基于循环神经网络的高超声速机动目标状态估计算法

针对高超声速机动目标运动状态的实时、高精确估计问题,提出了一种基于循环神经网络的高超声速机动目标状态估计算法。首先,将气动参数作为缓变量引入半速度系下的高超声速飞行器运动模型中,建立了高超声速机动目标状态估计模型,并给出了地面雷达测量模型;其次,考虑到高超目标的策略性机动带来的模型不确定性会超出常规滤波方法的处理能力,基于高超声速飞行器运动模型建立目标机动行为数据集,对其在实际作战过程中可能出现的运动行为进行模拟;然后,将多模型滤波与循环神经网络相结合设计了多滤波器联合估计算法;最后,在多种仿真条件下对本文所提出的算法进行了仿真验证。仿真结果表明,提出的高超声速机动目标状态估计算法可实现对高超声速目标机动变化的快速响应,并且算法的估计结果平滑、精度高,有利于后续预报。

三角交叉无源定位位置偏差估计滤波算法研究

提出了一种基于两观测站的三角交叉法对机动目标无源定位的算法 ,不是采用传统的对机动目标位置进行估计 ,而是直接对机动目标的位置偏差进行估计滤波 ,使定位的模糊区域迅速缩小 ,以达到快速定位跟踪的目的。列出了对位置偏差估计的状态方程和测量方程 ,用加权的最小二乘估计和卡尔曼滤波算法 ,实现了定位和跟踪。进行了计算机仿真 ,得到了良好的效果。

基于α-β滤波算法的目标运动信息滤波与估计

针对目标运动量测信息的特点,通过分析α-β滤波算法,指出该算法常增益的特点影响其对机动目标的跟踪效果。为了克服这个缺点,给出了一种新的目标运动信息滤波和估计的自适应算法,该算法可直接利用构造的函数关系自适应调整α,β的值。仿真结果表明,该自适应α-β滤波算法对目标运动信息的滤波和估计具有较好的性能。

自适应两步滤波算法在机载IRSTS被动定位中的应用

在机载红外搜索跟踪系统被动定位研究中,针对扩展卡尔曼滤波算法要求先验的噪声统计及存在系统观测模型线性化误差影响滤波精度的特点,利用两步滤波算法并结合Sage-Husa噪声估计器构建了适用于机载IRSTS被动定位特点的自适应两步滤波算法模型,算法不仅实时在线地估计了观测噪声的统计特性,而且避免了观测模型线性化误差。仿真结果表明,在完全相同的初始条件下,自适应两步滤波算法对目标运动参数的估计结果明显优于扩展卡尔曼滤波,从而提高了机载IRSTS被动定位的精度。

用于组合测姿的Kalman渐消因子自适应估计算法

提出了适用于姿态测量的Kalman滤波渐消因子自适应估计算法,滤波中采用序贯处理的方法计算出每个量测量对应的渐消因子,在位置速度组合导航系统中,只有位置、速度的误差状态是直接可观测的,用序贯滤波处理计算得到的渐消因子对协方差阵中对应于位置和速度误差状态的对角元素进行自适应控制,抑制滤波发散,提高位置、速度和姿态的测量精度。半实物仿真表明,与原来的算法相比,修改后的方法不仅能够提供高精度位置、速度信息,而且还可以提供高精度姿态信息,其中航向误差在0.08°以下,俯仰和横滚误差在0.02°左右。

锂电池SOC估计的实现方法分析与性能对比

锂电池荷电状态(state of charge,SOC)估计技术是保证电力储能和电动汽车合理应用的核心技术,也是锂电池系统控制运营、监测维护的基础。在锂电池实际应用中,其表现出非线性、时变性、影响因素复杂性和不确定性的问题,造成了荷电状态估计难度大、精度不高和适应能力不足。为此,众多锂电池荷电状态估计算法及改进策略应运而生。与此同时,部分研究人员针对不同估计方法和改进策略的实现方式和优缺点开展了分析与对比,但相关综述对估计方法的技术特点和适用性方面的论述不足且缺乏系统性总结。本文首先分析了锂电池荷电状态估计的影响因素和测试标准;然后从基于实验计算的传统方法、基于电池模型的滤波类算法、基于数据驱动的机器学习技术以及数模混合估计方法四个方面开展对比分析,归纳总结各类方法的技术特点、实现过程、适用条件、难题痛点以及应用优势,系统全面地论述了现有锂电池荷电状态估计技术的研究重点和应用现状;最后,展望了锂电池荷电状态估计算法的未来研究方向。

Filtering algorithm and direction identification in relative position estimation based on induction loop-cable

A filtering algorithm and direction identification method are presented for the positioning system of the mid-speed maglev train. Considering the special structure of the mid-speed maglev train, the ground position estimation method is adopted for its traction system. As the train is running, the induction loop-cable receives the signal sent by the on-board antenna to detect the position and direction of the train. But the height of the on-board antenna relative to the loop-cable is highly vulnerable to the change of the suspension height and the magnetic field produced by the traction during traveling, which may lead to amplitude fluctuation of the received signal. Consequently, the position estimation may be inaccurate. Therefore, a discrete second-order nonlinear trackdifferentiator is proposed based on the boundary characteristic curves, and the new differentiator could also extract the running direction of the train for the traction system. The experimental results show that the tracking differentiator can effectively filter out the signal interference and can provide accurate direction signal.

Multi-objective optimization sensor node scheduling for target tracking in wireless sensor network

Target tracking in wireless sensor network usually schedules a subset of sensor nodes to constitute a tasking cluster to collaboratively track a target.For the goals of saving energy consumption,prolonging network lifetime and improving tracking accuracy,sensor node scheduling for target tracking is indeed a multi-objective optimization problem.In this paper,a multi-objective optimization sensor node scheduling algorithm is proposed.It employs the unscented Kalman filtering algorithm for target state estimation and establishes tracking accuracy index,predicts the energy consumption of candidate sensor nodes,analyzes the relationship between network lifetime and remaining energy balance so as to construct energy efficiency index.Simulation results show that,compared with the existing sensor node scheduling,our proposed algorithm can achieve superior tracking accuracy and energy efficiency.

Estimation of catalytic activity using an unscented Kalman filtering in condensation reaction

The catalytic activity of cation exchange resins will be continuously reduced with its use time in a condensation reaction for bisphenol A(BPA).For online estimation of the catalytic activity,a catalytic deactivation model is studied for a production plant of BPA,state equation and observation equation are proposed based on the axial temperature distribution of the reactor and the acetone concentration at reactor entrance.A hybrid model of state equation is constructed for improving estimation precision.The unknown parameters in observation equation are calculated with sample data.The unscented Kalman filtering algorithm is then used for on-line estimation of the catalytic activity.The simulation results show that this hybrid model has higher estimation accuracy than the mechanism model and the model is effective for production process of BPA.