机载光电探测系统定位误差分析与标定研究

	机载光电探测系统定位误差分析与标定研究
其他题名	Research on Error Analysis and Calibration of Airborne Photoelectric Detecting System
	罗安治
学位类型	硕士
导师	贾宏光
	2015-11
学位授予单位	中国科学院大学
学位专业	机械制造及其自动化
关键词	光电探测系统定位误差误差建模指向误差参数标定误差修正
摘要	光电探测系统作为现代探测、成像系统的核心部件，主要作用是对目标的识别、定位、测量、跟踪。由于光机电一体化学科的快速发展，光电探测系统已普遍运用于侦测、定位、瞄准、跟踪、导航等各领域。定位精度作为系统工作性能的决定性评价指标，对于实现定位、成像、跟踪等工作具有重要意义，不仅关系图像的观测质量，还会影响到后续跟踪、瞄准的准确性。本文针对光电探测系统定位精度问题，分析了工作机理、总体结构与误差来源，对目标定位误差进行分析、建模、评价，并标定指向误差模型、修正指向误差。论文主要研究内容如下：一、总结了机载光电探测系统目标定位过程的原理、结构组成与综合性能评价指标。按照光电探测系统结构特征与目标定位流程，详细分析了目标定位误差中各分项误差的误差来源与分布规律。二、在总体结构与分项误差来源分析的基础上，需要建立关于机载光电探测系统目标定位过程的运动学综合模型。首先，将光电探测系统这个多轴复杂机械结构抽象成一个多体系统，利用多体系统理论，划分连续坐标系，按多体系统方法得到由若干齐次坐标变换矩阵组成的相邻坐标系间D-H（Denavit-Hartenberg）变换模型，进而推导出目标定位坐标的D-H基本参数模型。然后，若不考虑光电探测系统的具体结构，按照目标定位的大概运动过程，建立以李群、旋量理论为理论基础的局部指数积（Local POE）模型。同时，根据分项误差源分析和D-H基本参数模型，使用Monte Carlo方法与误差灵敏度公式，计算各分项误差灵敏度系数，对系统设计、装配与定位误差算法修正起到指导作用。三、为了达到高精度目标定位要求，需要修正定位误差中占主导作用的指向误差部分。首先，总结了光电探测系统指向误差的修正原理、研究流程、技术问题、实际应用领域。接着，基于传统线性误差因素的考虑，建立基本参数模型，使用最小二乘法标定模型参数。然后，基于线性误差因素与非线性误差因素共存的情况，建立半参数模型，使用补偿最小二乘法与双三次样条插值相结合的标定算法标定模型参数。最后，针对机载光电探测系统指向误差Local POE模型，改进现有智能算法，融合模拟退火与粒子群算法，并标定Local POE模型。四、搭建实验平台，针对某型光电探测系统进行指向误差分离、标定与修正实验，分离指向误差得到样本数据。分别使用最小二乘标定算法、补偿最小二乘法与双三次样条插值相结合的标定算法、基于模拟退火的粒子群优化算法分别标定光电探测系统指向误差基本参数模型、半参数模型与Local POE模型的待定参数。实验结果表明，各标定算法都能显著提高光电探测系统指向精度与稳定性。
其他摘要	Photoelectric detecting system is an important part of the modern detecting and reconnaissance system, and it is mainly used for recognition, orientation, detection and tracing. With the development of optical mechanical and electronic integration system, photoelectric detecting system has been widely used in the field of detecting, orientation, collimation, tracing and navigation. As the decisive evaluation index of the system performance, orientation precision is important to the photoelectric detecting system. Not only influence the observational quality of image, but also concern the precision of tracing and collimation. In order to improve the orientation precision of photoelectric detecting system, the paper analyses the working principle, overall structure, error sources, model theories, evaluations, then calibrates the model, and corrects the pointing error. The following studies are included in this paper: 1. Summarize the working principle, overall structure and evaluation index of orientation error. According to the structure feature and the flow of orientation, analyses the error sources and distribution rules of all orientation error in detail. 2. Based on the analysis of overall structure and orientation error, the orientation error kinematic model of photoelectric detecting system needs to be established. Firstly, regard the photoelectric detecting system as a complex multi-body system, and divide the continuation coordination system. Establish the Denavit-Hartenberg basic parametric model according to the homogenous coordinate transformation which is obtained by multi-body system theory. Secondly, if the overall structure is unknown, according to the process of orientation, establish the Local POE model by lie group theory and screw theory. Finally, based on the error analysis and basic parametric model, calculate the error sensitivity coefficient by Monte Carlo method and sensitivity analysis, which can tell the proportion of each error source in target orientation error, and the coefficient can direct the correction of orientation error. 3. In order to meet the high-precision target of the photoelectric detecting system, pointing error which takes up the chief proportion in orientation error needs to be corrected. Firstly, analyse the correcting principle, process and key problems. Secondly, in consideration of the traditional linear error factors, establish the basic parametric model, and calibrate the parameters by least square calibration algorithm. Then, in consideration of the linear and nonlinear error factors, semi-parametric model is established, and calibrate the parameters by penalized least square calibration algorithm and bicubic spline interpolation. Finally, optimize the traditional algorithm, and calibrate the Local POE model by particle swarm optimization algorithm based on simulated annealing. 4. A prototype experimental platform is built. Sample data of pointing error are obtained after pointing error separation, calibration and correction experiments. Calibrate the basic parametric model, semi-parametric model and Local POE model with least square calibration algorithm, penalized least square calibration algorithm and particle swarm optimization algorithm based on simulated annealing, and correct the pointing error of photoelectric detecting system by calibrated model. The experiments and results make it clear that all algorithms can improve the pointing precision and pointing stability of photoelectric detecting system.
语种	中文
文献类型	学位论文
条目标识符	http://ir.ciomp.ac.cn/handle/181722/49246
专题	中科院长春光机所知识产出
推荐引用方式 GB/T 7714	罗安治. 机载光电探测系统定位误差分析与标定研究[D]. 中国科学院大学,2015.

条目包含的文件
文件名称/大小	文献类型	版本类型	开放类型	使用许可
罗安治.pdf（2321KB）	学位论文		开放获取	CC BY-NC-SA	请求全文