback unit eliminates the phase deviation caused by factors such as environmental interference and temperature drift through real-time feedback and dynamic adjustment, improving the noise resistance and detection accuracy of the lidar system.

[0055] Next, the present application will describe in detail the dynamic adjustment process of the control parameters of the phase modulation unit in combination with specific data:

[0056] Step 1: The phase detector (PD) continuously receives the output signal (the signal to be calibrated) of the phase modulation unit and the reference clock signal. By comparing the phase differences of the two signals, an error voltage proportional to the time difference is generated. For example, the current error voltage is 1.32 V. The second microcontroller U2 converts it into a time difference Δt = 2 ns (the proportional coefficient is 1.65 V / ns) through the built-in analog-to-digital converter (ADC).

[0057] Step 2: The second microcontroller U2 adjusts the equivalent resistance and capacitance through the built-in PID control algorithm to make the time difference Δt approach 0. Specifically, when Δt > 1 ns, the MOS transistor combination is preferentially switched to change the equivalent capacitance. For example, the current equivalent capacitance C = 10 pF (Q1 off, Q2 on, Q3 off). If Δt = 2 ns, the equivalent capacitance C needs to be increased to 100 pF (Q3 conducts, Q1 and Q2 are closed). When Δt < 1 ns, the resistance is finely adjusted through the digital potentiometer (step 1 Ω). For example, if Δt = 0.3 ns, R needs to be increased from 500 Ω to 505 Ω to further reduce Δt.

[0058] Step 3: The second microcontroller U2 sends the target capacitance combination (such as "enable Q3") to the first microcontroller U1 through serial communication (such as UART). The first microcontroller U1 sets the GPIO pin 1 to high level (conducts Q3), and pins 2 and 3 to low level (closes Q1 and Q2). At this time, the equivalent capacitance is switched from 10 pF to 100 pF, and the phase offset τ increases by 10 times (for example, from 1 μs to 10 μs), and the time difference Δt decreases from 2 ns to 1.2 ns.

[0059] Step 4: The second microcontroller U2 sends a 16-bit instruction to the digital potentiometer via the SPI bus, which contains the target resistance value (e.g., 500 Ω). The internal resistance network of the digital potentiometer adjusts the resistance value according to the instruction, for example, increasing from 480 Ω to 500 Ω. At this time, the RC time constant τ = 500 Ω × 100 pF = 50 ns, and the phase shift is further compensated by 0.8 ns, and Δt decreases from 1.2 ns to 0.4 ns.

[0060] Step 5: The phase detector updates the error voltage, and the second microcontroller U2 reads the new Δt (such as 0.4 ns).

[0061] If Δt > 0.1 ns (the target threshold), continue to finely adjust R. Exemplarily, increase R from 500 Ω to 503 Ω, and Δt further decreases to 0.1 ns. When Δt < 0.1 ns is detected continuously for 3 times, the system enters the steady state and stops adjusting.

[0062] In another exemplary embodiment, the signal processing module includes a microprocessor.

[0063] In this embodiment, the microprocessor is built-in with a pre-trained target information recognition model (the microprocessor can use, for example, Texas Instruments TDA4VM processor, which integrates a deep learning accelerator (C7xDSP + MMA), supports real-time multitasking, and can efficiently run the pre-trained target information recognition model). The target information recognition model includes: an input layer, a backbone network, and a multi-task output layer. Among them, the input layer is used to input the electrically signal after phase adjustment and convert the electrically signal into a time-frequency diagram through complex Morlet wavelet transform;

[0064] The backbone network includes a dual-stream feature extraction module and a gated attention fusion module. Among them, the dual-stream feature extraction module includes an amplitude stream and a phase stream. The amplitude stream takes the spectral amplitude matrix of the time-frequency map (size: H×W×C) as input. First, the number of channels of the spectral amplitude matrix is expanded through 1×1 convolution (for example, the number of channels of the spectral amplitude matrix is C, and the number of channels after convolution expansion is 4C) to enhance the feature expression ability. Secondly, local feature extraction is performed on the spectral amplitude matrix with expanded number of channels through depthwise separable convolution. Specifically, the depthwise separable convolution includes depth convolution (3×3 convolution kernel) and pointwise convolution (1×1 convolution kernel). Among them, depth convolution is used to extract spatial features channel by channel for the spectral amplitude matrix with expanded number of channels, and pointwise convolution is used to fuse the spatial features extracted channel by channel to obtain fused channel information (the number of channels is compressed to 2C). A GeLU activation function is set after the depthwise separable convolution to enhance the non-linear expression ability. Then, global context capture is performed on the fused channel information through dilated convolution (5×5 convolution kernel (dilation rate = 2)), and the receptive field is expanded through the dilation rate to extract long-distance dependence features. Finally, the outputs of the separable convolution and the dilated convolution are concatenated by channel through a concatenation layer, and then compressed to C channels through 1×1 convolution, and finally an amplitude stream feature map with C channels is obtained.

[0065] The phase stream takes the phase angle matrix of the time-frequency map (generated by complex Morlet wavelet transform, containing the phase information of the signal at different frequencies and time points) as input. First, the phase angle matrix is filtered by a multi-directional Gabor filter bank (filter parameters: direction: four groups of filters (0°, 45°, 90°, 135°), covering horizontal and vertical symmetric directions. Wavelength (λ): 4 pixels, controlling the filter period and matching the typical lidar target size. Standard deviation (σ): 2 pixels, determining the smoothing range of the Gaussian window and balancing the positioning accuracy and noise resistance. Phase shift (φ): 0°, keeping the filter aligned with the input phase angle) to generate four groups of direction response feature maps (size: H×W×C), and local contrast normalization is performed on each group of direction response feature maps to suppress illumination or noise interference. Secondly, the four groups of normalized direction response feature maps are concatenated along the channel dimension (output size: H×W×4C); finally, the concatenated feature map is dimension-reduced through 1×1 convolution to generate a phase stream feature map (size: H×W×C).

[0066] The gated attention fusion module takes the amplitude stream feature map and the phase stream feature map as input, and the gated attention fusion module can be represented by the following formula:

[0067]

[0068] Among them, Denotes the fused feature map, including complementary enhancement of dual-stream information; Denotes the amplitude stream feature map, with a scale of height H, width W, and number of channels C, containing spectral amplitude information; Denotes the phase stream feature map, with the same size as the amplitude stream feature Figure 1 and contains direction-sensitive features of the phase angle; Denotes the spatial weight matrix, and , generates spatial weights through the phase stream feature to calibrate the important regions in the amplitude stream; Denotes the channel weight vector, and , generates channel weights through global average pooling of the amplitude stream feature to calibrate the key channels in the phase stream, Denotes the weight matrix of the fully connected layer, used to map the pooled feature to a channel weight vector; Denotes the adaptive gating coefficient, which generates a channel-level gating coefficient through a fully connected layer after concatenating the dual-stream features to dynamically adjust the fusion ratio, Denotes the weight matrix of the fully connected layer; Denotes element-wise multiplication; Denotes the activation function.

[0069] The gated attention fusion module achieves efficient feature fusion through a cross-weighting mechanism and dynamic gating adjustment. Its core advantages are as follows: First, by using the bidirectional complementary enhancement of the amplitude stream and the phase stream, it generates spatial attention weights through the phase stream to focus on the target edge region, while the amplitude stream filters the key channel information of the phase stream, significantly improving the analytical ability for complex scenes; Second, it adopts a lightweight design, relying only on 1×1 convolution, global pooling, and fully connected layers, reducing the number of parameters by about 2 / 3 compared with traditional attention mechanisms, and adaptively balancing the contributions of dual-stream features through dynamically generated gating coefficients to avoid overfitting problems caused by fixed weights; In addition, this module suppresses background noise through spatial attention, strengthens effective information through channel attention, and further filters out unstructured interference through the gating mechanism, greatly enhancing noise robustness.

[0070] The multi-task output layer uses the fused feature map as the output. The multi-task output layer includes a 1×1 convolutional layer, which compresses the number of channels of the fused feature map to C / 2 through 1×1 convolution to generate a shared feature basis F base . An adaptive feature routing module is set after the 1×1 convolutional layer. The adaptive feature routing module first processes the shared feature basis F baseChannel-level global average pooling (GAP) is performed to extract the spatial statistical vector v of the feature base. Then, a weighted feature map is generated through two fully connected layers and distributed to the juxtaposed object detection head and distance estimation head. The object detection head extracts the local sensitive features (such as edges, corners, textures, etc.) in the spatial dimension of each channel in the weighted feature map through 3×3 depthwise separable convolution, and further outputs the bounding box parameters (center offsets x / y, width and height scaling factors w / h) and confidence scores (range [0,1], indicating the probability of the presence of an object at that position) at each spatial position (H×W) through 1×1 convolution. The distance estimation head compresses the weighted feature map to C / 8 channels through 1×1 convolution, then extracts the significant features through global max pooling, and finally outputs a 1D distance value through a fully connected layer.

[0071] The effects of the above model are verified in this application. Specifically, on the KITTI dataset (including 7,481 training samples), the detection mAP (mean average precision) using only the amplitude flow (only spectral amplitude) is 78.3%, the mAP using only the phase flow (only phase angle direction features) is 72.1%, while the mAP after the fusion of the two flows is increased to 85.6%, with a relative increase of 9.3% compared to the single flow. In addition, the IoU (intersection over union) for occluded objects is increased from 0.68 for the single flow to 0.75, especially showing significant performance on small objects such as pedestrians and bicycles.

[0072] In this application, the amplitude information and phase information of the time-frequency map are independently analyzed through the two-stream feature extraction mechanism. By combining the global features of the spectral amplitude and the phase direction sensitive features, the ability to represent the shape, texture, and motion characteristics of the object can be enhanced, especially suitable for the accurate recognition of object contours in complex backgrounds. In addition, by focusing on the key regions of the object through spatial weights, screening out effective information through channel weights, and adaptively adjusting the fusion ratio of the two-stream features, background noise and interference signals can be significantly suppressed, and the robustness of the model to low-contrast objects or occlusion scenarios can be improved.

[0073] In summary, the object detection model can achieve high-precision and high-robustness object recognition and ranging capabilities, providing intelligent environmental perception core support for lidar systems.

[0074] The above are only the preferred embodiments of this application, and do not limit the patent scope of this application. Any equivalent structure or equivalent process transformation made using the content of the specification and drawings of this application, or directly or indirectly applied in other related technical fields, shall be included in the patent protection scope of this application by the same token.

Claims

1. A prototype of a lidar transmitter and receiver, characterized in that, The receiving prototype includes: a transmitting module, a receiving module, and a signal processing module, wherein, the transmitting module is used for transmitting an optical signal to the target to be measured; the receiving module is used for receiving the optical signal reflected from the target to be measured and converting it into an electrical signal; the receiving module includes: a grating receiving chip, a dynamically tunable grating, a beam combiner, a photodetector array, a preamplifier, and a phase adjustment network connected in sequence, wherein, the grating receiving chip is used for capturing the optical signal reflected by the target to be measured and focusing it to a specific channel for directional transmission; the dynamically tunable grating is used for dynamically optimizing the capture of the directionally transmitted optical signal; the beam combiner is used for integrating multiple optical signals optimized and captured by the dynamically tunable grating into a single optical signal; the photodetector array is used for converting the single optical signal into an electrical signal; the preamplifier is used for amplifying the electrical signal; the phase adjustment network is used for adjusting the phase of the amplified electrical signal and inputting the phase-adjusted electrical signal into the signal processing module; the phase adjustment network includes: an input unit, a phase modulation unit, and a feedback unit, wherein, the input unit is used for converting the voltage signal output by the preamplifier into a current signal; the phase modulation unit is used for modulating the current signal; the feedback unit is used for performing real-time phase monitoring on the modulated current signal and feeding back the monitoring result to the phase modulation unit to dynamically adjust the phase modulation parameters; the input unit includes: a fourth resistor, a fifth resistor, a transconductance amplifier, and an eleventh capacitor, wherein, the non-inverting input terminal of the transconductance amplifier is connected to the input signal through the fourth resistor, the inverting input terminal of the transconductance amplifier is connected to the third ground terminal through the eleventh capacitor, the output terminal of the transconductance amplifier is connected to the input terminal of the phase modulation unit through the fifth resistor, the positive pin of the transconductance amplifier is connected to the +5V power supply, and the negative pin is connected to the -5V power supply; the signal processing module is used for digitally processing the electrical signal to obtain information about the target to be measured.

2. The lidar transmitting and receiving prototype according to claim 1, wherein the transmitting module includes: a laser, an optical fiber coupler, a 3D transmitting chip, and an edge emitting unit connected in sequence, wherein, the laser is used for outputting an optical signal with a certain wavelength; the optical fiber coupler is used for coupling the optical signal into an optical fiber and transmitting it into the 3D transmitting chip; the 3D transmitting chip is used for splitting the coupled optical signal and performing phase modulation on the split optical signal; the edge emitting unit is used for adjusting the phase-modulated optical signal to a preset angle and emitting it to the target to be measured.

3. The lidar transmitting and receiving prototype according to claim 1, wherein, the input unit further includes: a sixth resistor, the sixth resistor is a variable resistor, including a first fixed terminal, a second fixed terminal, and a sliding terminal. The first fixed terminal is connected to the +5V power supply, the sliding terminal is connected to the bias current control pin of the transconductance amplifier, and the second fixed terminal is short-circuited with the sliding terminal.

4. The lidar transmitting and receiving prototype according to claim 1, wherein the feedback unit includes: a phase detector, a buffer operational amplifier, and a second microcontroller, wherein, the non-inverting input terminal of the buffer operational amplifier is connected to the output terminal of the phase modulation unit, The inverting input terminal of the buffer operational amplifier is connected to the first output terminal of the phase detector. The first input terminal of the phase detector is connected to the output terminal of the phase modulation unit. The second input terminal of the phase detector is used to receive a reference clock signal. The second output terminal of the phase detector is connected to the analog input terminal of the second microcontroller.

5. The lidar transmitting and receiving prototype according to claim 1, characterized in that, The signal processing module includes a microprocessor.

6. The lidar transmitting and receiving prototype according to claim 5, wherein The microprocessor is built-in with a pre-trained target information recognition model, and the target information recognition model includes: an input layer, a backbone network, and a multi-task output layer.

Images