CN-122017861-A - Quick high-precision scanning type laser fuze data processing method

Abstract

The invention relates to a rapid high-precision scanning type laser fuze data processing method, and belongs to the technical field of laser detection and fuze control. According to the method, echo signals of the line scanning laser radar are received, and after shaping, multi-echo screening, distance calculation, speed correction and elevation conversion, an improved one-dimensional clustering algorithm is adopted to rapidly analyze elevation data, and the target elevation and the confidence level thereof are output for predicting detonation time. The method abandons the traditional three-dimensional point cloud reconstruction process, has the advantages of high processing speed, strong anti-interference capability, less resource occupation and the like, and is suitable for real-time target detection and fuze control of the high-dynamic missile-borne platform in a complex environment.

Inventors

• XU SHIYUE
• Zhao Quanxiao
• ZHANG GUOJUAN
• Ran Chengyao
• YANG QIN
• MA RUI
• YANG ZHENYUAN

Assignees

• 西南技术物理研究所

Dates

Publication Date

20260512

Application Date

20251223

Claims (10)

1. 1. The rapid high-precision scanning type laser fuze data processing method is characterized by comprising the following steps of: s1, shaping an echo signal received by a line scanning laser radar, and extracting multi-echo pulses; S2, screening effective echo data from the multi-echo pulse; s3, calculating a target distance value based on the effective echo data; s4, correcting the distance value according to the projectile velocity and the flight angle, and converting the distance value into elevation data; s5, carrying out cluster analysis on all elevation data in one frame, and outputting a final height Cheng Jieguo and the confidence level thereof.
2. 2. The method of claim 1, wherein the echo signal shaping comprises filtering, thresholding to obtain a near gaussian shaped echo signal.
3. 3. The method of claim 1, wherein the valid echo data comprises at least one of a first echo, a last echo, and a strongest echo.
4. 4. The method according to claim 1, wherein the distance value calculation adopts a quadratic curve inflection point fitting formula, and the distance measurement precision is +/-2 cm under the conditions that the laser pulse width is 6-10ns and the sampling rate is 1 GHz.
5. 5. The method of claim 1, wherein the velocity correction comprises time compensating the distance value based on the projectile flight velocity and the scanning time stamp.
6. 6. The method of claim 1, wherein the cluster analysis employs a modified one-dimensional DBSCAN algorithm with low computational overhead, supporting parallel processing, and single frame processing times of no more than 10 milliseconds.
7. 7. The method of claim 1, wherein the cluster analysis output comprises a center point elevation, a data amount, a total number of clusters, and a confidence level for each cluster.
8. 8. The method according to any of claims 1-7, wherein the method is implemented in an FPGA or MCU embedded platform supporting parallel or pipelined processing.
9. 9. The method of claim 8, wherein the elevation data is obtained by projecting distance values in a direction perpendicular to the direction of flight of the projectile.
10. 10. The method of claim 8, wherein in the one-dimensional DBSCAN algorithm, the distance calculation uses absolute differences, and the cluster radius and the minimum number of samples are dynamically set according to the elevation data distribution.

Description

Quick high-precision scanning type laser fuze data processing method Technical Field The invention belongs to the field of laser detection and fuze control, and particularly relates to a rapid high-precision scanning type laser fuze data processing method. Background Traditional laser fuzes mostly adopt TOF mechanism ranging, and are assisted with point cloud generation and target modeling technology, so that accurate detection and identification are realized. However, the laser fuze based on single-point ranging is easy to be influenced by terrain due to small ranging area, and can not accurately judge the detonation moment in an area with complex ground conditions, and continuous frame data splicing and complex algorithm processing are often required for acquiring complete three-dimensional point clouds, such as point cloud classification and detection networks based on PointNet, PV-RCNN and the like, so that higher requirements are put on the calculation resources, power consumption and response speed of the missile-borne platform. Meanwhile, under the influence of interference factors such as rain, fog, smoke dust and the like in a complex battlefield environment, the laser reflection signals can be greatly fluctuated, the noise points in the final point cloud are more, and the misidentification situation is frequent. On the other hand, in the laser fuze application scenario, the complete target shape is often not required to be reconstructed, but the actual optimal detonation opportunity prediction is taken as a primary target. Therefore, the efficient time sequence analysis and trend prediction are developed only based on the distance information, the functional closed loop can be realized without point cloud modeling, and the stability and instantaneity of the system are obviously improved. At present, a lightweight processing algorithm and a robustness enhancing mechanism facing the requirement are not available, and become one of key bottlenecks for restricting laser fuze actual combat deployment. Disclosure of Invention The invention provides a rapid high-precision scanning type laser fuze data processing method, which takes a processing result as the output of a laser fuze and is used for calculating the detonation time. In order to solve the technical problems, the invention provides a rapid high-precision scanning type laser fuse data processing method, which is characterized by comprising the following steps: s1, shaping an echo signal received by a line scanning laser radar, and extracting multi-echo pulses; S2, screening effective echo data from the multi-echo pulse; s3, calculating a target distance value based on the effective echo data; s4, correcting the distance value according to the projectile velocity and the flight angle, and converting the distance value into elevation data; s5, carrying out cluster analysis on all elevation data in one frame, and outputting a final height Cheng Jieguo and the confidence level thereof. Further, the echo signal shaping includes filtering and threshold segmentation to obtain a near Gaussian echo signal. Further, the effective echo data comprises at least one of a first echo, a last echo and a strongest echo. Further, the distance value is calculated by adopting a quadratic curve inflection point fitting formula, and the distance measurement precision is +/-2 cm under the conditions that the laser pulse width is 6-10ns and the sampling rate is 1 GHz. Further, the speed correction comprises the step of performing time compensation on the distance value according to the projectile flying speed and the scanning time stamp. Further, the cluster analysis adopts an improved one-dimensional DBSCAN algorithm, so that the calculation cost is low, parallel processing is supported, and the single-frame processing time is not more than 10 milliseconds. Further, the cluster analysis output comprises a central point elevation, a data volume, a total number of clusters and a confidence level of each cluster. Further, the method is realized by an FPGA or MCU embedded platform, and parallel or pipeline processing is supported. In the one-dimensional DBSCAN algorithm, absolute difference is adopted in distance calculation, and clustering radius and minimum sample number are dynamically set according to the distribution of the elevation data. The method has the beneficial effects that the method can realize target approximation detection, anti-interference processing and ideal detonation point calculation only based on the distance information obtained by linear array laser scanning without generating a three-dimensional point cloud, thereby improving the environment adaptability and response speed of the fuze. The invention has the following advantages: And the point cloud data construction is not needed, and the calculation cost is low. According to the invention, data processing is only carried out based on the distance information obtained by linear array laser scan