【作者】 唐康华；

【导师】 黄新生；

【作者基本信息】 国防科学技术大学 ， 控制科学与工程， 2008， 博士

【摘要】 现代战争的高技术特点表明,精确制导武器已经逐渐成为现代战争的主导,低成本导航与制导技术是消耗型制导武器的关键技术之一。卫星/MIMU嵌入式组合导航系统具有成本低、体积小、重量轻、功耗低、动态性能高和抗干扰能力强等优点,因而在精确制导武器等军事应用方面具有非常重要的应用价值。本文以GPS为实例,开展卫星/MIMU嵌入式组合导航一体化关键技术的研究,完成了MIMU辅助的GPS信号捕获、MMU辅助的GPS信号跟踪、MIMU辅助的导航定位算法和GPS/MIMU深组合导航算法的分析,并给出了具体解决方案。主要的工作包括以下几个方面：(1)从卫星信号的检测概率、虚警率和捕获时间方面对时域相干和非相干捕获算法进行了分析比较：非相干捕获算法不受导航数据位的影响,而且多普勒频率误差对其影响较小,但是非相干捕获算法增加了平方损失；而相干捕获算法获取相同的捕获灵敏度,仅需较短的积分时间,但其受多普勒频移估计偏差和码移估计偏差的影响较大,并且受导航数据位的影响。基于时域标准相干和非相干捕获算法的特点,针对快速捕获的应用要求,提出了基于辅助的(卫星历书和星历辅助、卫星／MIMU估计的多普勒频率辅助、位置和时间的辅助)时域相干捕获算法,分析结果表明该方法能有效地缩短捕获时间(从几百秒缩短到几秒)和提高捕获效率(在检测概率为95%的条件下,对GPS信号的捕获能力提高了4.5dB／Hz)。(2)对接收机DLL/PLL/FLL的鉴别器进行了分析和比较研究,在综合考虑接收机跟踪环路中的各种误差源(热噪声、晶振误差、动态牵引误差等)的基础上,设计了一种适合于高动态的GPS环路结构：码跟踪环采用载波辅助的结构,载波环路采用双模态结构——2阶FLL辅助的3阶PLL结构。针对不同的晶振类型,综合考虑各种误差因素,设计了MIMU辅助的GPS接收机环路结构及MIMU辅助3阶PLL的最小带宽和最优带宽。根据设计的MIMU辅助的GPS环路结构,分析了其抗干扰性能,分析结果表明MIMU辅助的卫星接收机比一般的GPS接收机抗干扰性能至少有1 1dB的提高。(3)分析了基于Kalman滤波器的纯GPS导航定位算法,为了减少系统的状态维数,观测量采用伪距单差和伪距率单差(星间差)来抵消用户钟差和钟差漂移的影响,试验结果表明Kalman滤波导航算法使得导航定位结果更平滑,定位精度更高；同时分析了MMU辅助的GPS导航定位算法,观测量采用伪距单差和伪距率单差来抵消用户钟差和钟差漂移的影响；针对低成本MIMU系统,设计了一种基于最大似然估计的自适应滤波算法,通过试验验证了该算法,静态试验结果表明：MIMU辅助GPS定位系统的位置精度优于5m,速度精度优于0.1m／s,俯仰角和滚动角精度优于0.2°,航向角精度优于0.2°(航向角辅助)。(4)为了提高系统的动态性能和抗干扰能力,论文进一步研究了GPS/MIMU深组合导航算法。首先论述了GPS/MMU嵌入式深组合导航算法的基础——矢量跟踪环结构,着重分析了三种基带测量信号预处理方法；接着分析了GPS/MIMU嵌入式深组合导航算法中的导航滤波器设计,为了提高导航滤波性能,采用了降维的导航滤波算法；最后针对基带信号预滤波的强非线性问题,提出了基于UKF的基带测量信号预滤波方法,并对其性能进行了分析比较。(5)论文最后给出了GPS/MIMU嵌入式导航系统的实现方案及试验结果,客观评价了系统的关键性能指标,进一步验证了论文提出的有关算法模型。系统主要性能指标为：1)在GPS/MIMU嵌入式组合导航系统中,当有星历辅助时GPS首次定位时间优于10s。2)在静态条件下,研制的GPS接收机的位置精度优于5m,速度精度优于0.1 m／s；在动态车载条件下,水平位置误差优于3米,高程优于10米,与商用GPS接收机精度大体相当。3) GPS/MIMU嵌入式组合导航系统在静态条件下的精度重复性为：位置精度优于5m,速度精度优于0.1m／s,俯仰角和滚动角精度优于0.2°,航向角精度优于0.2°(航向角辅助)。论文虽以GPS为实例,但是研究的成果同样适用于其它的卫星导航系统(如GLONASS、GALLIEO、北斗Ⅱ卫星导航系统等)。更多还原

【Abstract】 High-tech modern wars declare that precision guided munitions have made a significant impact in recent armed conflicts, and the low-cost technology of navigation and guidance unit is a key in the consuming guided munitions. Having such advantages as high precision, low-cost, low-size, low-weight, low-power, powerful capability of anti-jamming and high dynamics, et al., embedded GPS/MIMU integrated navigation systems have extensive application value in the projectile and precision guided munitions. This dissertation addresses key technologies of embedded GPS/MIMU integration such as MIMU-aided GPS signal acquisition, MIMU-aided GPS signal tracking, MIMU-aided GPS navigation, GPS/MIMU ultra-tightly coupled navigation, proposing a full solution for the embedded GPS/MIMU integrated navigation system. The main contributions include the following aspects:(1) Benefits and drawbacks of non-coherent and coherent integration are analyzed from the aspects of probability of detection, probability of false alarm and the acquisition search time. Coherent integration requires a shorter integration time to achieve the same acquisition sensitivity versus a comparable non-coherent integration. Non-coherent integration is more tolerant to residual frequency errors and is not affected by the navigation data bits. According to their benefits and drawbacks, aided GPS signal coherent integration acquisition(ephemeris and almanac aiding, time and approximate user position aiding, GPS/MIMU estimated doppler aiding) is presented for the fast acquisition applications. The simulated results show that performance of the aided acquisition method can improve 4.5dB/Hz in the condition of 95% probability of detection than non-aiding GPS signal acquisition, and the time of aided acquisition is reduced from several hundred seconds to several seconds.(2) In order to obtain the best DLL/PLL/FLL performance, we select discriminators and their normalization algorithms with the highest processing gain and the least linearization effect. Considering the total tracking errors such as thermal noise, oscillator phase noise, dynamic stress error, et al., we design the structure of GPS receiver tracking loop for high dynamic applications. Doppler frequency from carrier is proportional to that from code, so Doppler information from the carrier tracking loop is selected to aid the code tracking loop. Dual mode structure is adopted as the high dynamic GPS receiver carrier tracking loop. One is a FLL-assisted-PLL mode for high dynamics, and the other is a PLL mode for low dynamics. The mode is switched according to the lock status.Considering the total tracking errors and different oscillators, the structure of MIMU-aided GPS receiver tracking loop is also designed. The structure of Doppler information from the carrier aided code tracking loop is adopted. The optimal loop bandwidth of MIMU-aided PLL is designed based on different oscillators such as a 1.1 Hz loop bandwidth of the MIMU-aided 3rd order PLL for OCXO, and 2.14Hz for TCXO. Finally according to the designed MIMU-aided GPS receiver tracking loop, the anti-jamming simulation is performed. Simulation results show that anti-jamming performance of the MIMU-aided receiver has 11dB improvement at least, compared to traditional GPS receiver.(3) The GPS navigation filter is investigated. In order to eliminate the clock components, that is, the receiver clock offset and clock offset rate, the pseudorange measurements and Doppler measurements are differenced across satellites. Compared to the least square algorithm, the navigation filter adopted can improve the positioning precision. According to the GPS navigation filter, the MIMU-aided GPS navigation filter is designed, in which the measurements are the pseudorange and Doppler difference across satellites. Also the structure of the adaptive Kalman filter is designed for low-cost, low-precision MEMS IMU. Finally, Test results show that using adaptive Kalman filter, perfect performance can be achieved for the low-cost MIMU aided GPS navigation so that the positioning accuracy is better than 5m, velocity accuracy is better than 0.1 m/s on the static condition, the level attitude errors are better than 0.2°, heading attitude error is better than 0.2°(with heading aiding).(4) In order to improve the performance of dynamics and anti-jamming, the dissertation presents a GPS/MIMU ultra-tightly coupled navigaton algorithm. The vector-based tracking structure of GPS receiver which is the foundation of GPS/MIMU ultra-tightly coupled navigaton algorithm is introduced firstly. Three local pre-processing algorithms for GPS correlated sample signals are presented for GPS vector-tracking architecture:least mean square curve fitting, linear Kalman pre-filter and non-linear Kalman pre-filter. The GPS/MIMU ultra-tightly coupled navigaton algorithm can be achieved by replacing the GPS navigaton filter in GPS vector-tracking structure with an integrated GPS/MIMU navigation filter. The local pre-processing algorithm is used to update an GPS/MIMU navigation filter. The navigation solution together with GPS parameter corrections are used in a tracking predictor to generate high-sampling-rate carrier and code replicas. The structure of reduced-dimension Kalman filter is adopted for the GPS/MIMU navigation filter in order to improve the performacne of GPS/MIMU navigation filter. Finally, regarding the strong nonlinearity of GPS corrlelator output signal, UKF is first presented for GPS local pre-processing algorithm, and the performance of GPS local pre-processing algorithm using UKF is evaluated.(5) Test methods and results are presented and the system performance evaluation are given. The main performances are as follows:1) In the embedded GPS/MIMU integrated navigation system, TTFF(Time To First Fix) of ephemeris aiding GPS is better than 10s. 2) The GPS positioning accuracy is better than 5m and velocity accuracy is better than 0.1 m/s on the static condition. On the dynamic condition, the accuracy of the designed GPS receiver is comparable to that of commercial GPS receiver.3) For the embedded GPS/MIMU integrated navigation system on the static condition, the positioning accuracy is better than 5m, velocity accuracy is better than 0.1 m/s, and the attitude errors are less than 0.2 degrees with yaw aiding.The contribution of this dissertation is not only applicable for GPS, but also for other GNSS such as GLONASS, GALLIEO, BDⅡ.更多还原