Method and system for self-calibration and dynamic compensation of multi-channel light source trigger link delay

Abstract

The present application relates to the technical field of multi-channel light source control, and particularly provides a multi-channel light source trigger link delay self-calibration and dynamic compensation method and system, which comprises the following steps: a master control unit measures a round-trip total delay value between transmission and reception, subtracts a known fixed delay from the round-trip total delay value to obtain a double one-way propagation delay, and then divides the double one-way propagation delay by two to obtain a one-way propagation delay of each channel; differences between receiving time instants of each channel and the one-way delay of each channel are compared, a relative delay deviation matrix is generated after all cross combinations are measured, and a channel-to-channel trigger time instant deviation calibration value is extracted from the relative delay deviation matrix; the channel-to-channel trigger time instant deviation calibration value is combined with an exponential weighted moving average filter to predict a delay drift trend; and a phase step control parameter of a programmable delay line of each channel is output by a delay drift trend adaptive proportional integral controller. The present application maintains synchronization performance of full-channel trigger signals in sub-nanosecond level through a dynamic compensation mechanism.

Description

Technical Field

[0001] This invention relates to the field of multi-channel light source control technology, and in particular to a self-calibration and dynamic compensation method and system for multi-channel light source trigger link delay. Background Technology

[0002] Multi-channel light source systems, including semiconductor laser arrays, LED arrays, and pulsed lasers, are widely used in fields such as lidar, machine vision light sources, high-power laser devices, and biomedical imaging. In applications requiring synchronous triggering of multiple light sources, such as lidar, high-speed photography, and particle physics experiments, ensuring strict synchronization of the leading edges of the optical pulses in each channel is a core performance indicator. Due to factors such as differences in the transmission paths of different channels, temperature drift, and device aging, the propagation delay of the trigger link changes dynamically over time. Traditional fixed delay compensation methods are difficult to meet the requirements of long-term stability and environmental adaptability.

[0003] Existing technologies include round-trip time measurement methods, cross-test synchronization systems, adaptive delay compensation devices, and delay calibration based on time-delay calibration (TDC). Traditional round-trip measurements require the assumption of symmetrical transmit and receive paths, and differences in actual line impedance and connector losses lead to deviations in single-trip delay calculations. Existing cross-test measurements rely on external clock references and cannot establish an absolute time base independently. Fixed-parameter PI controllers compensate for slow-varying delays caused by temperature by up to several minutes. Independent TDC solutions require measurement circuits for each channel. Summary of the Invention

[0004] To achieve the above objectives, the present invention adopts the following technical solution: One aspect of the present invention provides a self-calibration and dynamic compensation method for multi-channel light source triggering link delay, comprising the following steps: The main control unit emits a short measurement pulse with a width of less than 1 nanosecond. The short measurement pulse is coupled to the same transmission path via a bidirectional transceiver. The transmission path connects the main control unit to the light source driving modules of each channel. After receiving the short measurement pulse, the driving module triggers a reflection loop with a known fixed delay. The reflection loop generates an echo pulse and returns to the main control unit along the original transmission path. The main control unit measures the total round-trip delay between transmission and reception, subtracts the known fixed delay from the total round-trip delay to obtain twice the one-way propagation delay, and then divides by two to obtain the one-way propagation delay of each channel. Using the single-path propagation delay of each channel as a benchmark, one channel is selected sequentially as the active transmitter; the channel driving module sends cross-check pulses to adjacent channels or selected channels; the receiving channel measures the arrival time of the short pulse; the difference between the receiving time of each channel and its own single-path delay is compared; after all cross-combined measurements, a relative delay deviation matrix is ​​generated; and the trigger time deviation calibration value between channels is extracted from the relative delay deviation matrix. The calibration value of the inter-channel trigger timing deviation is input into the adaptive proportional-integral controller. Combined with the exponentially weighted moving average filter, the delay drift trend is predicted. The adaptive proportional-integral controller outputs the phase step control parameters of the programmable delay line of each channel. The phase step control parameters adjust the timing of the trigger signal application of each channel in real time, so as to achieve closed-loop synchronous compensation of the trigger pulse leading edge in the entire channel range.

[0005] Another aspect of the present invention provides a self-calibration and dynamic compensation system for multi-channel light source triggering link delay, comprising: The transmission path processing module is used by the main control unit to send short measurement pulses with a width of less than 1 nanosecond. The short measurement pulses are coupled to the same transmission path via a bidirectional transceiver. The transmission path connects the main control unit with the light source driving modules of each channel. After receiving the short measurement pulses, the driving modules trigger reflection loops with a known fixed delay. The reflection loops generate an echo pulse and return to the main control unit along the original transmission path. The main control unit measures the total round-trip delay between transmission and reception, subtracts the known fixed delay from the total round-trip delay to obtain twice the one-way propagation delay, and then divides by two to obtain the one-way propagation delay of each channel. The matrix verification generation module is used to select one channel as the active transmitter by taking the single-path propagation delay of each channel as a reference. The channel driving module sends cross-verification pulses to adjacent channels or selected channels. The receiving channel measures the arrival time of the short pulse. The difference between the receiving time of each channel and its own single-path delay is compared. After all cross-combined measurements, a relative delay deviation matrix is ​​generated. The trigger time deviation calibration value between channels is extracted from the relative delay deviation matrix. The trigger signal application module is used to input the calibration value of the trigger time deviation between channels into the adaptive proportional-integral controller. Combined with the exponentially weighted moving average filter, the delay drift trend is predicted. The adaptive proportional-integral controller outputs the phase step control parameters of the programmable delay line of each channel. The phase step control parameters adjust the timing of the trigger signal application of each channel in real time to achieve closed-loop synchronous compensation of the leading edge of the trigger pulse in the entire channel range.

[0006] This invention achieves self-calibration and real-time delay compensation of multi-channel light source triggering links through a three-level closed-loop control architecture of time delay measurement, cross-verification, and dynamic compensation; it can effectively eliminate inter-channel triggering jitter caused by factors such as transmission path differences and temperature drift, and meet the technical requirements of high-precision synchronous triggering systems. Attached Figure Description

[0007] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1This is a flowchart of the self-calibration and dynamic compensation method for multi-channel light source triggering link delay provided in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the self-calibration and dynamic compensation method for multi-channel light source triggering link delay provided in Embodiment 1 of the present invention; Figure 3 This is a process diagram of the main control unit measuring the total round-trip delay between transmission and reception, as provided in Embodiment 3 of the present invention; Figure 4 This is a process diagram of generating the relative delay deviation matrix after all cross-combination measurements, as provided in Embodiment 4 of the present invention; Figure 5 This is a flowchart illustrating the process of the adaptive proportional-integral controller providing the adaptive proportional-integral controller outputting the phase step control parameters of each channel's programmable delay line in Embodiment 7 of the present invention. Figure 6 This is a block diagram of the self-calibration and dynamic compensation system for multi-channel light source triggering link delay provided in Embodiment 11 of the present invention; Figure 7 A block diagram of the electronic device provided by the present invention; Figure 8 A block diagram of a computer-readable storage medium provided for this invention.

[0008] Reference numerals: 1. Transmission path processing module; 2. Matrix verification generation module; 3. Trigger signal application module; 4. Central processing unit / microprocessor / main control chip; 5. Storage medium; 6. Data bus; 7. Input / output bus / external bus / device bus; 8. Display; 9. Input / output device; 10. Computer-readable instructions; 11. Non-transitory computer-readable storage medium. Detailed Implementation

[0009] The technical solutions of the present invention will now be described with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0010] Hereinafter, the terms "first," "second," etc., are used for descriptive convenience only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "multiple" means two or more.

[0011] In this invention, unless otherwise explicitly specified and limited, the term "connection" should be interpreted broadly. For example, a connection can be a fixed mechanical connection, a detachable mechanical connection, or an integral part; or, a connection can be a direct connection or an indirect connection through an intermediate medium. Furthermore, unless otherwise explicitly specified and limited, the term "coupling" should be interpreted broadly. For example, coupling can be a direct electrical connection, such as physical contact and electrical conduction between two components; it can also be understood as an electrical connection between different components in a circuit structure through physical lines capable of transmitting electrical signals, such as copper foil or wires on a printed circuit board (PCB), to transmit electrical signals; or, coupling can be an indirect electrical connection between two components through an intermediate medium; or, coupling can be an electrical connection between two components in a non-contact manner, such as an electrical connection between two components using capacitive coupling to transmit electrical signals.

[0012] In this embodiment of the invention, directional terms such as up, down, left, and right may be defined relative to the orientation of the components shown in the accompanying drawings. It should be understood that these directional terms can be relative concepts, used for relative description and clarification, and can change accordingly depending on the orientation of the components in the accompanying drawings.

[0013] This invention comprises a main control unit transmitting narrow pulses, which are transmitted along the transmission path to the light source driving modules of each channel via a bidirectional transceiver, triggering a known fixed-delay reflection loop. The echo returns along the original path. The total round-trip delay is measured, and the fixed delay is subtracted to obtain twice the single-trip delay, which is then halved to obtain the single-trip propagation delay of each channel. Based on the single-trip delay, each channel polls and transmits cross-check pulses. A relative delay deviation matrix is ​​generated through all combined measurements, and the deviation calibration value of the triggering time between channels is extracted. The deviation calibration value is used by an adaptive PI controller and an exponentially weighted moving average filter to predict the drift trend, and the output phase step control parameters are used to adjust the triggering timing to achieve closed-loop synchronous compensation for all channels. This invention provides a self-calibration and dynamic compensation method for the triggering link delay of a multi-channel light source. It eliminates asymmetric errors by measuring the round-trip delay along the same path, constructs a cross-check matrix between channels to achieve referenceless calibration, and combines adaptive PI control and a drift prediction filter to achieve long-term sub-nanosecond synchronization of the leading edge of the trigger pulse of each channel without the need for an external clock source or additional measurement hardware.

[0014] Example 1: As Figure 1 As shown, this embodiment of the invention provides a self-calibration and dynamic compensation method for multi-channel light source triggering link delay, specifically including the following steps: Step S100: The main control unit sends out a short measurement pulse with a width of less than 1 nanosecond. The short measurement pulse is coupled to the same transmission path via a bidirectional transceiver. The transmission path connects the main control unit to the light source driving modules of each channel. After receiving the short measurement pulse, the driving module triggers a reflection loop with a known fixed delay. The reflection loop generates an echo pulse and returns to the main control unit along the original transmission path. The main control unit measures the total round-trip delay between transmission and reception, subtracts the known fixed delay from the total round-trip delay to obtain twice the one-way propagation delay, and then divides by two to obtain the one-way propagation delay of each channel. Step S200: Using the single-path propagation delay of each channel as a reference, select one channel as the active transmitter in sequence; the channel driving module sends cross-check pulses to adjacent channels or selected channels; the receiving channel measures the arrival time of the short pulse; compare the difference between the receiving time of each channel and its own single-path delay, generate a relative delay deviation matrix after all cross-combined measurements, and extract the inter-channel trigger time deviation calibration value from the relative delay deviation matrix. Step S300: Input the inter-channel trigger time deviation calibration value into the adaptive proportional-integral controller, combine it with the exponentially weighted moving average filter to predict the delay drift trend, and the adaptive proportional-integral controller outputs the phase step control parameters of the programmable delay line of each channel; the phase step control parameters adjust the timing of the trigger signal application of each channel in real time to achieve closed-loop synchronous compensation of the trigger pulse leading edge in the entire channel range.

[0015] The main control unit is an FPGA or microcontroller, integrating a high-precision time-to-digital converter or time interval measurement circuit. Its specific functions include: initiating measurement short pulses and accurately recording the transmission time; receiving echo pulses from each channel and recording their arrival time; calculating the total round-trip delay and subtracting the known fixed delay to obtain the one-way propagation delay; scheduling cross-validation measurement sequences between channels; running an adaptive proportional-integral controller and an exponentially weighted moving average filter; outputting phase step control parameters to the programmable delay line of each channel; and being directly related to the self-calibration of the multi-channel light source trigger link delay. Without the unified time base and computing power of the main control unit, centralized measurement and global compensation of the delay of all channels cannot be achieved. A measurement pulse less than 1 nanosecond is a digital or level pulse with a width strictly controlled within 1 nanosecond. The purpose of a narrow pulse is to ensure that there is almost no overlap between the leading and trailing edges of the pulse, which facilitates accurate measurement of round-trip time; to avoid the pulse width itself masking the tiny delay differences in the propagation link; and to ensure that the time resolution of round-trip measurement can reach the picosecond level. The pulse is coupled to the transmission path through a bidirectional transceiver and serves as the sole excitation signal for detecting delay. Its width of less than 1 nanosecond is a physical prerequisite for achieving high-precision round-trip measurement. A bidirectional transceiver is a three-terminal device: a main control terminal, a line terminal, and a driver module terminal. Internally, it includes a transmit buffer and a receive buffer, both sharing a single line terminal. Specific implementations can be on PCB traces or cables, using open-drain or tri-state gate circuits for time-division multiplexing of direction. In fiber optic links, wavelength division multiplexing or optical circulators at the same wavelength are used. The bidirectional transceiver achieves forward signal transmission and backward echo return using the same physical path, solving the hardware cost and installation complexity issues associated with traditional solutions requiring two independent paths (one for forward and one for backward). It also ensures that round-trip measurements are performed under identical media conditions, avoiding errors caused by path asymmetry. The transmission path refers to the physical signal channel from the main control unit to the driver modules of each channel light source. Specifically, it can be a coaxial cable, twisted pair, PCB microstrip / stripline, or single-mode / multimode fiber. The trigger signal is transmitted from the main control unit forward through this path to the driver module; the echo pulse returns from the driver module to the main control unit via the exact same path. The propagation characteristics of the path, such as unit length delay, temperature coefficient, and aging drift, are the objects to be measured and compensated in this embodiment. Since the round-trip measurements use the same path, the one-way propagation delay can be directly obtained from half of the total round-trip delay without assuming that the uplink and downlink paths are symmetrical.Each light source channel is equipped with a driver module, which includes: a signal receiving circuit to receive trigger or measurement pulses from the main control unit; a reflection loop with a known fixed delay; a cross-check pulse sending / receiving switch; and a light source driving power stage for ultimately illuminating the laser or LED. Its function is not only to drive the light source, but more importantly, to serve as the terminal node for delay measurement—it immediately triggers the reflection loop upon receiving a pulse, forming a closed-loop measurement loop with the main control unit. Without the cooperation of the module, the single-path propagation delay cannot be obtained online. A known fixed-delay reflection loop is a circuit with a precise and stable time delay. It can be implemented as follows: a fixed number of logic gate chains, such as an even number of inverters; a delay line of a specific length; or a clock-locked trigger toggle circuit. The delay value of the loop is precisely calibrated in advance, for example, 5 ns ± 10 ps, ​​and changes very little with temperature and voltage. After the drive module receives the measurement short pulse, it generates an echo pulse after the known fixed delay. The main control unit subtracts this known value from the total round-trip delay to obtain twice the one-way propagation delay. Converting the one-way delay, which cannot be directly measured, into the difference between the measurable round-trip delay and the known delay is the key to realizing the measurement of one-way delay without additional hardware. The echo pulse is an electrical or optical signal generated by the reflection loop and returned to the main control unit along the original transmission path via a bidirectional transceiver. Its width can be the same as or slightly wider than the measurement short pulse, but its leading edge must be stably detectable by the main control unit. The time interval between the echo pulse and the transmitted pulse includes: the bidirectional propagation delay of the transmission path + the known fixed delay of the reflection loop. By measuring this interval and subtracting the known fixed delay, the one-way propagation delay can be calculated. The existence of the echo pulse eliminates the need for the main control unit to independently install time measurement equipment on each channel side, greatly simplifying the system structure. In step S200, the active transmitter sequentially selects one channel from all N channels and configures its drive module to transmit mode; this is called the active transmitter. The channel emits a cross-check pulse, which can be a narrow pulse identical to the measurement short pulse. The pulse is transmitted to other channels via a dedicated connection between channels or through a common bus. The selection of the active transmitter is polling, ensuring that the relative delay between each pair of channels can be measured. This solves the reference deviation problem that may exist in the one-way propagation delay measurement, eliminating common errors through mutual testing. The channel driver module and the channel light source driver module in item 5 are the same entity. In the cross-checking step, the same module can be used as both an active transmitter and a receiver. The module integrates a signal switching switch, which can switch between reflection loop mode and cross-checking transmit / receive mode. This allows for hardware reuse and eliminates the need for additional dedicated cross-checking circuitry.The cross-check pulse is a test pulse with a width of less than 1 nanosecond, emitted by the active transmitter and received by another channel or receiver. The pulse bypasses the main control unit and is transmitted directly between the channel driver modules. By measuring the time the receiver receives the pulse and the time the transmitter emits the pulse, relative to their respective local clocks or relative to a common trigger edge, the relative delay between the two channels can be obtained. This feature complements the round-trip measurement in step S100: the round-trip measurement provides the absolute one-way delay, while the cross-check provides the relative delay between channels relative to the main control unit. Only by jointly solving both can the trigger time deviation calibration value of each channel relative to the reference channel be obtained. The receiver refers to the channel driver module currently configured to receive the cross-check pulse. It contains a high-speed comparator or trigger to detect the pulse leading edge and record that moment. The receiver also records its own one-way propagation delay obtained in step S100. By comparing the actual arrival time of the received pulse with the expected arrival time calculated based on its own one-way delay, the additional deviation relative to the active transmitter can be extracted. All measurement results from the receivers are summarized in the main control unit to form a relative delay deviation matrix. The relative delay deviation matrix is ​​an N×N matrix (N being the number of channels), where the (i, j)th element represents the normalized value of the propagation delay of the signal from channel i to channel j, measured with channel i as the active transmitter and channel j as the receiver, after calibration. Due to physical symmetry, the delay from i to j should be equal to the delay from j to i. The matrix is ​​usually symmetrical, but noise may occur in actual measurements. After all cross-combined measurements, the fixed delay deviation of each channel relative to a certain reference channel is extracted from the matrix using a least squares or consensus algorithm. The matrix is ​​the bridge connecting the original measurement data and the final calibration value; without it, the independent deviations of each channel cannot be separated. The inter-channel trigger time deviation calibration value is a vector [ΔT1, ΔT2, …, ΔTN] calculated from the relative delay deviation matrix, where ΔTk represents the amount of time that the trigger signal of the kth channel should be advanced or delayed relative to the reference channel (e.g., channel 1). For example, if the link delay of channel 2 is 3 ns greater than that of channel 1, then ΔT2 = -3 ns, which means that a trigger signal needs to be sent to channel 2 3 ns in advance; the calibration value is the result of the combination of static factory calibration and dynamic online measurement, and will be updated periodically by step S300 as the environment changes; the feature is the target quantity directly affected by the compensation algorithm.The proportional-integral (PI) controller is a closed-loop control algorithm whose output equals the sum of the proportional term, the current error, the integral term, and the accumulated error. Adaptability is reflected in the controller's ability to automatically adjust the proportional coefficient Kp and the integral coefficient Ki based on the speed and magnitude of the delay drift. In specific implementation, the main control unit monitors the rate of change of the deviation calibration value given in step S200 in real time. When the drift accelerates, Kp is increased to improve the response speed; when the drift slows down, Ki is decreased to reduce overshoot. The controller's function is to smoothly convert the original deviation calibration value into the adjustment amount of the delay line, avoiding frequent jittering of the delay line due to measurement noise. The exponentially weighted moving average filter is a recursive digital filter whose output... Where x(t) is the deviation calibration value of the current measurement, and α is the weighting coefficient (usually taken as 0.1~0.3); unlike ordinary average filtering, exponentially weighted moving average gives higher weight to the most recent measurement value, and can track the slow drift caused by environmental factors such as temperature more quickly; the filter preprocesses the deviation calibration value sequence generated in step S200, and obtains the delay drift trend after filtering out random jitter; the feature ensures that the subsequent proportional-integral controller will not make incorrect adjustments due to single measurement anomalies, while also not responding too slowly to the real drift. The delay drift trend refers to the overall direction and rate of the slow change of the propagation delay of each channel over time caused by factors such as temperature changes, device aging, and power supply voltage fluctuations; for example, the delay of optical fiber increases by 50 ps for every 10°C increase in temperature, and the delay of the driver chip increases by 100 ps after 1000 hours of aging; by fitting the results of multiple consecutive measurements through the exponentially weighted moving average filter, a drift curve or a slope value is obtained. This trend value is fed forward into the adaptive proportional-integral controller, enabling compensation to act in advance, rather than waiting until the deviation is already large before correcting it. This is a significant difference from traditional lookup-table temperature compensation – it does not rely on a preset model but is dynamically extracted entirely from actual measurement data. Programmable delay lines typically consist of a series of cascaded micro-delay units, each providing a fixed delay step, such as 15 ps, and a digital selector. The phase step control parameter is a binary number, such as 0101101, which tells the delay line how many delay units to pass through from input to output. Each channel is independently configured with a phase step control parameter. The parameter is calculated in real-time by the adaptive proportional-integral controller based on the deviation calibration value and drift trend, with a resolution of 15 ps or higher. The parameter directly determines the actual delay of the trigger signal on that channel and is the means of implementing synchronous compensation. The timing of trigger signal application refers to the actual point in time when the main control unit outputs the trigger pulse to the light source driver module of each channel, usually with the global clock of the main control unit as a reference. Without compensation, all channels emit trigger signals at the same time, but due to different link delays, the actual emission time of the light source varies. By adjusting the phase step control parameters of each channel in step S300, it is equivalent to independently offsetting the emission time of the trigger signal for each channel: for channels with large link delays, the trigger signal is emitted earlier; for channels with small link delays, the trigger signal is emitted later or not adjusted. The ultimate goal is to synchronize the leading edge of the light pulse of all channels at the same global time point, with an error less than the resolution of the programmable delay line. The characteristic is the final output performance of self-calibration and dynamic compensation, which is also a direct indicator of the success of the method.

[0016] In the above embodiments, the principle is referenced in the appendix. Figure 2The design of this embodiment, where the main control unit emits sub-nanosecond measurement short pulses and couples them to the transmission path via a bidirectional transceiver, accurately captures the signal propagation characteristics in the transmission medium. After the measurement short pulse triggers the fixed-delay reflection loop of the drive module, an echo pulse is generated. By performing a differential operation between the round-trip delay measurement and the known fixed delay, the interference of the reflection loop's own delay on the measurement can be eliminated, thus accurately extracting the single-path propagation delay parameter of the transmission path. The cross-validation pulse mechanism establishes a fully connected measurement network through active transmission and passive reception modes between channels. The accurate measurement of the short pulse arrival time at the receiving end, combined with the single-path propagation delay reference value, quantifies the relative time deviation between channels. The generation of the relative delay deviation matrix achieves a systematic characterization of the trigger deviation between channels, providing data support for compensation. The combined application of an adaptive proportional-integral controller and an exponentially weighted moving average filter constitutes a dual-closed-loop control system. The former performs proportional-integral operations based on the current deviation value to achieve a fast response, while the latter predicts the drift trend through weighted processing of historical delay data. The output parameters of both adjust the phase step of the programmable delay line in real time, ensuring that the synchronization accuracy of the trigger pulse's leading edge is unaffected by delay fluctuations caused by environmental factors. This embodiment maintains the synchronization performance of the full-channel trigger signals at the sub-nanosecond level through a dynamic compensation mechanism.

[0017] Example 2: As Figure 3 As shown, based on Embodiment 1, the process of the main control unit measuring the total round-trip delay between transmission and reception in step S100 of this embodiment of the invention specifically includes the following steps: Step S101: The main control unit transmits a short measurement pulse, simultaneously triggering the coarse counter to start counting and latching the current state of the tap delay line as the initial fine time vector; the initial value of the coarse counter and the initial fine time vector are merged into the initial timestamp; Step S102: The main control unit receives the leading edge of the echo pulse, latches the current value of the coarse counter, and latches the tap position where the tap delay line is triggered as the end fine time vector; the current value of the coarse counter and the end fine time vector are merged into the end timestamp; Step S103: The main control unit subtracts the start timestamp from the end timestamp to obtain the coarse count difference, and then converts the difference between the start fine time vector and the end fine time vector into fine delay; the coarse count difference plus the fine delay gives the total round-trip delay value.

[0018] In this embodiment, the basic parameters are defined as follows: Indicates the coarse counter clock cycle, in seconds or picoseconds; indicates the delay step of a single tap, in seconds or picoseconds. Represents the total number of taps, satisfying , which is a dimensionless integer; Step S101: Start timestamp calculation formula: Start timestamp equal to the initial value of the coarse counter Multiply by clock cycle plus the starting tap number Multiply by tap step ;Right now ;in It is a non-negative integer. The value is obtained from the initial fine time vector through a leading one detection, and its range is from 0 to... Integers; Step S102 End Timestamp Calculation Formula: End Timestamp Equal to the current value of the coarse counter Multiply by clock cycle Add the end tap number Multiply by tap step ;Right now ;in It is a non-negative integer. The value is obtained from the end time vector through a leading one detection, and its range is from 0 to... Integers; Step S103: Total Round-Trip Delay Calculation Formula: Total Round-Trip Delay It equals the end timestamp minus the start timestamp, i.e. After unfolding, you get ;because It may be negative; the correct calculation formula is: The modulo operation ensures that the fine part is always non-negative.

[0019] The coarse counter is a binary counter running inside the main control unit FPGA. Its clock frequency is typically 200MHz to 500MHz, corresponding to a counting period of 5ns to 2ns. The coarse counter is used to record the integer number of clock cycles between the transmitted measurement short pulse and the received echo pulse. Due to the limited resolution of the coarse counter (nanosecond level), it cannot meet the picosecond-level accuracy required for multi-channel light source synchronization compensation. Its function is only to provide a measurement range, covering the longest link round-trip time and the integer cycle count, while the remaining fine delay (less than one clock cycle) is completed by the tapped delay line. The coarse counter is relevant to this embodiment in that it forms a coarse-fine combined time measurement architecture with the tapped delay line, solving the problem of insufficient accuracy of a single clock count. This allows the main control unit to directly measure the total round-trip delay of each channel online without the need for external high-precision time measurement instruments, thereby obtaining the single-path propagation delay. A latched tap delay line is a fine-grained time measurement circuit based on the propagation delay of a signal in a gate chain. It consists of a series of cascaded buffers or inverters, with the output of each buffer called a tap. The total delay of the tap delay line typically covers a coarse counter clock cycle, such as 5 ns. The delay difference between each tap is fixed and known, such as 15 ps. In this embodiment, the latching operation means that when the short pulse transmission time or the echo pulse arrival time arrives, the main control unit simultaneously samples the level state of all taps on the tap delay line, freezing the position where the signal has propagated in the delay line. The latching result is a binary vector, and the high-level tap position of the vector indicates the fine position of the measurement event occurrence time within one clock cycle. The relevance of the latched tap delay line to this invention is that it enables the main control unit to record the occurrence time of transmission and reception events with picosecond-level resolution without increasing the system main clock frequency, thereby achieving high-precision round-trip delay measurement on a low-cost FPGA and directly supporting the calculation of the single-path propagation delay of each channel. The initial fine time vector refers to a set of binary state vectors obtained from the latched tap delay line at the moment of transmitting the short pulse measurement, such as 32-bit or 64-bit. The number of consecutive high-level taps or the position of the most significant tap in the vector corresponds to the fine time offset of the transmission moment within one coarse counter clock cycle. The resolution is equal to the delay of a single delay unit of the tap delay line, such as 15 ps. The initial fine time vector and the initial value of the coarse counter together constitute the initial timestamp, which has the physical meaning of: the absolute time of the transmission event = the initial value of the coarse counter × the clock cycle + the fine delay corresponding to the initial fine time vector. The initial fine time vector compensates for the deficiency of the coarse counter in distinguishing moments within a clock cycle, improving the positioning accuracy of the transmission moment from the nanosecond level to the picosecond level, thereby ensuring that the round-trip total delay measurement value can truly reflect the small delay differences of the trigger link.The start timestamp is a composite time stamp formed by merging the initial value of the coarse counter and the start fine time vector. The merging method involves shifting the initial value of the coarse counter to the left by a certain number of bits. For example, if the fine time vector has 64 levels, corresponding to 6 bits of binary precision, then the initial value of the coarse counter is shifted to the left by 6 bits, and then added to or concatenated with the fine delay code value corresponding to the start fine time vector. The numerical value of the start timestamp precisely represents the moment when the measured short pulse leaves the main control unit, and its resolution is equal to a single delay step of the tap delay line ≤ 15ps. The start timestamp provides a high-precision starting reference for round-trip measurements; the difference between the end timestamp and the start timestamp can be used to obtain the complete round-trip total delay without additional assumptions or external references. Without a start timestamp, the error in the transmission time would be directly carried into the one-way propagation delay, causing inaccuracies in the compensation parameters of each channel. The leading edge of the echo pulse refers to the edge where the echo pulse generated by the reflection circuit transitions from a low level to a high level, or from a high level to a low level. This edge is a time marker that the main control unit needs to accurately capture. In step S100, after the echo pulse returns to the main control unit along the original transmission path, the high-speed comparator or trigger of the main control unit detects this leading edge and immediately triggers a latching action. The detection accuracy of the leading edge of the echo pulse directly affects the accuracy of the end timestamp. The leading edge of the echo pulse is the termination marker of the round-trip measurement. The time interval between its capture time and the emission time of the measurement short pulse includes the bidirectional propagation delay of the transmission path and the known fixed delay of the reflection circuit. Only by accurately locking the leading edge of the echo pulse can a reliable total round-trip delay value be obtained, and then the one-way propagation delay can be calculated. The end timestamp is a composite time stamp formed by combining the current value of the coarse counter, or the value read at the arrival time of the echo pulse, with the end fine time vector. The merging method is the same as the start timestamp. The end timestamp accurately represents the moment when the leading edge of the echo pulse arrives at the main control unit, with a resolution of ≤15ps for a single step of the tap delay line. The end timestamp provides the termination reference for round-trip measurement, and the difference between it and the start timestamp directly gives the total round-trip delay. During dynamic compensation, the main control unit will repeat step S100 multiple times, generating a new end timestamp each time, thereby tracking the changes in link delay in real time. Without the end timestamp, it is impossible to quantify the precise arrival time of the echo, and the total round-trip delay measurement cannot be completed. Fine delay refers to the delay component that is less than one coarse counter clock cycle, which is calculated from the difference between the start fine time vector and the end fine time vector. The specific conversion method is: subtract the fine time value corresponding to the start fine time vector from the fine time value corresponding to the end fine time vector; if the result is negative, add one clock cycle.The resolution of fine delay is equal to the step value of a single delay unit of the tap delay line, such as 15 ps. The fine delay and the difference between the coarse and coarse counter values ​​must be added together to form the complete round-trip total delay value. If only the difference between the coarse and coarse counter values ​​is relied upon, the measurement error can reach up to one clock cycle (several nanoseconds) over an integer number of clock cycles, which is completely unacceptable for multi-channel light source systems that require picosecond-level synchronization. Fine delay improves the measurement accuracy from the nanosecond level to the picosecond level, enabling the main control unit to detect tiny delay drifts caused by factors such as temperature changes and device aging, thereby driving the adaptive proportional-integral controller and the exponentially weighted moving average filter in step S300 to perform accurate compensation.

[0020] In the above embodiments, the main control unit achieves high-precision time delay measurement by measuring the time difference between short pulse transmission and echo reception; it triggers a coarse counter to start counting and latches the state of the tap delay line as the initial fine time vector, ensuring accurate capture of the initial time reference; the initial value of the coarse counter and the initial fine time vector are merged into a starting timestamp, achieving seamless integration of coarse and fine time. When the leading edge of the echo pulse is triggered, the current value of the coarse counter is latched and the triggered position of the tap delay line is recorded as the ending fine time vector, ensuring accurate marking of the echo arrival time; the ending timestamp is generated by merging the current value of the coarse counter and the ending fine time vector, completing the complete recording of the receiving end time reference. The calculation process obtains the coarse count value difference by the difference between the ending timestamp and the starting timestamp, and at the same time converts the difference between the starting and ending fine time vectors into a fine time delay, realizing independent calculation of coarse and fine time; finally, the coarse count value difference and the fine time delay are added together to obtain the total round-trip time delay value, significantly improving the measurement accuracy through the combination of coarse and fine time; effectively solving the problem of insufficient accuracy of traditional single clock counting, and realizing nanosecond-level time measurement capability.

[0021] Example 3: Based on Example 2, the process of converting the difference between the initial fine time vector and the ending fine time vector into fine time delay in step S103 of this embodiment of the invention specifically includes the following steps: Step S1031: The main control unit executes a leading one detection instruction on the starting fine time vector, scans from the highest bit to the lowest bit of the vector, finds the position of the first 1, and records the position number as the starting tap number; Step S1032: The main control unit executes the same preamble detection instruction on the end fine time vector, scans from the highest bit to the lowest bit of the vector, finds the position of the first 1, and records the position number as the end tap number; Step S1033: The main control unit concatenates the start tap number and the end tap number into a composite index, uses the composite index to access the pre-built two-dimensional fine-delay data table, and stores the corresponding fine-delay value in advance at each index position in the data table, directly retrieving the fine-delay value as the fine-delay conversion result.

[0022] In this embodiment, the leading detection mapping function for steps S1031 and S1032 is: starting tap number. This is equivalent to the result of performing a leading-one detection function on the initial fine-time vector, i.e. End of tap sequence number This is equivalent to the result of performing a leading-one detection function on the final fine-time vector, i.e. ;in The function returns the index of the first 1 when scanning from the most significant bit to the least significant bit of a binary vector; Step S1033: Equivalent formula for table lookup: In the pre-constructed two-dimensional fine-grained time delay data table, the index position... Fine latency of storage satisfy Therefore, the fine-grained delay value can be directly obtained by looking up the table, i.e. ; Complete round-trip total delay formula: Total round-trip delay The coarse count difference provides the time for an integer number of clock cycles, while the lookup result provides the fine delay for less than one clock cycle. The two are added together to obtain the complete round-trip total delay with picosecond precision.

[0023] In the above embodiments, this embodiment achieves efficient positioning through a leading detection instruction: the most significant bit of the start and end fine time vectors is quickly located, and hardware instructions accelerate the position detection operation on the critical path, avoiding the timing overhead of traditional bit-by-bit comparisons. A combined index mechanism optimizes latency lookup: the start and end tap numbers are concatenated into a two-dimensional index key, replacing multiple calculations with a single table lookup operation, significantly reducing the computational complexity of fine latency conversion while avoiding the precision loss introduced by floating-point operations. A pre-built data table improves real-time performance: a two-dimensional lookup table structure that pre-stores calibration data is adopted, using a space-for-time strategy to ensure deterministic output of latency conversion results, meeting the low latency and high reliability requirements of high-precision time measurement systems.

[0024] In summary, this embodiment forms a processing pipeline: the start detection, end detection, and table lookup operation form a three-level pipeline architecture, which supports parallel processing of multiple time vector difference conversion tasks and is suitable for throughput optimization needs in batch timestamp processing scenarios.

[0025] Example 4: Figure 4 As shown, based on Example 1, the process of generating the relative delay deviation matrix after all cross-combination measurements in step S200 of this embodiment of the invention specifically includes the following steps: Step S201: Each receiving channel subtracts its own one-way propagation delay from the arrival time of the cross-check pulse to obtain the one-way relative delay value of the receiving end relative to the active transmitting end; arrange the one-way relative delay values ​​of all active transmitting ends and all receiving ends according to the transmitting end number as the row and the receiving end number as the column to generate an N-row N-column relative delay deviation matrix, where the element in the i-th row and j-th column of the matrix represents the measured relative delay from channel i to channel j; Step S202: Subtract the diagonal elements from each row of the relative delay deviation matrix to obtain the zeroed deviation matrix; then take the average of each column of the zeroed deviation matrix to obtain the preliminary deviation vector of each channel relative to channel 1. Step S203: Substitute the initial deviation vector into the consistency equation, and solve for the solution that minimizes the sum of squared mutual measurement residuals between channels by using the iterative least squares method. Output the solution as the trigger time deviation calibration value of each channel relative to the reference channel.

[0026] In the above embodiments, this embodiment constructs a complete relative delay relationship network using measured data: by subtracting the one-way propagation delay from the measured arrival times of each receiving channel, the interference of fixed link delay components on the relative deviation measurement is eliminated, forming a relative delay deviation matrix that includes all transmit and receive channel combinations, ensuring the comprehensiveness and symmetry of the deviation observation values. It achieves benchmark unification and coarse calibration: row zeroing eliminates the systematic influence of each transmitter's own trigger time offset on the matrix data; subsequent column averaging extracts the preliminary deviation estimate between channels relative to the benchmark channel 1, providing initial convergence conditions for subsequent optimization algorithms and reducing iterative computational complexity. It completes high-precision deviation calibration: based on the consistency equation, a multi-channel mutual measurement constraint relationship is established; iterative least squares method is used to optimize the sum of squared residuals of each channel, suppressing the cumulative effect of measurement noise and random errors in the deviation propagation process, and outputting channel trigger time calibration values ​​that conform to the global optimal solution, meeting the synchronization accuracy requirements of the multi-channel system. This embodiment achieves end-to-end processing from raw measurement data to high-confidence calibration values, forming a multi-channel delay calibration scheme with self-consistency and fault tolerance.

[0027] Example 5: Based on Example 4, the process of finding the solution that minimizes the sum of squared mutual measurement residuals between channels in step S203 of this embodiment of the invention using the iterative least squares method specifically includes the following steps: Step S2031: The main control unit sets the trigger time deviation of each channel as a one-dimensional unknown. Based on the linear relationship that each measured relative delay value in the relative delay deviation matrix is ​​equal to the sending end deviation minus the receiving end deviation, an overdetermined linear equation system covering all channels for the measured values ​​is established as the consistency observation equation. Step S2032: The main control unit substitutes the current estimated vector into the consistent observation equation, calculates the difference between the left and right sides of each equation to obtain the residual vector; then multiplies the residual vector with the transpose of the observation matrix and takes the negative to obtain the right-hand side term of the normal equation; at the same time, multiplies the transpose of the observation matrix with itself to obtain the coefficient matrix of the normal equation, thereby constructing a symmetric positive definite normal equation. Step S2033: The main control unit applies the Choliski decomposition to the normal equation to obtain the correction vector, adds the correction vector to the current estimated vector to generate the updated estimated vector, and repeats steps S2032 and S2033 until the change in the L2 norm of the residual vector is less than the preset threshold, and outputs the final estimated vector as the deviation calibration value at the trigger time.

[0028] Among them, step S2031 is the system of linear equations:

[0029] Matrix form (consistent observation equation):

[0030] Character meaning: Indicates the first Channel trigger timing deviation; Indicates the first Channel trigger timing deviation; This represents the measured relative delay value obtained from the relative delay deviation matrix, i.e., the channel. Send to channel The measured value; This represents the observation matrix, with each row corresponding to one... Yes, the bank in the Listed as In the Listed as The rest are ; Indicates by all The column vector formed; Indicates by all The column vectors are composed of these equations. By utilizing the linear relationship that each measured relative delay equals the difference between the transmitter deviation and the receiver deviation, an overdetermined linear equation system covering all channels for the measured values ​​is established. The equation system transforms the relative relationships between multiple channels into unified mathematical constraints, providing a foundation for least squares optimization. Step S2032: Residual vector representation of formula:

[0031] Normal equation representation:

[0032] Character meaning representation Indicates the first The current estimated vector in the next iteration; This represents the residual vector, indicating the deviation between the current estimate and the measured value for each equation; Represents the observation matrix Transpose of; This represents the correction vector to be solved, used to update the current estimate; The coefficient matrix of the normal equation is symmetric positive definite. The right-hand side of the normal equation is represented by the following term: Substituting the current estimate into the consistent observation equation yields the residual vector. Based on the least squares criterion, the normal equation is constructed by minimizing the sum of squared residuals. The normal equation transforms the overdetermined problem into a positive definite linear system. By solving for the correction vector and iteratively updating it, the optimal solution that minimizes the global sum of squared residuals is gradually approximated, thus obtaining a high-precision trigger time deviation calibration value.

[0033] In the above embodiments, this embodiment achieves high-precision calibration of trigger time deviations between multiple channels by systematically constructing an overdetermined system of equations and an iterative optimization algorithm. The technical feature of establishing an overdetermined linear system of equations as the consistency observation equations ensures that the global optimal solution is searched using measured relative delay data from all channel pairs, effectively suppressing the impact of single-measurement noise through redundant measurement data. Modeling based on the linear relationship of the relative delay deviation matrix transforms the nonlinear time synchronization problem into a linear estimation problem, reducing computational complexity. The combined effect of the residual vector and the normal equation construction process achieves standardized processing of the least squares problem. Generating a symmetric positive definite normal equation coefficient matrix through the transpose and multiplication of the observation matrix ensures the existence and uniqueness of the solution, providing a mathematical foundation for numerical stability. The negative gradient characteristic of the right-hand side of the normal equations ensures that the iteration direction always converges in the direction of decreasing residual sum of squares. Applying the technical feature of Choliski decomposition to solve the normal equations fully utilizes the numerical characteristics of the symmetric positive definite matrix, achieving higher computational efficiency and numerical accuracy compared to direct inversion. By setting the L2 norm change of the residual vector as a convergence criterion, quantitative control of the algorithm's termination condition is achieved, avoiding invalid iterations while ensuring calibration accuracy. The final output estimated vector is used as the deviation calibration value at the trigger time. Through iterative optimization, the mutual measurement residual between channels is minimized globally, thereby eliminating the cumulative effect of systematic and random errors during the measurement process and realizing unified calibration of the cross-channel time reference. While ensuring measurement accuracy, it also has good engineering feasibility.

[0034] Example 6: Based on Example 5, the process of obtaining the correction vector by applying Chorisky decomposition to the normal equation in step S2033 of this embodiment of the invention specifically includes the following steps: Step S20331: The main control unit performs Choliski decomposition on the coefficient matrix of the normal equation. The decomposition result is the product of the lower triangular matrix and the transpose of the lower triangular matrix, which yields the lower triangular matrix. Step S20332: The main control unit substitutes the lower triangular matrix and the right-hand side of the normal equation into the forward substitution process, and solves each component of the intermediate vector sequentially from the first equation to obtain the complete intermediate vector; Step S20333: The main control unit substitutes the transpose of the lower triangular matrix and the intermediate vector into the backward substitution process, and solves each component of the correction vector sequentially from the last equation to obtain the complete correction vector.

[0035] In the above embodiments, this embodiment decomposes the coefficient matrix of the normal equation into the product of a lower triangular matrix and its transpose. Utilizing the properties of positive definite matrices, it ensures numerical stability and uniqueness in the decomposition process, avoiding ill-conditioned problems that may occur in general matrix decomposition. The decomposition process transforms the original normal equation into a cascaded solution problem of triangular matrices, laying the mathematical foundation for solving linear systems. A forward substitution process is used to process the lower triangular matrix and the right-hand side of the normal equation. Its technical feature is that intermediate vector components are calculated sequentially and recursively, with each component depending only on the previously solved results, ensuring the serializability of the calculation process. The decomposed triangular matrix is ​​transformed into a linear mapping of intermediate results, achieving a lossless conversion from the right-hand side of the normal equation to the intermediate vector. This embodiment uses a backward substitution process combined with the transpose of the lower triangular matrix and the intermediate vector to solve the correction vector components in reverse recursion. It fully utilizes the sparsity of the upper triangular matrix, with each component calculation involving only a single row of matrix operations, reducing computational complexity while ensuring the numerical accuracy of the correction vector. The synergistic effect of forward and backward substitution constitutes a complete solution chain for the linear system, and the final output correction vector satisfies the optimal solution requirements of the normal equation. In summary, this embodiment combines decomposition and replacement calculations to reduce memory usage and computational load while maintaining computational stability, providing high-precision directional correction for iterative optimization, and ultimately ensuring the convergence and reliability of the deviation calibration results at the trigger moment.

[0036] Example 7: Figure 5 As shown, based on Embodiment 1, the process of the adaptive proportional-integral controller outputting the phase step control parameters of each channel programmable delay line in step S300 of the present invention specifically includes the following steps: Step S301: Exponential weighted moving average filter, recursively filters the deviation calibration values ​​of each channel's trigger time in continuous measurement, and outputs the current delay drift trend value of each channel as the feedforward compensation amount; Step S302: The adaptive proportional-integral controller takes the difference between the current deviation calibration value and the delay drift trend value as the error signal. The proportional path multiplies the error signal with the adaptive proportional coefficient to obtain the proportional component. The integral path accumulates the error signal and multiplies it with the adaptive integral coefficient to obtain the integral component. The proportional component and the integral component are superimposed to obtain the original control quantity. Step S303: The adaptive proportional-integral controller divides the original control quantity by the time resolution corresponding to a single step of the programmable delay line and rounds it to the nearest integer to obtain the number of phase steps that each channel needs to add. The number of phase steps is then converted into binary numbers and output as phase step control parameters to the programmable delay lines of each channel.

[0037] In the above embodiments, the delay drift trend value provided in this embodiment directly applies to the time-domain characteristic compensation of the trigger link; the exponential weighted filtering technology extracts the trend of the trigger time deviation data output by the calibration system, providing a feedforward reference for compensation control, effectively overcoming the inherent delay changes of the trigger link caused by light source aging and temperature drift. The designed adaptive proportional-integral control architecture performs closed-loop adjustment for the trigger delay differences of the multi-channel light source system; the proportional path quickly responds to the inter-channel trigger deviation detected by the calibration system, and the integral path gradually eliminates the steady-state error of the system; ensuring that the synchronization accuracy of the inter-channel trigger time can still be maintained under the conditions of light source intensity fluctuation or circuit parameter drift. The implemented phase step parameter conversion maps the control algorithm output to the physical adjustment amount of the programmable delay line; through quantization processing, the time-domain compensation amount is converted into a step instruction executable by the digital circuit, which directly applies to the delay unit of each channel trigger link, completing the full closed-loop control from calibration data to hardware compensation; ensuring the accurate matching between the compensation parameters output by the self-calibration system and the physical implementation of the trigger link.

[0038] In summary, this embodiment constructs a complete technical link from delay calibration to dynamic compensation. Through the synergistic effect of feedforward prediction and feedback control, it realizes automatic calibration and stability maintenance of the trigger moment of the multi-channel light source system during long-term operation; together, it ensures the real-time performance and accuracy of the trigger link delay self-calibration system.

[0039] Example 8: Based on Example 7, the process of obtaining the original control quantity by superimposing the proportional component and the integral component in step S302 of this embodiment of the invention specifically includes the following steps: Step S3021: The adaptive proportional-integral controller subtracts the current channel trigger time deviation calibration value from the channel delay drift trend value output by the exponentially weighted moving average filter to obtain the channel's net error signal; the net error signal is multiplied by the current adaptive proportional coefficient of the channel to obtain the channel's proportional compensation component. Step S3022: The adaptive proportional-integral controller adds the historical cumulative error value of the channel to the current net error signal to obtain the updated cumulative error value; the updated cumulative error value is multiplied by the current adaptive integral coefficient of the channel to obtain the integral compensation component of the channel. Step S3023: The adaptive proportional-integral controller adds the proportional compensation component of the channel to the integral compensation component to generate the original control quantity of the channel; after the original control quantities of all channels are calculated in parallel, they are uniformly quantized and mapped.

[0040] The net error signal refers to the difference between the current channel's trigger time deviation calibration value (output in real time by step S200) and the channel delay drift trend value output by the exponentially weighted moving average filter. The delay drift trend value represents the long-term slow change component caused by factors such as temperature and aging, while the net error signal eliminates the trend and retains the instantaneous random fluctuations and model prediction errors. The signal is directly input to the proportional and integral paths of the proportional-integral controller as the basis for compensation adjustment. The net error signal enables the controller to distinguish between predictable drift and unpredictable instantaneous deviation, thereby avoiding repeated adjustment of the pre-compensated trend and improving compensation accuracy and stability. The adaptive proportional gain is a gain factor that dynamically adjusts according to the rate of change of the current delay drift trend of the channel, typically ranging from 0.1 to 2.0. The main control unit monitors the amplitude change rate of the net error signal in real time: when the rate of change exceeds a preset threshold, indicating rapid drift in the link, the adaptive proportional gain increases, enhancing the proportional compensation component and accelerating the response speed; when the rate of change is small, the gain decreases to prevent overshoot. The gain is closely related to the self-calibration of the multi-channel light source-triggered link delay: different channels have different link lengths, media types, and drive circuit characteristics, resulting in varying delay drift speeds and amplitudes. The adaptive proportional gain is configured independently for each channel, ensuring that the compensation action of each channel matches its own dynamic characteristics, achieving personalized correction. The proportional compensation component equals the net error signal multiplied by the adaptive proportional gain, reflecting the controller's immediate response strength to the current instantaneous deviation. In the light source triggering link, sudden interference, such as power supply ripple or electromagnetic pulse, causes a short-term jump in the net error signal. The proportional compensation component immediately generates an adjustment proportional to the jump amplitude to quickly offset the triggering time offset caused by the interference. This component is responsible for suppressing high-frequency random jitter and sudden disturbances, ensuring that the synchronization of multi-channel light sources on the microsecond or even nanosecond time scale is not disrupted by transient events. The historical error accumulation value is the algebraic sum of all net error signals of the channels up to the previous control cycle, stored in the non-volatile register of the main control unit. It represents the total amount of long-term accumulated deviation since the system started or the last reset. In the light source triggering link, due to factors such as device aging and slow fiber creep, a single compensation may not be able to completely eliminate steady-state errors, and these residual deviations are continuously integrated. The historical error accumulation value provides a long-term memory function, enabling the controller to eliminate the accumulated synchronization error caused by chronic drift, ensuring that multi-channel light sources can maintain sub-nanosecond triggering consistency even after running continuously for hours or even days.The cumulative error value is the updated result obtained by adding the historical cumulative error value to the current net error signal, i.e., the new sum. In each control cycle, the controller first reads the historical cumulative error value saved in the previous cycle, adds the current net error signal to it to form the cumulative error value, and then writes it into the register to replace the original historical value. It is directly used to calculate the integral compensation component. This feature realizes the real-time integral update of the error, enabling the controller to track the slowly changing offset trend and avoid compensation lag caused by outdated historical data. It is the core link of the continuous operation of the dynamic self-calibration closed loop of the multi-channel light source. The adaptive integral coefficient is a gain factor that dynamically adjusts with the absolute value of the accumulated error of a channel, typically ranging from 0.01 to 0.5. When the absolute value of the accumulated error is large, indicating significant steady-state error, the adaptive integral coefficient increases to accelerate the elimination of accumulated deviation. When the accumulated error approaches zero, the coefficient decreases to prevent integral saturation. Compared to a fixed integral coefficient, the adaptive integral coefficient effectively suppresses overshoot and oscillation. In multi-channel light source triggering links, the temperature drift coefficients and aging rates of different channels vary greatly. The adaptive integral coefficient is adjusted independently for each channel, ensuring that the integral compensation component neither causes oscillation due to an excessively large coefficient nor results in long-term residual steady-state error due to an excessively small coefficient. The integral compensation component equals the accumulated error multiplied by the adaptive integral coefficient, reflecting the controller's integral response to long-term accumulated deviation. Unlike the proportional compensation component, which is fast but may leave residual steady-state error, the integral compensation component continuously increases until the net error signal is completely eliminated, achieving steady-state error-free regulation. In the light source triggering link, even if the proportional compensation component has reduced the deviation to within a single step of the programmable delay line, the small residual steady-state error will still accumulate over time. The integral compensation component continuously fine-tunes the triggering timing, eventually locking the leading edge of the light pulses of all channels at the same global time point. The component and the proportional compensation component work together to form a proportional-integral dual-mode control architecture, which simultaneously meets the dual requirements of fast response and steady-state accuracy of the multi-channel light source triggering link. This is the key guarantee for achieving sub-nanosecond synchronization accuracy in dynamic compensation.

[0041] In the above embodiments, this embodiment eliminates the interference of the drift trend on the real-time deviation assessment by subtracting the trigger time deviation calibration value from the delay drift trend value, and extracts the net error signal reflecting the true synchronization error. The proportional compensation component generated by multiplying the net error signal by the adaptive proportional coefficient can quickly suppress the instantaneous offset of the trigger time caused by sudden fluctuations in the light source or circuit noise, significantly improving the dynamic response performance of the multi-channel system. An error accumulation mechanism is used to integrate the net error signal in the time domain, and its update process continuously tracks the systematic deviation caused by long-term drift. The integral compensation component generated by multiplying the accumulated value by the adaptive integral coefficient can progressively correct the trigger link reference offset caused by component aging or environmental temperature drift, ensuring the steady-state accuracy of delay compensation for each channel. Through the linear superposition of the proportional component and the integral component, an original control quantity with both fast response and continuous correction capability is formed. The proportional component prioritizes the suppression of high-frequency disturbances, while the integral component dominates the elimination of low-frequency drift. The synergistic effect of the two achieves full-band dynamic compensation of the trigger link delay error. The parallel computing architecture ensures the real-time synchronous update of the control quantities of all channels, providing a unified time base for quantization mapping.

[0042] In summary, this embodiment constitutes the core calculation of the composite control strategy, and its output directly determines the accuracy of the compensation parameters of the programmable delay line; it realizes closed-loop coupling between calibration data and control response in a multi-channel light source triggering system, and ultimately achieves rapid convergence and long-term stability of the triggering time deviation.

[0043] Example 9: Based on Example 8, the process of adding the proportional compensation component and the integral compensation component in step S3023 of this embodiment of the invention specifically includes the following steps: Step S30231: The proportional compensation component and the integral compensation component of the channel are respectively input into their respective nonlinear shaping programs. The proportional compensation component is subjected to exponential decay shaping to obtain the proportionally shaped value, and the integral compensation component is subjected to saturation compression shaping to obtain the integral shaped value. Step S30232: Convert the current delay drift trend value of the channel into a dynamic weight coefficient, multiply the dynamic weight coefficient by the proportional shaping value to obtain the proportional contribution value, subtract the dynamic weight coefficient from the integral shaping value to obtain the integral contribution value, and then superimpose the proportional contribution value and the integral contribution value to obtain the combined control quantity. Step S30233: Perform a weighted moving average of the combined control quantity of the channel and the output control quantity of the channel in the previous cycle to obtain the smoothed control quantity; then multiply the smoothed control quantity by the link health factor of the channel and output the original control quantity.

[0044] Exponential decay shaping involves multiplying the proportional compensation component by a decay factor that decreases exponentially over time. Mathematically, the proportional shaping value equals the proportional compensation component multiplied by the negative time of the natural constant, divided by the time constant raised to the power of the time constant. Here, time is the time interval between the current control cycle and the last significant adjustment, and the time constant is a preset value. In a multi-channel light source triggering link, the proportional compensation component is responsible for rapidly responding to sudden delay transitions, such as nanosecond-level offsets caused by power fluctuations or electromagnetic interference. Applying full compensation for every transition would lead to frequent and significant adjustments to the programmable delay line, introducing additional phase jitter. Exponential decay shaping retains most of the compensation strength in the initial few cycles after a transition to quickly correct the offset, subsequently decreasing the compensation strength exponentially over time to avoid over-adjustment and oscillation. Shaping allows the proportional compensation to maintain a fast response while automatically suppressing repeated impacts from high-frequency noise and transient interference on the triggering link, thus ensuring long-term stability of the multi-channel light source trigger pulse leading edge while achieving sub-nanosecond synchronization accuracy. Saturated compression shaping limits the integral compensation component to a preset upper and lower threshold range, compressing any excess to the threshold value. The specific operation is as follows: if the integral compensation component is greater than the positive saturation threshold, the integral shaping value is equal to the positive saturation threshold; if it is less than the negative saturation threshold, the integral shaping value is equal to the negative saturation threshold; otherwise, it remains unchanged. In the multi-channel light source triggering link, the integral compensation component is obtained by multiplying the historical error accumulation value by the adaptive integral coefficient. As time accumulates, it may become very large, resulting in integral saturation. This causes the controller to continue to output a large adjustment amount even when the link has actually stabilized, leading to overshoot or even oscillation. The function of saturation compression shaping is to force the integral compensation component not to exceed the physical adjustment range of the programmable delay line and to prevent runaway caused by integral saturation. Shaping ensures that the integral path always outputs a control quantity within the legal range, while retaining the fine adjustment capability for small steady-state errors, ensuring the reliability and repeatability of trigger synchronization during long-term operation of multi-channel light sources. The dynamic weighting coefficient is a real number ranging from 0 to 1, calculated in real time from the current channel's delay drift trend value using a preset mapping function. A larger delay drift trend value indicates a rapidly changing link delay, and the dynamic weighting coefficient is closer to 1; a smaller delay drift trend value indicates a more stable link, and the dynamic weighting coefficient is closer to 0. In a multi-channel light source triggering link, the coefficient adaptively allocates the proportions of proportional and integral compensation components in the final control quantity. When the link drifts rapidly due to a sudden temperature rise or accelerated aging, the weight of the proportional component is increased to achieve rapid tracking. When the link is stable, the weight of the proportional component is decreased to reduce noise sensitivity, while the effect of the integral component is relatively enhanced to ensure steady-state accuracy. The proportional contribution value equals the dynamic weighting coefficient multiplied by the proportional shaping value.The proportional shaping value is the result of exponentially decaying shaping of the proportional compensation component; the proportional contribution value represents the effective control contribution from the proportional path in the current control cycle, after shaping and weighting; in a multi-channel light source triggering link, the proportional path is responsible for quickly responding to instantaneous deviations, and the proportional contribution value determines the intensity of the quick response; because the dynamic weighting coefficient adapts to the drift trend, the proportional contribution value dominates when the link fluctuates drastically, ensuring that the triggering timing can promptly correct sudden delay deviations and prevent short-term loss of synchronization in multi-channel light sources. The integral contribution value is equal to the difference obtained by subtracting the dynamic weighting coefficient from the integral shaping value, which is the result of saturated compression shaping of the integral compensation component. The integral contribution value represents the effective control contribution from the integral path in the current control cycle, after shaping and weighting. The integral path is responsible for eliminating long-term accumulated steady-state errors, and the integral contribution value determines the degree of intervention of the integral action. When the link is stable, the difference obtained by subtracting the dynamic weight coefficient is large, the integral contribution value is enhanced, and the triggering timing is finely adjusted to eliminate residual small deviations, achieving long-term synchronous locking of multi-channel light sources. The integral contribution value complements the proportional contribution value, together forming a complete control output. The combined control quantity is the algebraic sum of the proportional contribution value and the integral contribution value, that is, the result of directly adding the two. It combines the fast response of the proportional path and the steady-state error-free characteristics of the integral path, and is a preliminary control signal without smoothing and health factor correction. In the multi-channel light source triggering link, the combined control quantity directly reflects the total adjustment direction and magnitude that should be applied in the current cycle. As intermediate variables, the input weighted moving average and health factor correction links are the links connecting the original proportional-integral calculation and the final output. Weighted moving average is a process of weighted summing of the combined control quantity of the current cycle and the original control quantity output of the previous cycle. The specific weighting method is as follows: the smoothed control quantity is equal to the smoothing coefficient multiplied by the combined control quantity, plus the difference obtained by subtracting the smoothing coefficient and multiplying by the original control quantity of the previous cycle. The smoothing coefficient is a preset constant, usually between 0.3 and 0.7. The operation is essentially a first-order low-pass filter used to suppress the drastic jumps in the control quantity from cycle to cycle. In a multi-channel light source triggering link, if the adjustment of the programmable delay line is too frequent or the amplitude changes abruptly, it will introduce additional phase jitter, which will worsen the synchronization performance. Weighted moving average makes the change of control quantity continuous and smooth, avoids the appearance of spikes or oscillations at the leading edge of the trigger pulse, and maintains the ability to track the real drift.The link health factor is a coefficient ranging from 0 to 1, dynamically evaluated by the main control unit based on indicators such as the historical communication bit error rate, echo pulse loss rate, and cross-check consistency deviation of the channel. The factor of a healthy channel is close to 1, while the factor of a faulty or severely degraded channel is close to 0. The smoothed control quantity is multiplied by the link health factor to output the original control quantity. If the health factor of a channel is lower than a preset threshold, such as 0.2, its original control quantity is significantly attenuated or even set to zero to prevent the abnormal compensation requirement of that channel from interfering with other channels or causing further loss of control. The connection between this feature and the multi-channel light source triggered link delay self-calibration is that it realizes fault isolation and fault-tolerant control, ensuring that when some channels in the system fail, the remaining channels can still maintain normal synchronization, and the faulty channel will not cause a chain reaction due to incorrect compensation parameters. The physical meaning of the difference obtained by subtracting the dynamic weighting coefficient is the weight of the integral path. Since the dynamic weighting coefficient controls the weight of the proportional path, and the sum of the two must be 1 (i.e., the total control energy is normalized), subtracting the dynamic weighting coefficient automatically becomes the weight of the integral path. When the dynamic weighting coefficient is 0.8, the proportional contribution accounts for 80% and the integral contribution accounts for 20%. When the dynamic weighting coefficient is 0.2, the proportional contribution accounts for 20% and the integral contribution accounts for 80%. The core function of the design is to dynamically balance the contradiction between fast response and steady-state accuracy without changing the overall control sensitivity. In a multi-channel light source triggering link, when the delay drift trend is fast, priority is given to ensuring the response speed, i.e., the proportional weight is large; when the drift trend is slow, priority is given to ensuring steady-state error-free operation, i.e., the integral weight is large. This allows the same controller to adapt to the performance requirements of different channels and different operating conditions without manual parameter tuning, significantly improving the robustness and deployment convenience of the self-calibration system.

[0045] In the above embodiments, this embodiment processes the proportional compensation component and the integral compensation component separately through a nonlinear shaping program. The proportional component undergoes exponential decay shaping to suppress high-frequency noise and transient interference, while the integral component undergoes saturation compression shaping to prevent integral saturation and overcompensation. This ensures the stability and anti-disturbance capability of the control signal in a high dynamic range and reduces unnecessary adjustments to the trigger link. A dynamic weight allocation mechanism is adopted to adaptively adjust the contribution ratio of the proportional and integral components based on the current delay drift trend value. When the trend is significant, the proportional weight is increased to improve the response speed; when the trend is flat, the integral weight is enhanced to improve steady-state accuracy. The dynamic optimization strategy achieves efficient adaptive adjustment of trigger delay compensation, taking into account both the need for rapid suppression of disturbances and long-term stability maintenance. By using weighted moving average and historical feedback to smooth the changes in control quantity, the impact of high-frequency chattering on the trigger link is reduced. The introduction of a link health factor enables adaptive weight reduction of faulty channels, avoiding abnormal compensation interference with the overall system stability. The final output original control quantity ensures both the continuous smoothness of delay adjustment in each channel and fault tolerance capability, ensuring the robustness of the self-calibration and dynamic compensation system under complex operating conditions.

[0046] In summary, this embodiment achieves an optimized balance in terms of fast response, steady-state accuracy, anti-interference capability, and fault tolerance in multi-channel light source trigger delay compensation, effectively supporting the reliable operation of high-precision triggering links.

[0047] Example 10: Based on Example 9, the process of converting the current delay drift trend value of the channel into a dynamic weighting coefficient in step S30232 of this embodiment of the invention specifically includes the following steps: Step S302321: The main control unit reads the current delay drift trend value of the channel, compares the current delay drift trend value with the pre-stored high drift threshold value and low drift threshold value respectively. If the high drift threshold value is greater than the low drift threshold value, three drift levels are obtained: fast drift level greater than the high drift threshold value, medium drift level between the high drift threshold value and the low drift threshold value, and slow drift level less than the low drift threshold value. Step S302322: Select the fusion method of the corresponding proportional shaping value and integral shaping value according to the drift level. For fast drift level, select the fusion method in which the proportional shaping value completely occupies the output and the integral shaping value is ignored. For medium drift level, select the fusion method in which the proportional shaping value and the integral shaping value are added in a one-to-one ratio. For slow drift level, select the fusion method in which the integral shaping value completely occupies the output and the proportional shaping value is ignored, so as to obtain the original fusion control quantity. Step S302323: The original fused control quantity is fed into the smoothing filter and weighted and averaged with the output control quantity of the previous cycle to obtain the final combined control quantity. The smoothing coefficient is dynamically determined by the drift level. The fast drift level uses a low smoothing coefficient for fast response, and the slow drift level uses a high smoothing coefficient to suppress noise.

[0048] The high drift threshold and low drift threshold are a pair of preset time rate constants, measured in picoseconds per second or nanoseconds per minute, pre-stored in the non-volatile registers of the main control unit; the high drift threshold is greater than the low drift threshold. The high drift threshold is used to define whether the link delay is in a state of drastic fluctuation, for example, set to 10 picoseconds per second; the low drift threshold is used to define whether the link is in a basically stable state, for example, set to 1 picosecond per second. In a multi-channel light source triggered link, different channels have different delay drift rates due to differences in transmission media and driving circuits. By comparing the current delay drift trend value with these two thresholds, the main control unit can automatically classify the current operating condition of the channel into three levels: fast drift, medium drift, and slow drift, thereby matching the most suitable compensation strategy for each channel. The fusion method, where proportional shaping values ​​completely dominate the output while integral shaping values ​​are ignored, refers to using only proportional shaping values ​​and setting integral shaping values ​​to zero when calculating combined control quantities; that is, the output equals the proportional shaping value, without introducing an integral component. In a multi-channel light source triggering link, when the detected delay drift trend value is greater than the high drift threshold, indicating a rapid drift level, it means the link is experiencing sudden interference, such as power spikes, strong electromagnetic pulses, or sudden temperature rises. At this time, the control quantity needs to respond quickly to immediately correct the deviation; the integral component responds slowly and may even cause overshoot due to accumulated errors. Therefore, the integral shaping value is completely discarded, and only the proportional shaping value provides rapid correction, ensuring that the leading edge of the multi-channel light source trigger pulse can recover to near the synchronization reference in the shortest possible time, preventing short-term loss of synchronization. The fusion method of adding proportional and integral shaped values ​​in a one-to-one ratio means that when calculating the combined control quantity, half of the proportional shaped value and half of the integral shaped value are added together. In a multi-channel light source triggering link, when the delay drift trend value is between the high and low drift thresholds (i.e., at a medium-speed drift level), the link exhibits some fluctuations but not to a severe degree. At this point, both fast response and steady-state error elimination need to be considered. The one-to-one addition fusion method balances the contributions of the proportional and integral paths, moderately suppressing fluctuations while gradually eliminating accumulated deviations. This ensures that the triggering synchronization of multi-channel light sources under normal operating conditions will not lose synchronization due to slow response or oscillate due to excessive integral strength. The fusion method where the integral shaped value completely occupies the output while the proportional shaped value is ignored means that when calculating the combined control quantity, only the integral shaped value is used, and the proportional shaped value is set to zero. In other words, the output equals the integral shaped value, without introducing a proportional component.In a multi-channel light source triggering link, when the delay drift trend value is less than the low drift threshold (i.e., in the slow drift level), the link is basically stable, with only slow temperature creep or aging drift. At this point, the proportional component is prone to unnecessary frequent adjustments due to measurement noise, which worsens synchronization accuracy. Therefore, the proportional shaping value is completely ignored, and only the integral shaping value provides slow and fine adjustment, gradually eliminating small steady-state errors. This ensures that the trigger pulse leading edge of the multi-channel light source is locked at the same global moment for a long time, achieving steady-state high-precision synchronization. The low smoothing coefficient is a coefficient used in the weighted moving average to calculate the smoothed control quantity, typically ranging from 0.1 to 0.3. When the drift level is fast drift, the system uses a low smoothing coefficient. The low smoothing coefficient results in a higher proportion of the combined control quantity in the current cycle in the weighted average, while the output control quantity in the previous cycle has a lower proportion. This allows the output control quantity to quickly follow the changes in control requirements of the current cycle. In a multi-channel light source triggering link, at the fast drift level, the trigger timing needs to be quickly adjusted to catch up with sudden delay changes. The low smoothing coefficient ensures the controller's agile response and avoids compensation lag caused by smoothing filtering. The high smoothing coefficient is a factor used in weighted moving averages to calculate the smoothed control quantity, typically ranging from 0.7 to 0.9. When the drift level is slow, the system uses a high smoothing coefficient. This coefficient results in a lower proportion of the combined control quantity in the current cycle within the weighted average, while the output control quantity from the previous cycle has a higher proportion. This leads to extremely smooth changes in the output control quantity, effectively suppressing jitter caused by random measurement noise and quantization errors. In multi-channel light source triggering links, the link is stable under slow drift levels and requires no significant adjustments. The high smoothing coefficient ensures high repeatability and low noise in trigger timing adjustments, preventing frequent changes in the step size of the programmable delay line due to minor noise, thus maintaining the purity and stability of the leading edge of the multi-channel light source trigger pulse.

[0049] In the above embodiments, the main control unit of this embodiment classifies drift levels by comparing the current delay drift trend value with a preset threshold, which can dynamically distinguish delay fluctuations of different intensities, laying the foundation for subsequent differentiated compensation strategies. The identification of fast drift levels ensures the system responds quickly to sudden high-amplitude delay fluctuations, the classification of medium-speed drift levels provides a balanced handling capability for regular fluctuations, and the determination of slow drift levels provides fine-grained suppression of low-frequency, small drifts. The binding of drift levels and fusion methods enables dynamic adaptation of control algorithm parameters. The design of exclusive output of proportional-shaped values ​​under fast drift levels enhances the system's transient response capability and effectively suppresses short-term, large-amplitude delay fluctuations; the use of equal-proportional fusion for medium-speed drift levels balances dynamic performance and stability; and the characteristic of integral-shaped values ​​dominating the output under slow drift levels eliminates long-term accumulated errors and improves steady-state accuracy. The introduction of a smoothing filter suppresses output jumps through weighted averaging of historical control values. Its dynamic smoothing coefficient mechanism is interlocked with the drift level: the low smoothing coefficient of the fast drift level prioritizes response speed, the high smoothing coefficient of the slow drift level focuses on noise suppression, and the medium drift level achieves a compromise between the two; while ensuring dynamic response capability, it reduces high-frequency jitter introduced by the compensation process.

[0050] In summary, this embodiment achieves accurate tracking and real-time compensation of multi-channel light source triggering links under full-band delay fluctuations through a three-level linkage mechanism of drift level detection, dynamic switching of fusion strategy, and adaptive adjustment of smoothing coefficient, ensuring the synchronization accuracy and stability of triggering timing between multiple channels.

[0051] Example 11: As Figure 6 As shown, based on Embodiments 1-10, the self-calibration and dynamic compensation system for multi-channel light source triggering link delay provided in this embodiment of the invention includes: Transmission path processing module 1 is used by the main control unit to send a short measurement pulse with a width of less than 1 nanosecond. The short measurement pulse is coupled to the same transmission path through a bidirectional transceiver. The transmission path connects the main control unit with the light source driving modules of each channel. After receiving the short measurement pulse, the driving module triggers a reflection loop with a known fixed delay. The reflection loop generates an echo pulse and returns to the main control unit along the original transmission path. The main control unit measures the total round-trip delay between transmission and reception, subtracts the known fixed delay from the total round-trip delay to obtain twice the one-way propagation delay, and then divides by two to obtain the one-way propagation delay of each channel. The matrix verification generation module 2 is used to select one channel as the active transmitter by taking the single-path propagation delay of each channel as a reference; the channel driving module sends cross-verification pulses to adjacent channels or selected channels; the receiving channel measures the arrival time of the short pulse; compares the difference between the receiving time of each channel and its own single-path delay, and generates a relative delay deviation matrix after all cross-combined measurements, and extracts the inter-channel trigger time deviation calibration value from the relative delay deviation matrix. The trigger signal application module 3 is used to input the calibration value of the trigger time deviation between channels into the adaptive proportional-integral controller. Combined with the exponentially weighted moving average filter to predict the delay drift trend, the adaptive proportional-integral controller outputs the phase step control parameters of the programmable delay line of each channel. The phase step control parameters adjust the timing of the trigger signal application of each channel in real time to achieve closed-loop synchronous compensation of the leading edge of the trigger pulse in the entire channel range.

[0052] In the above embodiments, the self-calibration and dynamic compensation system for the multi-channel light source trigger link delay in this embodiment achieves high-precision synchronous control through modular design. The transmission path processing module 1 utilizes sub-nanosecond-level measurement of short pulses and a bidirectional transceiver architecture. By calculating the difference between round-trip delay measurement and the known reflection loop delay, it accurately quantifies the single-path transmission delay of each channel signal, achieving automated calibration of the channel's basic delay parameters. The matrix verification generation module 2 establishes a relative delay relationship model between channels, employing a cross-verification pulse transmission and multi-point reception time comparison mechanism. Through full combination measurement, it generates a relative delay deviation matrix, solving the problem of eliminating residual system errors in single-path propagation delay measurement and extracting accurate deviation calibration values ​​for the trigger times between channels. The trigger signal application module 3 constructs a closed-loop compensation system based on adaptive control theory. It inputs the deviation calibration values ​​into a proportional-integral controller and combines it with an exponentially weighted moving average filtering algorithm to predict and compensate for delay drift trends. Through dynamic adjustment of the phase step parameters of the programmable delay line, it achieves nanosecond-level synchronization accuracy of the leading edge of the trigger pulse for all channels.

[0053] In summary, this embodiment achieves self-calibration and real-time delay compensation of multi-channel light source triggering links through a three-level closed-loop control architecture of time delay measurement, cross-verification, and dynamic compensation; it can effectively eliminate inter-channel triggering jitter caused by factors such as transmission path differences and temperature drift, and meet the technical requirements of high-precision synchronous triggering systems.

[0054] Figure 7 A block diagram of an exemplary electronic device suitable for implementing embodiments of the present invention is shown.

[0055] The electronic device may include a central processing unit / microprocessor / main control chip 4; and a storage medium 5 coupled to the central processing unit / microprocessor / main control chip 4 and storing computer-executable instructions therein for performing the steps of various methods of embodiments of the present invention when executed by the processor.

[0056] The central processing unit / microprocessor / main control chip 4 may include, but is not limited to, one or more processors or microprocessors.

[0057] Storage medium 5 may include, but is not limited to, random access memory (RAM), read-only memory (ROM), flash memory, EPROM memory, EEPROM memory, registers, computer storage media (e.g., hard disk, floppy disk, solid-state drive, removable disk, CD-ROM, DVD-ROM, Blu-ray disc, etc.).

[0058] In addition, the electronic device may include (but is not limited to) a data bus 6, an input / output bus / external bus / device bus 7, a display 8, and input / output devices 9 (e.g., keyboard, mouse, speaker, etc.).

[0059] The central processing unit / microprocessor / main control chip 4 can communicate with external devices (8, 9, etc.) via wired or wireless networks (not shown) through the input / output bus / external bus / device bus 7.

[0060] The storage medium 5 may also store at least one computer-executable instruction for performing the steps of various functions and / or methods in the embodiments described herein when the central processing unit / microprocessor / main control chip 4 is running.

[0061] In one embodiment, the at least one computer-executable instruction may also be compiled into or comprise a software product, wherein one or more computer-executable instructions are executed by a processor to perform the steps of the various functions and / or methods in the embodiments described herein.

[0062] Figure 8 A schematic diagram of a computer-readable storage medium according to an embodiment of the present invention is shown.

[0063] like Figure 8 As shown, the non-transitory computer-readable storage medium 11 stores instructions, such as computer-readable instructions 10. When the computer-readable instructions 10 are executed by a processor, the various methods described above can be performed. The non-transitory computer-readable storage medium includes, but is not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Non-transitory non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. For example, the non-transitory computer-readable storage medium 11 can be connected to a computing device such as a computer, and then, when the computing device executes the computer-readable instructions 10 stored on the non-transitory computer-readable storage medium 11, the various methods described above can be performed.

[0064] In a large-scale high-power laser device, in order to achieve the precise energy superposition required for inertial confinement fusion experiments, the front-end seed light pulse needs to be simultaneously split into dozens of main amplification chains, and the time synchronization accuracy of each laser at the target point needs to be better than 20 picoseconds. The device is equipped with an independent pump source drive module for each amplification chain, which is responsible for triggering the xenon lamp or laser diode pump source at a precise moment.

[0065] However, the following problems were found in actual operation: the transmission paths of trigger signals from the central control unit to each drive module, including optical fibers or coaxial cables, vary in length by tens of meters, resulting in different inherent delays; during device operation, changes in ambient temperature and the thermal effect of high-current pumping cause inconsistent drift in the delay of each transmission path, typically ranging from 1 to 5 picoseconds per degree Celsius; as some drive modules age, the response delay of their internal circuits also changes slowly; these problems cause the actual emission time of the pump pulses in each channel to be asynchronous with the ideal time by several nanoseconds, directly degrading the amplification synchronization of the main laser and affecting the target accuracy.

[0066] To address the aforementioned issues, the device employs a self-calibration and dynamic compensation system for the multi-channel light source trigger link delay. The specific implementation is as follows: First, initial delay calibration. After the device is powered on and warmed up, the main control unit, using an FPGA and a built-in time-to-digital converter, first sends a short measurement pulse with a width of 800 picoseconds. This pulse is coupled to the fiber optic transmission path via a bidirectional transceiver, such as an optical circulator, and transmitted to each channel driver module. The fixed-delay reflection loop within the module, with a calibrated delay of 5 nanoseconds and a deviation controlled within ±10 picoseconds, generates an echo pulse that returns to the main control unit along the original fiber. The main control unit accurately measures the total round-trip delay of the transmission and echo, subtracts the known 5-nanosecond reflection loop delay, and divides by 2 to obtain the real-time one-way propagation delay of each channel; no path symmetry assumption is required, and no additional hardware is needed. Second, cross-validation and deviation matrix generation. Using the obtained single-path propagation delay of each channel as a benchmark, the main control unit sequentially configures each drive module as an active transmitter. Each active transmitter sends a cross-validation pulse with a width of picosecond to the adjacent channel. Each receiver module records the pulse arrival time and compares it with its own single-path delay benchmark. After all N multiplied by N minus one cross-validation measurements are completed, the main control unit summarizes the data to generate a relative delay deviation matrix. This matrix is ​​solved using iterative least squares to obtain the precise trigger time deviation calibration value of each channel relative to the benchmark channel; for example, channel 5 is 3.2 nanoseconds slower than the benchmark channel. Third, closed-loop dynamic compensation: the main control unit sends the above-mentioned deviation calibration value to the adaptive proportional-integral controller; at the same time, an exponentially weighted moving average filter filters the deviation sequence of multiple consecutive measurements and predicts the delay drift trend in real time, such as the linear increase caused by temperature rise; the controller combines the current deviation and the predicted trend to dynamically adjust the proportional and integral coefficients and generate corresponding phase step control parameters for each channel, for example, determining the number of steps to advance or delay in 15 picoseconds; the parameters are written to the programmable delay line in each channel's drive module in real time, and the timing of the trigger pulse application is independently fine-tuned.

[0067] Through continuous and rapid closed-loop operation using the aforementioned method, the system updates compensation parameters every second, ensuring that the pump source triggering fronts of all channels of the high-power laser device are stably synchronized within a deviation range of ±15 picoseconds. Even during continuous target-hitting experiments lasting several hours, with ambient temperature fluctuations exceeding 15 degrees Celsius, the system can automatically suppress trigger delay drift without manual intervention or periodic shutdowns for calibration, significantly improving the device's operational efficiency and target-hitting success rate.

[0068] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between systems or units may be electrical, mechanical, or other forms.

[0069] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0070] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0071] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for executing all or part of the steps of the methods of the various embodiments of this invention through a computer device (which may be a personal computer, server, or network device, etc.). The aforementioned storage medium includes: USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media capable of storing program code.

[0072] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A self-calibration and dynamic compensation method for multi-channel light source triggering link delay, characterized in that, Includes the following steps: The main control unit sends a narrow pulse, which is transmitted to the light source drive module of each channel via a bidirectional transceiver along the transmission path. This triggers a known fixed-delay reflection circuit, and the echo returns along the original path. The total round-trip time delay is measured, and the fixed delay is subtracted to obtain twice the one-way time delay. Then, half of this result is taken to obtain the one-way propagation delay of each channel. Based on the single-way delay, each channel polls and sends cross-check pulses. After all the combined measurements, a relative delay deviation matrix is ​​generated, and the calibration value of the trigger time deviation between channels is extracted. The calibration value of the inter-channel trigger timing deviation is input into the adaptive proportional-integral controller. Combined with the exponentially weighted moving average filter, the delay drift trend is predicted. The adaptive proportional-integral controller outputs the phase step control parameters of the programmable delay line of each channel. The phase step control parameters adjust the timing of the trigger signal application of each channel in real time, so as to achieve closed-loop synchronous compensation of the trigger pulse leading edge in the entire channel range.

2. The self-calibration and dynamic compensation method for multi-channel light source triggering link delay as described in claim 1, characterized in that, The main control unit emits a short measurement pulse with a width of less than 1 nanosecond. The short measurement pulse is coupled to the same transmission path via a bidirectional transceiver. The transmission path connects the main control unit to the light source driving modules of each channel. After receiving the short measurement pulse, the driving module triggers a reflection loop with a known fixed delay. The reflection loop generates an echo pulse and returns to the main control unit along the original transmission path. The main control unit measures the total round-trip delay between transmission and reception, subtracts the known fixed delay from the total round-trip delay to obtain twice the one-way propagation delay, and then divides by two to obtain the one-way propagation delay of each channel.

3. The self-calibration and dynamic compensation method for multi-channel light source triggering link delay as described in claim 1, characterized in that, Using the single-path propagation delay of each channel as a benchmark, one channel is selected sequentially as the active transmitter; the channel driving module sends cross-check pulses to adjacent channels or selected channels; the receiving channel measures the arrival time of the short pulse; the difference between the receiving time of each channel and its own single-path delay is compared; after all cross-combined measurements, a relative delay deviation matrix is ​​generated; and the inter-channel trigger time deviation calibration value is extracted from the relative delay deviation matrix.

4. The self-calibration and dynamic compensation method for multi-channel light source triggering link delay as described in claim 1, characterized in that, The process of an adaptive proportional-integral controller outputting phase step control parameters for each channel's programmable delay line includes the following steps: The exponentially weighted moving average filter recursively filters the deviation calibration values ​​of each channel's trigger time in continuous measurements, and outputs the current delay drift trend value of each channel as the feedforward compensation amount. The adaptive proportional-integral controller uses the difference between the current deviation calibration value and the delay drift trend value as the error signal. The proportional path multiplies the error signal with the adaptive proportional coefficient to obtain the proportional component. The integral path accumulates the error signal and multiplies it with the adaptive integral coefficient to obtain the integral component. The proportional component and the integral component are superimposed to obtain the original control quantity. The adaptive proportional-integral controller divides the original control quantity by the time resolution corresponding to a single step of the programmable delay line and rounds it to the nearest integer to obtain the number of phase steps that each channel needs to add. The number of phase steps is then converted into binary numbers and output as phase step control parameters to the programmable delay lines of each channel.

5. The self-calibration and dynamic compensation method for multi-channel light source triggering link delay as described in claim 4, characterized in that, The process of obtaining the original control quantity by superimposing the proportional component and the integral component includes the following steps: The adaptive proportional-integral controller subtracts the current channel trigger time deviation calibration value from the channel delay drift trend value output by the exponentially weighted moving average filter to obtain the channel's net error signal. Multiply the net error signal by the current adaptive scaling factor of the channel to obtain the scaling compensation component of the channel; The adaptive proportional-integral controller adds the historical cumulative error value of the channel to the current net error signal to obtain the updated cumulative error value; Multiply the updated cumulative error value by the current adaptive integral coefficient of the channel to obtain the integral compensation component of the channel; The adaptive proportional-integral controller adds the proportional compensation component of the channel to the integral compensation component to generate the original control quantity of the channel. After the original control values ​​of all channels are calculated in parallel, they are uniformly quantized and mapped.

6. The self-calibration and dynamic compensation method for multi-channel light source triggering link delay as described in claim 5, characterized in that, The process of adding the proportional compensation component and the integral compensation component includes the following steps: The proportional compensation component and the integral compensation component of the channel are respectively input into their respective nonlinear shaping programs. The proportional compensation component is subjected to exponential decay shaping to obtain the proportionally shaped value, and the integral compensation component is subjected to saturation compression shaping to obtain the integral shaped value. The current delay drift trend value of the channel is converted into a dynamic weight coefficient. The dynamic weight coefficient is multiplied by the proportional shaping value to obtain the proportional contribution value. The dynamic weight coefficient is subtracted from the integral shaping value to obtain the integral contribution value. The proportional contribution value and the integral contribution value are then superimposed to obtain the combined control quantity. The combined control quantity of the channel is weighted and averaged with the output control quantity of the channel in the previous cycle to obtain the smoothed control quantity. The original control quantity is then output after multiplying the smooth control quantity by the channel's link health factor.

7. The self-calibration and dynamic compensation method for multi-channel light source triggering link delay as described in claim 6, characterized in that, Exponential decay shaping refers to multiplying the proportional compensation component by a decay factor that decreases exponentially over time. The proportional shaping value is equal to the proportional compensation component multiplied by the negative time of the natural constant divided by the power of the time constant, where time is the time interval between the current control cycle and the last large adjustment, and the time constant is a preset value. Saturated compression shaping limits the integral compensation component to a preset upper and lower threshold range, with the excess portion being compressed to the threshold value; If the integral compensation component is greater than the positive saturation threshold, then the integral shaping value is equal to the positive saturation threshold. If the value is less than the negative saturation threshold, the integral-shaped value is equal to the negative saturation threshold; otherwise, it remains unchanged.

8. The self-calibration and dynamic compensation method for multi-channel light source trigger link delay as described in claim 7, characterized in that, The process of converting the current delay drift trend value of a channel into a dynamic weighting coefficient includes the following steps: The main control unit reads the current delay drift trend value of the channel, compares the current delay drift trend value with the pre-stored high drift threshold value and low drift threshold value respectively. If the high drift threshold value is greater than the low drift threshold value, three drift levels are obtained. Based on the drift level, select the corresponding fusion method of proportional and integral shaping values. For fast drift levels, select a fusion method where the proportional shaping value completely occupies the output while the integral shaping value is ignored. For medium drift levels, select a fusion method where the proportional and integral shaping values ​​are added in a one-to-one ratio. For slow drift levels, select a fusion method where the integral shaping value completely occupies the output while the proportional shaping value is ignored, and obtain the original fusion control quantity. The original fused control quantity is fed into a smoothing filter and weighted averaged with the output control quantity of the previous cycle to obtain the final combined control quantity. The smoothing coefficient is dynamically determined by the drift level. A low smoothing coefficient is used for fast drift level to achieve fast response, while a high smoothing coefficient is used for slow drift level to suppress noise.

9. The self-calibration and dynamic compensation method for multi-channel light source triggering link delay as described in claim 8, characterized in that, The drift levels are: fast drift (above the high drift threshold), medium drift (between the high and low drift thresholds), and slow drift (below the low drift threshold).

10. A self-calibration and dynamic compensation system for multi-channel light source trigger link delay, used to implement the self-calibration and dynamic compensation method for multi-channel light source trigger link delay as described in any one of claims 1 to 9, characterized in that, include: The transmission path processing module is used by the main control unit to send short measurement pulses with a width of less than 1 nanosecond. The short measurement pulses are coupled to the same transmission path via a bidirectional transceiver. The transmission path connects the main control unit with the light source driving modules of each channel. After receiving the short measurement pulses, the driving modules trigger reflection loops with a known fixed delay. The reflection loops generate an echo pulse and return to the main control unit along the original transmission path. The main control unit measures the total round-trip delay between transmission and reception, subtracts the known fixed delay from the total round-trip delay to obtain twice the one-way propagation delay, and then divides by two to obtain the one-way propagation delay of each channel. The matrix verification generation module is used to select one channel as the active transmitter by taking the single-path propagation delay of each channel as a reference. The channel driving module sends cross-verification pulses to adjacent channels or selected channels. The receiving channel measures the arrival time of the short pulse. The difference between the receiving time of each channel and its own single-path delay is compared. After all cross-combined measurements, a relative delay deviation matrix is ​​generated. The trigger time deviation calibration value between channels is extracted from the relative delay deviation matrix. The trigger signal application module is used to input the calibration value of the trigger time deviation between channels into the adaptive proportional-integral controller. Combined with the exponentially weighted moving average filter, the delay drift trend is predicted. The adaptive proportional-integral controller outputs the phase step control parameters of the programmable delay line of each channel. The phase step control parameters adjust the timing of the trigger signal application of each channel in real time to achieve closed-loop synchronous compensation of the leading edge of the trigger pulse in the entire channel range.

Images