交互多模型在车辆组合定位系统中的应用（英文）

为解决扩展卡尔曼滤波器(extended Kalman filter,EKF)在车辆组合定位系统中因车辆加减速、转弯(以下简称机动)而存在的精度低、稳定性差等问题,设计了一种将交互多模型(interacting multiple model,IMM)算法与非线性卡尔曼滤波器相融合的自适应滤波算法。该算法使用三种状态空间模型来描述车辆的运动模式,采用多个非线性滤波器对每个模型并行滤波,通过模型匹配似然函数对滤波结果进行加权融合,最终得到系统的定位信息。该方法具备非线性系统滤波器优点,克服了单一模型滤波算法对机动目标定位效果差的缺点。利用该方法和EKF算法分别对GPS/INS/DR车辆组合定位系统中进行了仿真实验,结果表明,该算法的滤波定位精度明显优于目前组合定位系统中所用的EKF滤波器,大幅提高了组合定位系统的稳定性和定位精度。     

Novel hybrid of strong tracking Kalman filter and wavelet neural network for GPS/INS during GPS outages

Aiming to improve positioning precision of the GPS/INS integrated navigation system during GPS outages, a novel model combined with strong tracking Kalman filter (STKF) and wavelet neural network (WNN) algorithms for INS errors compensation is proposed and tested. STKF is used to estimate INS errors as a replacement of Kalman filter (KF), and WNN is applied to establish a highly accurate model based on STKF when GPS works well and to predict INS errors during GPS outages. Performance of the proposed model has been experimentally verified using GPS and INS data collected in a land vehicle navigation test. The comparison results indicate that the proposed model combined with STKF/WNN algorithms can effectively provide high accurate corrections to the standalone INS during GPS outages.     

水下潜航器的惯导/超短基线/多普勒测速信息融合及容错验证

针对水下潜航器惯导系统的定位误差积累和容错性差等问题,分析了水声超短基线的相位差定位方法,推导了基于惯导提供实时位置、姿态误差角信息的惯导/超短基线(INS/USBL)导航解算过程及其坐标转换。结合惯导/多普勒测速(INS/DVL)滤波器,给出INS/USBL/DVL组合导航联邦滤波在3种信息融合算法下的应用。通过MATLAB仿真对导航算法进行了验证,结果表明该导航算法能够抑制惯导系统误差随时间发散的问题,能充分利用了3种导航系统提供的参数信息,且状态维数低,滤波收敛速度快,其中基于精度因子信息分配方法的导航系统误差最小。容错性验证结果显示,当超短基线出现故障时,重构后的组合导航系统在较高航速情况下依旧能提供有效的导航参数。所提出的INS/USBL/DVL组合导航联邦滤波方法能够精确地提供水下潜航器的各位导航参数信息,且具有较高的容错性和稳定性。     

Outdoor position estimation based on a combination system of GPS-INS by using UPF

This paper proposes a technique that global positioning system(GPS)combines inertial navigation system(INS)by using unscented particle filter(UPF)to estimate the exact outdoor position.This system can make up for the weak point on position estimation by the merits of GPS and INS.In general,extended Kalman filter(EKF)has been widely used in order to combine GPS with INS.However,UPF can get the position more accurately and correctly than EKF when it is applied to real-system included non-linear,irregular distribution errors.In this paper,the accuracy of UPF is proved through the simulation experiment,using the virtual-data needed for the test.     

Improving unmanned ground vehicle navigation by exploiting virtual sensors and multisensor fusion

This paper investigates techniques on improving navigation accuracy using multiple sensors mounted on a mobile platform and exploiting the inherent characteristic of a ground vehicle that does not move along the cross-track and off the ground, often termed nonholonomic constraints (NHC) for car-like vehicles that assume no slip or skid. The forward velocity of the vehicle is obtained using a wheel encoder. The 3D velocity vector becomes observable during the normal moving state of the vehicle by using NHC, which produces one virtual sensor. Another virtual sensor is the zero-velocity update (ZVU) condition of the vehicle; when the condition is true, the 3D velocity vector (which is zero) becomes observable. These observables were employed in an extended Kalman filter (EKF) update to limit the growth of inertial navigation system error. We designed an EKF for data fusion of inertial measurement units, global positioning systems (GPS), and motion constraints (i.e., NHC and ZVU). We analyzed the effects of utilizing these constraints on improving navigation accuracy in stationary and dynamic cases. Our proposed navigation suite provides reliable accuracy for unmanned ground vehicle applications in a GPS-denied environment (e.g., forest canopy and urban canyon).     

An adaptive navigation method for a small unmanned aerial rotorcraft under complex environment

Focusing on low navigation performance of small unmanned aerial rotorcraft under complex environment, a composite navigation method combined with adaptive Kalman filtering and radial basis function neural network prediction method is proposed to improve navigation performance during GPS outages. When the GPS signal is available, an adaptive Kalman filter based on covariance scaling is introduced to deal with the process noise in real time. Meanwhile, a radial basis function neural network is trained on line to construct the projection among input (output of the inertial measurement unit, attitude and GPS losing time) and output (position error and velocity error). During GPS outages, the radial basis function neural network can provide high performance error estimation for position and velocity to improve state information. Finally, a land vehicle test and a flight test have confirmed that the proposed method can improve the navigation performance largely under complex environment.     

SINS/LBL组合导航序贯滤波方法

在长基线水声定位系统(LBL)实际应用过程中,由于信标作用距离限制、障碍物遮挡等多种原因,使得水下无人航行器(UUV)可能无法接收到所有信标的应答信号而产生量测更新延迟问题。对UUV在无法接收到所有信标信号时的导航滤波算法进行了研究;结合捷联惯性导航系统(SINS)误差模型和声速误差,建立了SINS/LBL组合导航模型,并将异步量测序贯处理方法引入该模型,实现了SINS/LBL组合导航滤波算法的实时量测更新;通过湖上试验数据分析,对比了SINS/LBL组合导航序贯滤波方法与常规方法的位置误差。结果表明,该方法即使在应答信号有缺失的情况下,仍然能够利用有限的应答信号量测值进行实时量测更新,保障了组合导航的精度。     

An approach to improve the GPS positioning performance under urban environment conditions

Satellite deficiency degrades the performance of GPS single point positioning (SPP) in city urban areas. If the number of observed satellites is smaller than that of the unknown independent parameters in the data processing, extended Kalman filter (EKF) methods will usually be employed to transit the variable values epoch by epoch. In this case, the prediction model in EKF is very important for the positioning results. This paper focuses on the receiver clock offset models in the prediction equation. The investigation shows that, the clock offset model with clock shift is able to improve the accuracy of the clock offset prediction for internal crystal clocks within a short time span, compared with the clock offset model without clock shift. Besides, in order to further improve the performance of positioning in the case of satellite deficiency, an adaptive EKF is developed to enhance the clock prediction. The experiments show that this approach can improve the positioning accuracy in the scenarios that only fewer satellites are available.     

High accuracy navigation information estimation for inertial system using the multi-model EKF fusing adams explicit formula applied to underwater gliders

The underwater navigation system, mainly consisting of MEMS inertial sensors, is a key technology for the wide application of underwater gliders and plays an important role in achieving high accuracy navigation and positioning for a long time of period. However, the navigation errors will accumulate over time because of the inherent errors of inertial sensors, especially for MEMS grade IMU (Inertial Measurement Unit) generally used in gliders. The dead reckoning module is added to compensate the errors. In the complicated underwater environment, the performance of MEMS sensors is degraded sharply and the errors will become much larger. It is difficult to establish the accurate and fixed error model for the inertial sensor. Therefore, it is very hard to improve the accuracy of navigation information calculated by sensors. In order to solve the problem mentioned, the more suitable filter which integrates the multi-model method with an EKF approach can be designed according to different error models to give the optimal estimation for the state. The key parameters of error models can be used to determine the corresponding filter. The Adams explicit formula which has an advantage of high precision prediction is simultaneously fused into the above filter to achieve the much more improvement in attitudes estimation accuracy. The proposed algorithm has been proved through theory analyses and has been tested by both vehicle experiments and lake trials. Results show that the proposed method has better accuracy and effectiveness in terms of attitudes estimation compared with other methods mentioned in the paper for inertial navigation applied to underwater gliders.     

基于环境特征跟踪的移动机器人定位

提出了一种基于环境特征跟踪来实现移动机器人定位的方法。对传感器、环境观测和机器人的运动建立了相应的模型，并以扩展卡尔曼滤波技术将多种传感器的信息进行融合，从而最终实现了移动机器人的精确定位。     

基于目标特征提取和EKF的位置跟踪方法

针对设备对接等特殊环境对导航定位系统的需求,提出了一种新的基于激光测距传感器进行位置跟踪的方法,主要包含激光雷达成像特征匹配和EKF滤波跟踪两部分,相比传统的雷达与视觉定位具备高精高速等特点。在搭建的实验平台上通过进行室内跟踪实验,最终验证了该方法的有效性。     

A new method of seamless land navigation for GPS/INS integrated system

For the last few years, integrated navigation systems have been widely used to calculate positions and attitudes of vehicles. The strapdown inertial navigation system (SINS) provides velocity, attitudes and position information, whereas the global positioning system (GPS) provides velocity and position information. A method using neural network (NN) and wavelet-based de-noising technology is introduced into the SINS/GPS/magnetometer integrated navigation system, because system accuracy may decrease during GPS outages. When the GPS is working well, NN is trained using the velocity and position information provided by SINS as input and the corresponding errors as output. Wavelet multi-resolution analysis (WMRA) is also introduced to de-noise the errors, the desired output of NN. Test results showed that velocity accuracies improved using NN, but other accuracies remained poor. By re-training NN with WMRA, the system accuracies improved to the level of using normal GPS signal. In addition, NN trained with WMRA also improved the approximation to the actual model, further enhancing alignment accuracy.     

INS/SFGPS PPP紧组合系统动态补偿滤波算法

惯性/卫星组合导航系统结合精密单点定位技术可有效提高导航定位精度。但精密单点定位技术一般需采用双频接收机,成本较高;同时该系统中采用载波相位作为部分或全部观测量,极容易受到周跳的影响而导致精度下降和系统不稳定。针对上述问题,设计了一种惯性/卫星精密定位紧组合导航系统以及基于动态周跳补偿的鲁棒滤波算法。该系统采用低成本的单频接收机(SFGPS),以精密单点定位技术(PPP)处理过的伪距和载波相位作为观测信息,与惯性导航系统(INS)等效观测量进行紧组合,建立了相应紧组合观测模型并引入周跳作为信息融合滤波状态模型中的状态量,以滤波器信息构建周跳检测统计量并对周跳幅值进行识别和估计,实时补偿观测量以提高观测信息精度,同时以前述周跳估计的结果对状态模型中周跳状态量部分滤波参数进行实时调节。上述方法通过动态补偿周跳误差提高导航精度,通过滤波器参数自适应调节提高滤波稳定性。仿真结果验证了该系统模型及算法的有效性。     

基于视线测量的航天器相对导航滤波方法研究

基于视线测量的航天器相对导航精度会受到相对轨迹形状和滤波算法设计等因素的共同影响。以低轨卫星近距离编队飞行为任务背景,设计了环航飞行、共面漂移和共线保持3种不同轨迹的相对运动模式。对3种模式建立了基于星间非线性相对运动模型的系统状态方程,并引入了J2项地球非球形摄动力的影响;建立了基于视线测量的观测方程,观测量包括星间相对距离、相对俯仰角和相对航向角。结合系统模型和观测模型均为高斯分布的非平稳随机过程的特点,分别在上述3种模式下设计了基于扩展卡尔曼滤波(extended Kalman filter,EKF)和无迹卡尔曼滤波(unscented Kalman filter,UKF)的相对导航滤波算法,对各自的相对运动轨迹进行了数值仿真,并在半物理硬件环境下进行了验证,分析了不同模式下EKF和UKF对于高斯非平稳随机过程的估计精度和稳定性,并结合EKF和UKF的运算复杂度,提出了3种相对运动模式下的滤波器优选方案,对工程设计提供了理论参考。     

Initial orientation of inertial navigation system realized through nonlinear modeling and filtering

Alignment is the process whereby the orientation of the axes of an inertial navigation system is determined with respect to the reference system. In this paper, the initial alignment error equations of the strapdown inertial navigation system (SINS) with large initial azimuth error have been derived with inclusion of nonlinear characteristics. Simulations have been carried out to validate and corroborate the stationary alignment case employing a strapdown inertial measurement unit (SIMU). A performance comparison between the extended Kalman filter (EKF), the unscented Kalman filter (UKF) and the second-order divided difference filter (DDF2) demonstrate that the accuracy of attitude error estimation using the DDF2 is better than that of using the EKF or the UKF.     

基于高斯混合容积卡尔曼滤波的UUV自主导航定位算法

针对过程噪声为非理想高斯分布时无人水下航行器(UUV)自主导航定位存在噪声模型失配的问题,将高斯混合密度模型与容积卡尔曼滤波(CKF)相结合,设计了基于高斯混合容积卡尔曼滤波(GM-CKF)的UUV导航定位算法。建立了UUV运动模型及观测模型,利用CKF完成各高斯分量的预测更新,并将更新结果进行融合缩减与加权求和,从而实现UUV自主导航定位。通过与EKF、UKF和CKF算法仿真对比实验,验证了GM-CKF可以提高估计精度;通过UUV湖试试验,验证了基于GM-CKF的UUV自主导航定位精度和稳定性优于传统算法,其计算时间满足实时导航定位的要求。     

基于鱼类侧线感知机理的流场辨识方法及仿真研究

在研究鱼类侧线感知机理的基础上,综合运用了流体动力学,边界层理论,耦合仿生学等理论,建立类似于鱼神经丘模型的神经网络模型,从数值计算和仿真分析两方面对鱼体侧线系统进行模拟,并将研究结论应用于自治水下机器人(Autonomous underwater vehicle,AUV)的导航及目标定位,为AUV导航和环境辨识提供新思路。计算机仿真结果表明,基于鱼类侧线感知机理构建的感知模型可以有区分度地对一个新的水流工况进行识别,可为AUV的导航及目标定位提供支持。     

An attitude determination algorithm for a spacecraft using nonlinear filter

In this paper, the algorithm for a real time attitude estimation of a spacecraft motion is investigated. The proposed algorithm for attitude estimation is the second order nonlinear filter form not containing truncation error in estimation values. The proposed second order nonlinear filter has improved performance compared with the EKF (extended Kalman filter), because the algorithm does not contain any truncation bias and covariance of the estimator is compensated by the nonlinear terms of the system. Therefore, the proposed second order nonlinear filter is a suboptimal estimator. However, the proposed estimator requires a lot of computation because of an inherent nonlinearity and complexity of the system model. For more efficient computation, this paper introduces a new attitude estimation algorithm using the state divided technique for a real time processing which is developed to provide an accurate attitude determination capability under a highly maneuverable dynamic environment. To compare the performance of the proposed algorithm with the EKF, simulations have been performed with various initial values and measurement covariances. Simulation results show that the proposed second order nonlinear algorithm outperforms the EKF. The proposed algorithm is useful for a real time attitude estimation since it has better accuracy compared with the EKF and requires less computing time compared with any existing nonlinear filters.     

面向移动机器人自定位的无线网络构造算法及实现

针对移动机器人的自定位问题,研制一种无线定位网络结点,该结点由超声模块和无线射频模块组成,根据超声与无线射频信号的到达时间差计算机器人与网络结点的相对位置关系,借助扩展的卡尔曼滤波器实现机器人的自定位和对网络结点的同步定位.然后,提出一种无线定位网络的自主构建算法,通过机器人动态分配网络结点,实现对环境的完备最小覆盖.仿真与实验表明该网络构造算法及定位技术能够满足较大规模环境下移动机器人的自定位要求.     

引入滑模观测器的GPS/INS组合导航滤波方法

由低成本器件组成的卫星/惯性(GPS/INS)组合导航系统中,存在较大的非线性与不确定性,为改善这一问题,本文提出一种引入滑模观测器(SMO)的滤波方法。首先,该方法建立了组合导航系统模型,介绍了扩展卡尔曼滤波(EKF)计算过程并分析存在的不足。然后,介绍了滑模观测器的基本原理,根据系统构建观测器。最后,说明了引入滑模观测器的EKF组合导航算法实现流程,滑模观测器将模型误差、状态估计以及均值方差融入EKF算法,修正系统输出。通过轨迹仿真实验与车载实验验证了所提方法优于传统EKF算法,具有更高的滤波精度。在车载实验中,卫星信号失锁15 s情况下,与EKF方法相比,所提方法的东向位置误差降低了53%,北向位置误差降低了37%,证明该方法能够有效抑制GPS/INS组合导航误差发散,为以后工程实践提供一定的参考价值。