High-speed signal acquisition and processing system and method based on laser radar

Abstract

The invention relates to the technical field of laser radar signal processing, in particular to a high-speed signal collecting and processing system and method based on a laser radar, and the system comprises a high-speed data receiving module which receives and caches the original sampling data of the laser radar; the data recombination module is used for carrying out overlapping block recombination on the data; the parallel peak detection module is used for carrying out parallel peak detection according to a self-adaptive peak degree calculation method; the boundary arbitration module is used for carrying out conflict arbitration on the boundary candidate peak points; the dynamic threshold updating module is used for updating the dynamic threshold; and the assembly line scheduling module is used for carrying out multi-stage assembly line parallel scheduling. Through a data overlapping block parallel detection architecture, a self-adaptive peak degree calculation mechanism and pipeline parallel scheduling, high-speed real-time processing of laser radar signals is realized, the processing throughput and the environmental adaptability are improved, and the technical problems of low processing speed, serious boundary effect and low resource utilization rate in the prior art are solved.

Description

Technical Field

[0001] This invention relates to the field of lidar signal processing technology, and in particular to a high-speed signal acquisition and processing system and method based on lidar. Background Technology

[0002] With the rapid development of autonomous driving, intelligent transportation, and 3D mapping, LiDAR, as a core sensing device, directly impacts system performance through its detection accuracy and response speed. LiDAR measures target distance by emitting laser pulses and receiving echo signals. Peak detection, a crucial step in LiDAR signal processing, requires real-time extraction of target echo positions from massive amounts of data acquired by high-speed ADC analog-to-digital converters. Currently, identifying LiDAR signal peaks mainly relies on fixed threshold detection or simple parallel block detection. The former is inefficient and highly dependent on preset thresholds; the latter fails to adequately consider boundary effects between overlapping data blocks, adaptive adjustments to local statistical features, and dynamic optimization of pipeline parallel scheduling, resulting in boundary peak point conflicts, poor environmental adaptability, and low resource utilization, making reliable decisions impossible in complex real-time processing environments.

[0003] Chinese Patent Publication (Announcement) No. CN119044923B discloses an FMCW lidar signal processing method and system. The method includes: dividing the laser signal generated by the laser into a measurement optical signal and a local oscillator optical signal; inputting the measurement optical signal to an optical modulator and modulating it with a modulated sinusoidal sweep frequency signal; amplifying the modulated measurement optical signal and transmitting it to the detection target through a circulator and collimating lens, and receiving the reflected signal; coupling the reflected signal with an optical signal to enhance the energy of the signal light and performing photoelectric signal conversion; amplifying the converted electrical signal and performing electrical mixing in a mixer to output an intermediate frequency (IF) signal; performing low-pass filtering on the IF signal; and converting the IF signal into a pulse square wave through a high-speed comparator in the signal processing system, performing capacitance analog interpolation, measuring the IF signal period, and calculating the IF signal frequency value. However, this scheme still suffers from problems such as the processing architecture being essentially serial, lacking an adaptive threshold adjustment mechanism, failing to address boundary effect issues, insufficient resource utilization optimization, and limited burst interference suppression capabilities, leading to a decrease in lidar signal processing speed and accuracy, and consequently, low lidar signal processing efficiency. Summary of the Invention

[0004] To address these issues, this invention provides a high-speed signal acquisition and processing system and method based on lidar, which overcomes the technical problems of slow processing speed, poor environmental adaptability, low resource utilization, and severe boundary effects in the prior art.

[0005] To achieve the above objectives, in one aspect, the present invention provides a high-speed signal acquisition and processing system based on lidar, comprising:

[0006] A high-speed data receiving module is used to receive and buffer the raw sampling data from the lidar.

[0007] The data recombination module is used to reassemble the original sampling data of the lidar by overlapping blocks according to a preset overlap length, and to obtain a number of data overlap blocks, wherein the data overlap blocks include internal data points and boundary data points.

[0008] The parallel peak detection module is used to perform parallel peak detection on each data overlap block according to the adaptive kurtosis calculation method to obtain candidate peak points of each data overlap block. The candidate peak points include internal candidate peak points and boundary candidate peak points.

[0009] The boundary arbitration module is used to arbitrate the conflict of candidate peak points at the boundary of adjacent overlapping data blocks and obtain the peak points after arbitration.

[0010] The dynamic threshold update module is used to update the dynamic threshold of the adaptive maxima calculation method based on the peak point after arbitration.

[0011] The pipeline scheduling module is used to perform multi-level pipeline parallel scheduling of the processes of data loading, peak degree calculation, neighborhood comparison and result output, and also to dynamically adjust the block strategy of the data reassembly module according to the pipeline status.

[0012] Furthermore, the high-speed data receiving module includes:

[0013] A high-speed ADC interface unit is used to sample the lidar echo analog signal to obtain the lidar raw sampling data, wherein the sampling rate is ≥1GHz;

[0014] A circular buffer unit is used to cyclically buffer the raw sampling data of the lidar, with a buffer depth ≥4×L;

[0015] The data alignment unit is used to perform bit width alignment and clock domain conversion on the cached raw LiDAR sampling data, and output the aligned sampling data to the data reconstruction module.

[0016] Furthermore, the data reconstruction module performs overlapping block reconstruction on the original LiDAR sampling data according to a preset overlap length to obtain several overlapping data blocks, specifically including:

[0017] Step D1: Obtain the sampling point sequence X={x[0],x[1],…,x[N-1]} of the original sampling data of the lidar, where N is the total number of sampling points;

[0018] Step D2: Set the data block length L and the preset overlap length L0, where L0 ≥ 2;

[0019] Step D3: Overlapping and reorganizing the sampling point sequence X to obtain a data overlap block set B={B[0],B[1],…,B[M-1]}, where the sampling point range of the k-th data overlap block B[k] is x[k×(L-L0)] to x[k×(L-L0)+L-1], and M is the number of data overlap blocks, M=⌈(NL) / (L-L0)⌉+1.

[0020] Furthermore, the parallel peak detection module performs parallel peak detection on each data overlap block according to the adaptive kurtosis calculation method to obtain candidate peak points for each data overlap block, specifically including:

[0021] Step P1: For each data overlap block B[k], extract the local statistical features of the data overlap block, including the local mean μk and the local standard deviation σk;

[0022] Step P2: Calculate the adaptive parameter based on the local statistical characteristics, and set αk = μk / σk to obtain the first adaptive parameter αk;

[0023] Step P3: Calculate the kurtosis for each sampling point x[i] in the data overlap block B[k]. Set the formula for calculating the kurtosis P(x[i]) as follows:

[0024] P(x[i])=(x[i]−μk) / (αk×σk+β×∣Δx[i]∣+γ);

[0025] Where Δx[i]=x[i]-x[i-1] is the rate of change of the sampling point, β is the rate of change suppression coefficient, and γ is the stability constant;

[0026] Step P4: Compare the calculated peak value P(x[i]) with the dynamic threshold T, and mark the candidate peak points according to the comparison result, where:

[0027] When P(x[i]) > T and x[i-1] < x[i] and x[i] > x[i+1], the sampling point x[i] is marked as a candidate peak point;

[0028] When P(x[i])≤T or x[i-1]≥x[i] or x[i]≤x[i+1], the sampling point x[i] is not marked as a candidate peak point.

[0029] Furthermore, the boundary arbitration module performs conflict arbitration on the candidate peak points of adjacent overlapping data blocks to obtain the arbitration peak points, specifically including:

[0030] Step A1: Identify the boundary region of adjacent overlapping data blocks B[k] and B[k+1]. The boundary region is the overlapping region formed by the last L0 sampling points of B[k] and the first L0 sampling points of B[k+1].

[0031] Step A2: Extract the set of candidate peak points PB={pb[0],pb[1],…,pb[Q-1]} in the boundary region, where Q is the number of candidate peak points.

[0032] Step A3: For each boundary candidate peak point pb[q], obtain its peak degree set {Pleft, Pright} in the adjacent data overlapping block, where Pleft is the peak degree calculated in B[k] and Pright is the peak degree calculated in B[k+1].

[0033] Step A4: Arbitrate the candidate peak points pb[q] at the boundary based on the set of peak degrees, where:

[0034] When Pleft > T and Pright > T, compare Pleft and Pright, and confirm the boundary candidate peak point corresponding to the one with the larger kurtosis as the arbitrated peak point, and discard the other boundary candidate peak point.

[0035] When Pleft>T and Pright≤T, the boundary candidate peak points in B[k] are confirmed as the arbitrated peak points;

[0036] When Pleft≤T and Pright>T, the boundary candidate peak points in B[k+1] are confirmed as the arbitrated peak points;

[0037] When Pleft≤T and Pright≤T, the boundary candidate peak point pb[q] is removed from the candidate peak points.

[0038] Furthermore, the dynamic threshold update module updates the dynamic threshold of the adaptive kurtosis calculation method based on the arbitrated peak point, specifically including:

[0039] Step G1: Count the number of peak points Np and the total number of sampling points Ns after arbitration in the current processing cycle, and calculate the peak density ρ=Np / Ns;

[0040] Step G2: Compare the peak density ρ with the preset density threshold ρ0, and determine the adjustment direction of the dynamic threshold T based on the comparison result, wherein:

[0041] When ρ > ρ0, the peak density is determined to be too high, and the dynamic threshold T is adjusted upwards. The adjusted dynamic threshold is then set. , where δup is the up-adjustment coefficient, 0 < δup ≤ 0.1;

[0042] When ρ < ρ0, the peak density is determined to be too low, and the dynamic threshold T is adjusted downwards. The adjusted dynamic threshold is then set. Where δdown is the down-adjustment coefficient, 0 < δdown ≤ 0.1;

[0043] When ρ=ρ0, the peak density is determined to be normal, and the dynamic threshold T is not adjusted.

[0044] Step G3: Output the adjusted dynamic threshold T' as the dynamic threshold T for the next processing cycle.

[0045] Furthermore, the pipeline scheduling module performs multi-level pipeline parallel scheduling for the processes of data loading, peak degree calculation, neighborhood comparison, and result output, specifically including:

[0046] Step H1: Set the number of pipeline stages R=4, which are data loading stage S1, peak value calculation stage S2, neighborhood comparison stage S3 and result output stage S4.

[0047] Step H2, in the nth clock cycle, each stage of the pipeline executes simultaneously:

[0048] Data loading level S1: Load the nth data overlap block B[n] into the cache;

[0049] Peak degree calculation level S2: Calculate the peak degree of the (n-1)th data overlap block B[n-1];

[0050] Neighborhood comparison level S3: Perform neighborhood comparison and candidate peak point marking on the (n-2)th data overlap block B[n-2];

[0051] Output stage S4: Outputs the arbitrated peak point of the (n-3)th data overlap block B[n-3];

[0052] Step H3 involves passing data at each stage through pipeline registers to achieve uninterrupted processing of continuous data streams.

[0053] Furthermore, the pipeline scheduling module dynamically adjusts the block division strategy of the data reassembly module according to the pipeline status, specifically including:

[0054] Step V1: Monitor the processing delays {Δt1,Δt2,Δt3,Δt4} of each stage of the pipeline and calculate the pipeline imbalance index Imbalance=max(Δtr)-min(Δtr), where r∈{1,2,3,4};

[0055] Step V2: Compare the pipeline imbalance index Imbalance with the preset imbalance threshold Imbalance0, and determine the need to adjust the segmentation strategy based on the comparison result, wherein:

[0056] When Imbalance ≤ Imbalance0, the pipeline is considered balanced and no adjustment is made to the block strategy;

[0057] When Imbalance > Imbalance0, the pipeline is determined to be unbalanced, and the block partitioning strategy is adjusted. The adjustment includes:

[0058] If Δt1 = max(Δtr), increase the data block length L, and set the adjusted data block length L' = L × (1 + ε), where 0 < ε ≤ 0.2;

[0059] If Δt2=max(Δtr), reduce the preset overlap length L0, and set the adjusted preset overlap length L0'=L0×(1-ε);

[0060] If Δt3=max(Δtr), reduce the preset overlap length L0, and set the adjusted preset overlap length L0'=L0×(1-ε);

[0061] If Δt4 = max(Δtr), reduce the data block length L, and set the adjusted data block length L' = L × (1 - ε).

[0062] Furthermore, the parallel peak detection module also includes a burst interference suppression unit to detect and suppress outliers in the rate of change of sampling points Δx[i], specifically including:

[0063] Step U1: Calculate the rate of change of the current sampling point Δx[i] = x[i] - x[i-1];

[0064] Step U2, obtain the historical rate of change set ΔXhistory={Δx[i-1],Δx[i-2],…,Δx[iW]}, where W is the length of the historical window, W≥3;

[0065] Step U3: Calculate the historical rate of change mean μΔ and the historical rate of change standard deviation σΔ;

[0066] Step U4: Compare |Δx[i]| with the burst interference threshold Thburst=μΔ+k×σΔ, where k is the confidence coefficient and k≥3;

[0067] Step U5: Dynamically adjust the rate of change inhibition coefficient β based on the comparison results, where:

[0068] When |Δx[i]|≤Thburst, it is determined to be a normal change, and β=βbase is set;

[0069] When |Δx[i]|>Thburst, it is determined to be a sudden disturbance, and β=βburst is set, where βburst>βbase.

[0070] On the other hand, the present invention also provides a high-speed signal acquisition and processing method based on lidar, including:

[0071] Step J1: Receive and buffer the raw sampling data from the lidar;

[0072] Step J2: The original sampling data of the lidar is recombined by overlapping blocks according to the preset overlap length to obtain several data overlap blocks, wherein the data overlap blocks include internal data points and boundary data points;

[0073] Step J3: Perform parallel peak detection on each data overlap block according to the adaptive kurtosis calculation method to obtain candidate peak points for each data overlap block. The candidate peak points include internal candidate peak points and boundary candidate peak points.

[0074] Step J4: Perform conflict arbitration on the candidate peak points of the boundary of adjacent overlapping data blocks to obtain the arbitration peak points;

[0075] Step J5: Update the dynamic threshold of the adaptive kurtosis calculation method based on the peak point after arbitration;

[0076] Step J6 involves performing multi-level pipeline parallel scheduling for the processes of data loading, peak degree calculation, neighborhood comparison, and result output, and dynamically adjusting the block strategy based on the pipeline status.

[0077] Compared with existing technologies, the beneficial effects of this invention are as follows: The system achieves seamless connection between data blocks through the data overlap block design of the data reassembly module and the conflict arbitration mechanism of the boundary arbitration module, solving the boundary effect problem during parallel block processing, thereby ensuring the continuity and accuracy of peak detection; The system also achieves adaptive suppression of environmental noise and sudden interference through the adaptive peak degree calculation method of the parallel peak detection module, comprehensively considering signal amplitude, local noise statistical characteristics, and sampling point change rate, thereby improving the detection reliability in complex environments; The system also achieves the processing capacity of one data block per clock cycle through the multi-level pipeline parallel scheduling of the pipeline scheduling module, overlapping the data loading, calculation, comparison, and output processes, thereby improving the system throughput; The system also achieves optimal hardware resource configuration through the dynamic adjustment of the block strategy by the pipeline scheduling module, adaptively optimizing the data block length and overlap length according to the processing delay of each pipeline level, thereby improving the utilization rate of FPGA resources; The system also achieves adaptive balance of detection sensitivity through the peak density feedback mechanism of the dynamic threshold update module, dynamically adjusting the detection threshold according to the detection results, thereby reducing the risk of over-detection and under-detection. Attached Figure Description

[0078] Figure 1This is a schematic diagram of the high-speed signal acquisition and processing system based on lidar in this embodiment;

[0079] Figure 2 This is a schematic diagram of the multi-level pipeline parallel scheduling of the pipeline scheduling module in this embodiment;

[0080] Figure 3 This is a flowchart illustrating the high-speed signal acquisition and processing method based on lidar in this embodiment. Detailed Implementation

[0081] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0082] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0083] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0084] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0085] Please see Figure 1 As shown, this is a schematic diagram of the high-speed signal acquisition and processing system based on lidar in this embodiment. The system includes:

[0086] A high-speed data receiving module is used to receive and buffer the raw sampling data from the lidar.

[0087] The data reconstruction module is used to reconstruct the original sampling data of the lidar by overlapping blocks according to a preset overlap length, resulting in several data overlap blocks. The data overlap blocks include internal data points and boundary data points. The data reconstruction module is connected to the high-speed data receiving module.

[0088] A parallel peak detection module is used to perform parallel peak detection on each data overlap block according to the adaptive kurtosis calculation method to obtain candidate peak points of each data overlap block. The candidate peak points include internal candidate peak points and boundary candidate peak points. The parallel peak detection module is connected to the data reconstruction module.

[0089] The boundary arbitration module is used to arbitrate the conflict of candidate peak points at the boundary of adjacent overlapping data blocks to obtain the peak points after arbitration. The boundary arbitration module is connected to the parallel peak detection module.

[0090] A dynamic threshold update module is used to update the dynamic threshold of the adaptive maxima calculation method based on the peak point after arbitration. The dynamic threshold update module is connected to the boundary arbitration module.

[0091] The pipeline scheduling module is used to perform multi-level pipeline parallel scheduling of the processes of data loading, peak degree calculation, neighborhood comparison and result output. It is also used to dynamically adjust the block strategy of the data reassembly module according to the pipeline status. The pipeline scheduling module is connected to the dynamic threshold update module.

[0092] Specifically, the high-speed signal acquisition and processing system based on lidar is applied in a lidar signal processing terminal. This terminal includes an FPGA processing board, a high-speed ADC acquisition card, and a lidar optical front-end. Through a data overlap block parallel detection architecture, an adaptive peak degree calculation mechanism, and pipelined parallel scheduling, it achieves high-speed real-time processing of lidar signals, improving processing throughput and environmental adaptability. Specifically, the system achieves seamless connection between data blocks through the data overlap block design of the data reassembly module and the conflict arbitration mechanism of the boundary arbitration module, solving the boundary effect problem during block parallel processing and thus ensuring the continuity and accuracy of peak detection. Furthermore, the system uses an adaptive peak degree calculation method in the parallel peak detection module to comprehensively consider signal amplitude and local noise. The system utilizes statistical characteristics and sampling point change rates to adaptively suppress environmental noise and sudden interference, thereby improving detection reliability in complex environments. Furthermore, through multi-level pipeline parallel scheduling of the pipeline scheduling module, the data loading, calculation, comparison, and output processes are overlapped, achieving a single-clock-cycle throughput of one data block, thus increasing system throughput. The system also dynamically adjusts the block segmentation strategy through the pipeline scheduling module, adaptively optimizing the data block length and overlap length based on the processing delay of each pipeline level, achieving optimal hardware resource allocation and improving FPGA resource utilization. Finally, the system employs a peak density feedback mechanism in the dynamic threshold update module to dynamically adjust the detection threshold based on the detection results, achieving an adaptive balance in detection sensitivity and reducing the risk of over-detection and under-detection.

[0093] Specifically, the high-speed data receiving module includes:

[0094] A high-speed ADC interface unit is used to sample the lidar echo analog signal to obtain the lidar raw sampling data, wherein the sampling rate is ≥1GHz;

[0095] A circular buffer unit is used to cyclically buffer the raw sampling data of the lidar, with a buffer depth ≥4×L;

[0096] The data alignment unit is used to perform bit width alignment and clock domain conversion on the original LiDAR sampling data after cyclic buffering to obtain aligned sampling data, and output the aligned sampling data to the data reconstruction module.

[0097] Specifically, the raw sampling data of the lidar refers to the raw echo signal data acquired by the lidar through a high-speed ADC analog-to-digital converter. The high-speed ADC interface unit refers to the hardware interface unit that implements high-speed analog-to-digital conversion. This embodiment does not limit the specific model of the high-speed ADC interface unit; those skilled in the art can set it according to actual conditions. For example, this embodiment uses TI. The ADC12DJ3200 chip, the sampling rate refers to the sampling frequency of the analog signal by the high-speed ADC interface unit. This embodiment does not limit the specific value of the sampling rate; those skilled in the art can set it according to actual conditions. For example, this embodiment sets the sampling rate to ≥1GHz. The ring buffer unit refers to a data buffer unit using a circular buffer structure. The buffer depth refers to the number of sampling points that the ring buffer unit can store. This embodiment does not limit the specific value of the buffer depth; those skilled in the art can set it according to actual conditions. For example, this embodiment sets the buffer depth to ≥4×L. The data alignment unit refers to a hardware unit that realizes data bit width alignment and cross-clock domain conversion. This embodiment does not limit the specific implementation method of aligning the bit width of the original LiDAR sampling data; those skilled in the art can set it according to actual conditions. For example, expanding 12-bit sampling data to 16-bit for bit width alignment. This embodiment does not limit the specific implementation method of clock domain conversion of the buffered original LiDAR sampling data; those skilled in the art can set it according to actual conditions. For example, cross-clock domain conversion can be completed through asynchronous FIFO.

[0098] Specifically, the data reconstruction module performs overlapping block reconstruction on the original LiDAR sampling data according to a preset overlap length to obtain several overlapping data blocks, including:

[0099] Step D1: Obtain the sampling point sequence X={x[0],x[1],…,x[N-1]} of the original sampling data of the lidar, where N is the total number of sampling points;

[0100] Step D2: Set the data block length L and the preset overlap length L0, where L0 ≥ 2;

[0101] Step D3: Overlapping and reorganizing the sampling point sequence X to obtain a data overlap block set B={B[0],B[1],…,B[M-1]}, where the sampling point range of the k-th data overlap block B[k] is x[k×(L-L0)] to x[k×(L-L0)+L-1], and M is the number of data overlap blocks, M=⌈(NL) / (L-L0)⌉+1.

[0102] Specifically, the sampling point sequence refers to a set of discrete sampling points arranged in chronological order from the original LiDAR sampling data. The data block length L refers to the number of sampling points contained in each data overlap block. This embodiment does not limit the specific value of the data block length L; those skilled in the art can set it according to actual conditions. For example, in this embodiment, L=1024. The preset overlap length L0 refers to the number of overlapping sampling points between adjacent data overlap blocks. This embodiment does not limit the specific value of the preset overlap length L0; those skilled in the art can set it according to actual conditions. For example, in this embodiment, L0=2. The data overlap block refers to a data block containing internal data points and boundary data points, obtained through overlapping block recombination. Internal data points refer to sampling points in the data overlap block that do not overlap with adjacent data blocks. Boundary data points refer to sampling points in the data overlap block that overlap with adjacent data blocks. The symbol ⌈⌉ represents the floor function, which takes the smallest integer greater than or equal to the value in parentheses. For example, ⌈2.3⌉=3, ⌈5⌉=5. In the expression ⌈(NL) / (L-L0)⌉, it means that the value of (NL) / (L-L0) is calculated first, and then the value is rounded up. For example, in this embodiment, N=12500, L=1024, L0=2, then ⌈(12500-1024) / 1022⌉≈11.2. Rounding 11.2 up gives 12, so the value of M=⌈(12500-1024) / 1022⌉+1 is 13.

[0103] Specifically, the data reorganization module achieves seamless connection between data blocks through the design of overlapping data blocks, solves the boundary effect problem when processing blocks in parallel, and thus ensures the continuity and accuracy of peak detection.

[0104] Specifically, the parallel peak detection module performs parallel peak detection on each data overlap block according to the adaptive kurtosis calculation method to obtain candidate peak points for each data overlap block, specifically including:

[0105] Step P1: For each data overlap block B[k], extract the local statistical features of the data overlap block, including the local mean μk and the local standard deviation σk;

[0106] Step P2: Calculate the adaptive parameter based on the local statistical characteristics, and set αk = μk / σk to obtain the first adaptive parameter αk;

[0107] Step P3: Calculate the kurtosis for each sampling point x[i] in the data overlap block B[k]. Set the formula for calculating the kurtosis P(x[i]) as follows:

[0108] P(x[i])=(x[i]−μk) / (αk×σk+β×|Δx[i]|+γ);

[0109] Where Δx[i]=x[i]-x[i-1] is the rate of change of the sampling point, β is the rate of change suppression coefficient, and γ is the stability constant;

[0110] Step P4: Compare the calculated peak value P(x[i]) with the dynamic threshold T, and mark the candidate peak points according to the comparison result, where:

[0111] When P(x[i]) > T and x[i-1] < x[i] and x[i] > x[i+1], the sampling point x[i] is marked as a candidate peak point; otherwise, the sampling point x[i] is not marked as a candidate peak point.

[0112] Specifically, the local mean μk refers to the arithmetic mean of all sampling points in the data overlap block B[k], the local standard deviation σk refers to the standard deviation of all sampling points in the data overlap block B[k], the first adaptive parameter αk is the ratio of the local mean to the local standard deviation, used to characterize the local signal-to-noise ratio, the sampling point change rate Δx[i] is the difference between the current sampling point and the previous sampling point, used to characterize the signal change trend, and the change rate suppression coefficient β is a coefficient used to suppress the influence of sudden interference on the peak value calculation. This embodiment does not limit the specific value setting of the change rate suppression coefficient β. Those skilled in the art can set it according to the actual situation. For example, in this embodiment, βbase=0.5 and βburst=2 are set. The stability constant γ is a constant term used to prevent the denominator from being zero. This embodiment does not limit the specific value of the stability constant γ. Those skilled in the art can set it according to the actual situation. For example, in this embodiment, γ=65. The dynamic threshold T is an adaptive threshold used to judge candidate peak points. This embodiment does not limit the initial value T0 of the dynamic threshold T. Those skilled in the art can set it according to the actual situation. For example, in this embodiment, T0=3.0. The candidate peak point is a sampling point that satisfies the peak degree being greater than the dynamic threshold and satisfies the neighborhood comparison condition. The internal candidate peak point is a candidate peak point among the data points inside the data overlap block. The boundary candidate peak point is a candidate peak point among the data points at the boundary of the data overlap block.

[0113] Specifically, the parallel peak detection module uses an adaptive peak degree calculation method to comprehensively consider signal amplitude, local noise statistical characteristics, and sampling point change rate, thereby achieving adaptive suppression of environmental noise and sudden interference, thus improving the detection reliability in complex environments.

[0114] Specifically, the boundary arbitration module arbitrates conflicts between candidate peak points at the boundaries of adjacent overlapping data blocks to obtain the arbitrated peak points, including:

[0115] Step A1: Identify the boundary region of adjacent overlapping data blocks B[k] and B[k+1]. The boundary region is the overlapping region formed by the last L0 sampling points of B[k] and the first L0 sampling points of B[k+1].

[0116] Step A2: Extract the set of candidate peak points PB={pb[0],pb[1],…,pb[Q-1]} in the boundary region, where Q is the number of candidate peak points.

[0117] Step A3: For each boundary candidate peak point pb[q], obtain its peak degree set {Pleft, Pright} in the adjacent data overlapping block, where Pleft is the peak degree calculated in B[k] and Pright is the peak degree calculated in B[k+1].

[0118] Step A4: Arbitrate the candidate peak points pb[q] at the boundary based on the set of peak degrees, where:

[0119] When Pleft > T and Pright > T, compare Pleft and Pright, and confirm the boundary candidate peak point corresponding to the one with the larger kurtosis as the arbitrated peak point, and discard the other boundary candidate peak point.

[0120] When Pleft>T and Pright≤T, the boundary candidate peak points in B[k] are confirmed as the arbitrated peak points;

[0121] When Pleft≤T and Pright>T, the boundary candidate peak points in B[k+1] are confirmed as the arbitrated peak points;

[0122] When Pleft≤T and Pright≤T, the boundary candidate peak point pb[q] is removed from the candidate peak points.

[0123] Specifically, the boundary region refers to the overlapping sampling point region between adjacent data overlapping blocks, the boundary candidate peak point set PB refers to the sampling point set marked as candidate peak points in the boundary region, the peak degree Pleft refers to the peak degree of the boundary candidate peak point calculated in the left data overlapping block B[k], the peak degree Pright refers to the peak degree of the boundary candidate peak point calculated in the right data overlapping block B[k+1], and the arbitrated peak point refers to the boundary peak point that is confirmed after conflict arbitration and has no repetition.

[0124] Specifically, the boundary arbitration module resolves the boundary effect problem during block parallel processing by arbitrating the boundary candidate peak points of adjacent overlapping data blocks, avoiding repeated detection or missed detection of boundary peak points, thereby ensuring the continuity and accuracy of peak detection.

[0125] Specifically, the dynamic threshold update module updates the dynamic threshold of the adaptive kurtosis calculation method based on the arbitrated peak point, including:

[0126] Step G1: Count the number of peak points Np and the total number of sampling points Ns after arbitration in the current processing cycle, and calculate the peak density ρ=Np / Ns;

[0127] Step G2: Compare the peak density ρ with the preset density threshold ρ0, and determine the adjustment direction of the dynamic threshold T based on the comparison result, wherein:

[0128] When ρ > ρ0, the peak density is determined to be too high, and the dynamic threshold T is adjusted upward. The adjusted dynamic threshold T' = T × (1 + δup) is set, where δup is the adjustment coefficient, 0 < δup ≤ 0.1.

[0129] When ρ < ρ0, the peak density is determined to be too low, and the dynamic threshold T is adjusted downward. The adjusted dynamic threshold T' = T × (1 - δdown) is set, where δdown is the adjustment coefficient, 0 < δdown ≤ 0.1.

[0130] When ρ=ρ0, the peak density is determined to be normal, and the dynamic threshold T is not adjusted.

[0131] Step G3: Output the adjusted dynamic threshold T' as the dynamic threshold T for the next processing cycle.

[0132] Specifically, the current processing cycle refers to the time window for dynamic threshold updates. This embodiment does not limit the specific duration of the current processing cycle; those skilled in the art can set it according to actual conditions. For example, in this embodiment, the processing cycle is set to 100 data blocks. The number of peak points after arbitration, Np, refers to the total number of peak points confirmed after boundary arbitration within the current processing cycle. The total number of sampling points, Ns, refers to the total number of sampling points processed within the current processing cycle. The peak density, ρ, refers to the ratio of the number of peak points after arbitration to the total number of sampling points. The preset density threshold, ρ0, refers to the benchmark value used to determine whether the peak density is normal. This embodiment does not specify the exact value of the preset density threshold, ρ0. The numerical values ​​are limited, but those skilled in the art can set them according to actual conditions. For example, in this embodiment, ρ0 is set to 0.01. The upward adjustment coefficient δup refers to the upward adjustment ratio of the dynamic threshold when the peak density is too high. This embodiment does not limit the specific value of the upward adjustment coefficient δup, but those skilled in the art can set it according to actual conditions. For example, in this embodiment, δup is set to 0.05. The downward adjustment coefficient δdown refers to the downward adjustment ratio of the dynamic threshold when the peak density is too low. This embodiment does not limit the specific value of the downward adjustment coefficient δdown, but those skilled in the art can set it according to actual conditions. For example, in this embodiment, δdown is set to 0.05.

[0133] Specifically, the dynamic threshold update module uses a peak density feedback mechanism to dynamically adjust the detection threshold based on the detection results, thereby achieving an adaptive balance in detection sensitivity and reducing the risk of over-detection and under-detection.

[0134] Please see Figure 2 As shown, this is a schematic diagram of the multi-level pipeline parallel scheduling of the pipeline scheduling module in this embodiment. The diagram shows the timing coordination relationship of the four-level pipelines.

[0135] Specifically, the pipeline scheduling module performs multi-level pipeline parallel scheduling for the processes of data loading, peak degree calculation, neighborhood comparison, and result output, including:

[0136] Step H1: Set the number of pipeline stages R=4, which are data loading stage S1, peak value calculation stage S2, neighborhood comparison stage S3 and result output stage S4.

[0137] Step H2, in the nth clock cycle, each stage of the pipeline executes simultaneously:

[0138] Data loading level S1: Load the nth data overlap block B[n] into the cache;

[0139] Peak degree calculation level S2: Calculate the peak degree of the (n-1)th data overlap block B[n-1];

[0140] Neighborhood comparison level S3: Perform neighborhood comparison and candidate peak point marking on the (n-2)th data overlap block B[n-2];

[0141] Output stage S4: Outputs the arbitrated peak point of the (n-3)th data overlap block B[n-3];

[0142] Step H3 involves passing data at each stage through pipeline registers to achieve uninterrupted processing of continuous data streams.

[0143] Specifically, the pipeline stage R refers to the number of stages in the pipeline parallel scheduling. This embodiment does not limit the specific value of the pipeline stage R. Those skilled in the art can set it according to the actual situation. For example, in this embodiment, R=4 is set. The data loading stage S1 refers to the processing stage that loads the data overlap block from the external memory to the cache. The peak degree calculation stage S2 refers to the processing stage that performs adaptive peak degree calculation on the data overlap block. The neighborhood comparison stage S3 refers to the processing stage that performs neighborhood comparison and candidate peak point marking on the data overlap block. The result output stage S4 refers to the processing stage that outputs the peak point after arbitration. The pipeline register refers to the register group used to transfer data between the pipeline stages.

[0144] Specifically, the pipeline scheduling module uses multi-level pipeline parallel scheduling to overlap the data loading, calculation, comparison and output processes, thereby achieving the processing capacity of processing one data block per clock cycle, thus improving system throughput.

[0145] Specifically, the pipeline scheduling module dynamically adjusts the block division strategy of the data reassembly module according to the pipeline status, including:

[0146] Step V1: Monitor the processing delays {Δt1,Δt2,Δt3,Δt4} of each stage of the pipeline and calculate the pipeline imbalance index Imbalance=max(Δtr)-min(Δtr), where r∈{1,2,3,4};

[0147] Step V2: Compare the pipeline imbalance index Imbalance with the preset imbalance threshold Imbalance0, and determine the need to adjust the segmentation strategy based on the comparison result, wherein:

[0148] When Imbalance ≤ Imbalance0, the pipeline is considered balanced and no adjustment is made to the block strategy;

[0149] When Imbalance > Imbalance0, the pipeline is determined to be unbalanced, and the block partitioning strategy is adjusted. The adjustment includes:

[0150] If Δt1 = max(Δtr), increase the data block length L, and set the adjusted data block length L' = L × (1 + ε), where 0 < ε ≤ 0.2;

[0151] If Δt2=max(Δtr), reduce the preset overlap length L0, and set the adjusted preset overlap length L0'=L0×(1-ε);

[0152] If Δt3=max(Δtr), reduce the preset overlap length L0, and set the adjusted preset overlap length L0'=L0×(1-ε);

[0153] If Δt4 = max(Δtr), reduce the data block length L, and set the adjusted data block length L' = L × (1 - ε).

[0154] Specifically, Δt1 is the processing delay of the data loading stage S1, Δt2 is the processing delay of the peak degree calculation stage S2, Δt3 is the processing delay of the neighborhood comparison stage S3, and Δt4 is the processing delay of the result output stage S4. The processing delay Δtr refers to the processing time of the r-th stage pipeline. The pipeline imbalance index Imbalance refers to the difference between the maximum and minimum processing delays of each stage pipeline, used to characterize the degree of imbalance of the load at each stage of the pipeline. The preset imbalance threshold Imbalance0 is a critical value used to determine whether the pipeline is unbalanced. This embodiment does not limit the specific value of the preset imbalance threshold Imbalance0. Those skilled in the art can set it according to the actual situation. For example, in this embodiment, Imbalance0 is set to 5 clock cycles. The adjustment coefficient ε refers to the magnitude coefficient of the block strategy adjustment. This embodiment does not limit the specific value of the adjustment coefficient ε. Those skilled in the art can set it according to the actual situation. For example, in this embodiment, ε is set to 0.1.

[0155] Specifically, the pipeline scheduling module dynamically adjusts the block segmentation strategy and adaptively optimizes the data block length and overlap length based on the processing delay of each pipeline level, thereby achieving optimal hardware resource allocation and improving FPGA resource utilization.

[0156] Specifically, the parallel peak detection module further includes a burst interference suppression unit, used to detect and suppress outliers in the rate of change of sampling points Δx[i], specifically including:

[0157] Step U1: Calculate the rate of change of the current sampling point Δx[i] = x[i] - x[i-1];

[0158] Step U2, obtain the historical rate of change set ΔXhistory={Δx[i-1],Δx[i-2],…,Δx[iW]}, where W is the length of the historical window, W≥3;

[0159] Step U3: Calculate the historical rate of change mean μΔ and the historical rate of change standard deviation σΔ;

[0160] Step U4: Compare |Δx[i]| with the burst interference threshold Thburst=μΔ+k×σΔ, where k is the confidence coefficient and k≥3;

[0161] Step U5: Dynamically adjust the rate of change inhibition coefficient β based on the comparison results, where:

[0162] When |Δx[i]|≤Thburst, it is determined to be a normal change, and β=βbase is set;

[0163] When |Δx[i]|>Thburst, it is determined to be a sudden disturbance, and β=βburst is set, where βburst>βbase.

[0164] Specifically, the historical rate of change set ΔXhistory refers to the set of rates of change of the W sampling points preceding the current sampling point. The historical window length W refers to the number of rates of change contained in the historical rate of change set. This embodiment does not limit the specific value of the historical window length W; those skilled in the art can set it according to actual conditions. For example, in this embodiment, W=8. The historical rate of change mean μΔ refers to the arithmetic mean of the historical rate of change set. The historical rate of change standard deviation σΔ refers to the standard deviation of the historical rate of change set. The sudden disturbance threshold Thburst refers to the threshold used to determine whether the rate of change of the current sampling point is abnormal. The confidence coefficient k refers to the confidence level coefficient of the sudden disturbance threshold. This embodiment does not limit the specific value of the confidence coefficient k; those skilled in the art can set it according to actual conditions. For example, in this embodiment, k=3, i.e., the 3σ criterion is adopted.

[0165] Specifically, the sudden interference suppression unit detects and suppresses abnormal values ​​of the rate of change of sampling points and dynamically adjusts the rate of change suppression coefficient to achieve adaptive suppression of sudden interference, thereby improving the detection reliability in complex environments.

[0166] Please see Figure 3 As shown, this is a flowchart illustrating the high-speed signal acquisition and processing method based on lidar in this embodiment. The method includes:

[0167] Step J1: Receive and buffer the raw sampling data from the lidar;

[0168] Step J2: The original sampling data of the lidar is recombined by overlapping blocks according to the preset overlap length to obtain several data overlap blocks, wherein the data overlap blocks include internal data points and boundary data points;

[0169] Step J3: Perform parallel peak detection on each data overlap block according to the adaptive kurtosis calculation method to obtain candidate peak points for each data overlap block. The candidate peak points include internal candidate peak points and boundary candidate peak points.

[0170] Step J4: Perform conflict arbitration on the candidate peak points of the boundary of adjacent overlapping data blocks to obtain the arbitration peak points;

[0171] Step J5: Update the dynamic threshold of the adaptive kurtosis calculation method based on the peak point after arbitration;

[0172] Step J6 involves performing multi-level pipeline parallel scheduling for the processes of data loading, peak degree calculation, neighborhood comparison, and result output, and dynamically adjusting the block strategy based on the pipeline status.

[0173] Specifically, in step J1, the raw sampling data of the lidar is received and buffered.

[0174] Specifically, in step J2, the original sampling data of the lidar is recombined by overlapping blocks according to a preset overlap length to obtain several data overlap blocks, wherein the data overlap blocks include internal data points and boundary data points.

[0175] Specifically, in step J3, parallel peak detection is performed on each data overlap block according to the adaptive kurtosis calculation method to obtain candidate peak points for each data overlap block. The candidate peak points include internal candidate peak points and boundary candidate peak points.

[0176] Specifically, in step J4, conflict arbitration is performed on the candidate peak points of the boundary of adjacent overlapping data blocks to obtain the arbitration peak points.

[0177] Specifically, in step J5, the dynamic threshold of the adaptive peak degree calculation method is updated based on the peak point after arbitration.

[0178] Specifically, in step J6, the processes of data loading, peak degree calculation, neighborhood comparison and result output are scheduled in parallel by a multi-level pipeline, and the block strategy is dynamically adjusted according to the pipeline status.

[0179] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A high-speed signal acquisition and processing system based on lidar, characterized in that, include: A high-speed data receiving module is used to receive and buffer the raw sampling data from the lidar. The data recombination module is used to reassemble the original sampling data of the lidar by overlapping blocks according to a preset overlap length, and to obtain a number of data overlap blocks, wherein the data overlap blocks include internal data points and boundary data points. The parallel peak detection module is used to perform parallel peak detection on each data overlap block according to the adaptive kurtosis calculation method to obtain candidate peak points of each data overlap block. The candidate peak points include internal candidate peak points and boundary candidate peak points. The boundary arbitration module is used to arbitrate the conflict of candidate peak points at the boundary of adjacent overlapping data blocks and obtain the peak points after arbitration. The dynamic threshold update module is used to update the dynamic threshold of the adaptive maxima calculation method based on the peak point after arbitration. The pipeline scheduling module is used to perform multi-level pipeline parallel scheduling of the processes of data loading, peak degree calculation, neighborhood comparison and result output, and also to dynamically adjust the block strategy of the data reassembly module according to the pipeline status.

2. The high-speed signal acquisition and processing system based on lidar according to claim 1, characterized in that, The high-speed data receiving module includes: A high-speed ADC interface unit is used to sample the lidar echo analog signal to obtain the lidar raw sampling data, wherein the sampling rate is ≥1GHz; A circular buffer unit is used to cyclically buffer the raw sampling data of the lidar, with a buffer depth ≥4×L; The data alignment unit is used to perform bit width alignment and clock domain conversion on the cached raw LiDAR sampling data, and output the aligned sampling data to the data reconstruction module.

3. The high-speed signal acquisition and processing system based on lidar according to claim 1, characterized in that, The data reconstruction module performs overlapping block reconstruction on the original sampling data of the lidar according to a preset overlap length, resulting in several overlapping data blocks, specifically including: Step D1: Obtain the sampling point sequence X={x[0],x[1],…,x[N-1]} of the original sampling data of the lidar, where N is the total number of sampling points; Step D2: Set the data block length L and the preset overlap length L0, where L0 ≥ 2; Step D3: Overlapping and reorganizing the sampling point sequence X to obtain a data overlap block set B={B[0],B[1],…,B[M-1]}, where the sampling point range of the k-th data overlap block B[k] is x[k×(L-L0)] to x[k×(L-L0)+L-1], and M is the number of data overlap blocks, M=⌈(NL) / (L-L0)⌉+1.

4. The high-speed signal acquisition and processing system based on lidar according to claim 3, characterized in that, The parallel peak detection module performs parallel peak detection on each data overlap block according to the adaptive kurtosis calculation method to obtain candidate peak points for each data overlap block, specifically including: Step P1: For each data overlap block B[k], extract the local statistical features of the data overlap block, including the local mean μk and the local standard deviation σk; Step P2: Calculate the adaptive parameter based on the local statistical characteristics, and set αk = μk / σk to obtain the first adaptive parameter αk; Step P3: Calculate the kurtosis for each sampling point x[i] in the data overlap block B[k]. Set the formula for calculating the kurtosis P(x[i]) as follows: P(x[i])=(x[i]−μk) / (αk×σk+β×∣Δx[i]∣+γ); Where Δx[i]=x[i]-x[i-1] is the rate of change of the sampling point, β is the rate of change suppression coefficient, and γ is the stability constant; Step P4: Compare the calculated peak value P(x[i]) with the dynamic threshold T, and mark the candidate peak points according to the comparison result, where: When P(x[i]) > T and x[i-1] < x[i] and x[i] > x[i+1], the sampling point x[i] is marked as a candidate peak point; otherwise, the sampling point x[i] is not marked as a candidate peak point.

5. The high-speed signal acquisition and processing system based on lidar according to claim 4, characterized in that, The boundary arbitration module arbitrates conflicts between candidate peak points at the boundaries of adjacent overlapping data blocks to obtain the arbitration peak points, specifically including: Step A1: Identify the boundary region of adjacent overlapping data blocks B[k] and B[k+1]. The boundary region is the overlapping region formed by the last L0 sampling points of B[k] and the first L0 sampling points of B[k+1]. Step A2: Extract the set of candidate peak points PB={pb[0],pb[1],…,pb[Q-1]} in the boundary region, where Q is the number of candidate peak points. Step A3: For each boundary candidate peak point pb[q], obtain its peak degree set {Pleft, Pright} in the adjacent data overlapping block, where Pleft is the peak degree calculated in B[k] and Pright is the peak degree calculated in B[k+1]. Step A4: Arbitrate the candidate peak points pb[q] at the boundary based on the set of peak degrees, where: When Pleft > T and Pright > T, compare Pleft and Pright, and confirm the boundary candidate peak point corresponding to the one with the larger kurtosis as the arbitrated peak point, and discard the other boundary candidate peak point. When Pleft>T and Pright≤T, the boundary candidate peak points in B[k] are confirmed as the arbitrated peak points; When Pleft≤T and Pright>T, the boundary candidate peak points in B[k+1] are confirmed as the arbitrated peak points; When Pleft≤T and Pright≤T, the boundary candidate peak point pb[q] is removed from the candidate peak points.

6. The high-speed signal acquisition and processing system based on lidar according to claim 5, characterized in that, The dynamic threshold update module updates the dynamic threshold of the adaptive kurtosis calculation method based on the arbitrated peak point, specifically including: Step G1: Count the number of peak points Np and the total number of sampling points Ns after arbitration in the current processing cycle, and calculate the peak density ρ=Np / Ns; Step G2: Compare the peak density ρ with the preset density threshold ρ0, and determine the adjustment direction of the dynamic threshold T based on the comparison result, wherein: When ρ > ρ0, the peak density is determined to be too high, and the dynamic threshold T is adjusted upward. The adjusted dynamic threshold T' = T × (1 + δup) is set, where δup is the adjustment coefficient, 0 < δup ≤ 0.

1. When ρ < ρ0, the peak density is determined to be too low, and the dynamic threshold T is adjusted downward. The adjusted dynamic threshold T' = T × (1 - δdown) is set, where δdown is the adjustment coefficient, 0 < δdown ≤ 0.

1. When ρ=ρ0, the peak density is determined to be normal, and the dynamic threshold T is not adjusted. Step G3: Output the adjusted dynamic threshold T' as the dynamic threshold T for the next processing cycle.

7. The high-speed signal acquisition and processing system based on lidar according to claim 6, characterized in that, The pipeline scheduling module performs multi-level pipeline parallel scheduling for the processes of data loading, peak degree calculation, neighborhood comparison, and result output, specifically including: Step H1: Set the number of pipeline stages R=4, which are data loading stage S1, peak value calculation stage S2, neighborhood comparison stage S3 and result output stage S4. Step H2, in the nth clock cycle, each stage of the pipeline executes simultaneously: Data loading level S1: Load the nth data overlap block B[n] into the cache; Peak degree calculation level S2: Calculate the peak degree of the (n-1)th data overlap block B[n-1]; Neighborhood comparison level S3: Perform neighborhood comparison and candidate peak point marking on the (n-2)th data overlap block B[n-2]; Output stage S4: Outputs the arbitrated peak point of the (n-3)th data overlap block B[n-3]; Step H3 involves passing data at each stage through pipeline registers to achieve uninterrupted processing of continuous data streams.

8. The high-speed signal acquisition and processing system based on lidar according to claim 7, characterized in that, The pipeline scheduling module dynamically adjusts the block division strategy of the data reassembly module according to the pipeline status, specifically including: Step V1: Monitor the processing delays {Δt1,Δt2,Δt3,Δt4} of each stage of the pipeline and calculate the pipeline imbalance index Imbalance=max(Δtr)-min(Δtr), where r∈{1,2,3,4}; Step V2: Compare the pipeline imbalance index Imbalance with the preset imbalance threshold Imbalance0, and determine the need to adjust the segmentation strategy based on the comparison result, wherein: When Imbalance ≤ Imbalance0, the pipeline is considered balanced and no adjustment is made to the block strategy; When Imbalance > Imbalance0, the pipeline is determined to be unbalanced, and the block partitioning strategy is adjusted. The adjustment includes: If Δt1 = max(Δtr), increase the data block length L, and set the adjusted data block length L' = L × (1 + ε), where 0 < ε ≤ 0.2; If Δt2=max(Δtr), reduce the preset overlap length L0, and set the adjusted preset overlap length L0'=L0×(1-ε); If Δt3=max(Δtr), reduce the preset overlap length L0, and set the adjusted preset overlap length L0'=L0×(1-ε); If Δt4 = max(Δtr), reduce the data block length L, and set the adjusted data block length L' = L × (1 - ε).

9. The high-speed signal acquisition and processing system based on lidar according to claim 4, characterized in that, The parallel peak detection module also includes a burst interference suppression unit, used to detect and suppress outliers in the rate of change of sampling points Δx[i], specifically including: Step U1: Calculate the rate of change of the current sampling point Δx[i] = x[i] - x[i-1]; Step U2, obtain the historical rate of change set ΔXhistory={Δx[i-1],Δx[i-2],…,Δx[iW]}, where W is the length of the historical window, W≥3; Step U3: Calculate the historical rate of change mean μΔ and the historical rate of change standard deviation σΔ; Step U4: Compare |Δx[i]| with the burst interference threshold Thburst=μΔ+k×σΔ, where k is the confidence coefficient and k≥3; Step U5: Dynamically adjust the rate of change inhibition coefficient β based on the comparison results, where: When |Δx[i]|≤Thburst, it is determined to be a normal change, and β=βbase is set; When |Δx[i]|>Thburst, it is determined to be a sudden disturbance, and β=βburst is set, where βburst>βbase.

10. A method applied to a high-speed signal acquisition and processing system based on lidar as described in any one of claims 1-9, characterized in that, include: Step J1: Receive and buffer the raw sampling data from the lidar; Step J2: The original sampling data of the lidar is recombined by overlapping blocks according to the preset overlap length to obtain several data overlap blocks, wherein the data overlap blocks include internal data points and boundary data points; Step J3: Perform parallel peak detection on each data overlap block according to the adaptive kurtosis calculation method to obtain candidate peak points for each data overlap block. The candidate peak points include internal candidate peak points and boundary candidate peak points. Step J4: Perform conflict arbitration on the candidate peak points of the boundary of adjacent overlapping data blocks to obtain the arbitration peak points; Step J5: Update the dynamic threshold of the adaptive kurtosis calculation method based on the peak point after arbitration; Step J6 involves performing multi-level pipeline parallel scheduling for the processes of data loading, peak degree calculation, neighborhood comparison, and result output, and dynamically adjusting the block strategy based on the pipeline status.

Images