永磁同步电机的自适应云模型控制研究

针对PID或其改进的算法鲁棒性偏低问题,提出了永磁同步电机(PMSM)的自适应云模型控制算法研究。在分析了自适应云模型结构后设计了PMSM自适应云模型控制器。搭建了基于自适应云模型控制算法的PMSM试验平台,试验结果表明提出的PMSM的自适应云模型控制算法精度高、性能稳定。     

Application of Gram-Schmidt algorithm to fully adaptive arrays

The Gram-Schmidt orthogonalization (GSO) algorithm has excellent numerical performance and is readily applicable to systolic implementations such as in a field of adaptive cancellation systems. A modified GSO algorithm for a fully adaptive array is proposed and computer simulations show that the proposed algorithm gives superior performance. A systolic implementation of the proposed GSO algorithm for fully adaptive array is presented. A feedback mode GSO algorithm for use with analog weights is also presented and has been shown to have excellent performance in the presence of weight errors     

Partially Adaptive STAP using the FRACTA Algorithm

A partially adaptive space-time adaptive processor (STAP) utilizing the recently developed FRACTA algorithm is presented which significantly reduces the high computational complexity and large sample support requirements of fully adaptive STAP. Multi-window post-Doppler dimensionality reduction techniques are employed to transform the data prior to application of the FRACTA algorithm. The FRACTA algorithm is a reiterative censoring (RC) and detection algorithm which has been shown to provide excellent detection performance in nonhomogeneous interference environments. Two multi-window post-Doppler dimensionality reduction techniques are considered: PRI-staggered and adjacent-bin. The partially adaptive FRACTA algorithm is applied to the KASSPER I (Knowledge-Aided Sensor Signal Processing & Expert Reasoning) challenge datacube. The pulse repetition interval (PRI)-staggered approach with D=6 filters per Doppler bin is found to provide the best detection performance, outperforming the fully adaptive case while simultaneously reducing the runtime by a factor of ten. Using this implementation, partially adaptive FRACTA detects 197 out of 268 targets with one false alarm. The clairvoyant processor (the covariance matrix for each range cell is known) detects 198 targets with one false alarm. In addition, the partially adaptive FRACTA algorithm is shown to be resilient to jamming, and performs well for reduced sample support situations. When compared with partially adaptive STAP using traditional sliding window processing (SWP), the runtime of partially adaptive FRACTA is 14 times faster, and the detection performance is significantly increased (SWP detects 46 out of 268 targets with one false alarm).     

小波域的自适应波束形成算法

在分析传统自适应波束形成的基础上,首次提出了一种小波域的自适应波束形成算法。与通常的自适应波束形成算法相比,该算法利用小波变换对小波空间进行了分解,信号经小波变换自相关性会下降,收敛速度提高,同时在此分解过程中,根据信号与白噪声在不同尺度上的小波变换模极大值表现完全不同的特性进行信号的消噪。理论分析和仿真结果表明了该算法收敛速度较快,且计算量增加较少,易于实时实现,而且具有较好性能。同时仿真实验表明该算法收敛速度与小波基和尺度的选择有关,尺度越大收敛速度越快;对于同一小波基系列,小波基的正则性越好收敛速度越快。     

飞机防滑刹车系统自适应模糊控制的应用研究

提出了一种自适应模糊控制算法的防滑刹车系统设计方案,并在DSP和CPLD硬件上实现了这种智能算法。结果表明:采用自适应模糊控制算法在改善飞机刹车性能和提高刹车效率方面有良好的效果,并增强了系统的鲁棒性。     

Efficient robust AMF using the FRACTA algorithm

The FRACTA algorithm has been shown to be an effective space-time adaptive processing (STAP) methodology for the airborne radar configuration in which there exists nonhomogeneous clutter, jamming, and dense target clusters. Further developments of the FRACTA algorithm are presented here in which the focus is on the robust, efficient implementation of the FRACTA algorithm. Enhancements to the FRACTA algorithm include a censoring stopping mechanism, an alternative data blocking approach for adaptive power residue (APR) censoring, and a fast reiterative censoring (RC) procedure. Furthermore, a coherent processing interval (CPI) segmentation scheme for computing the adaptive weights is presented as an alternative approach to computing the adaptive matched filter (AMF) weight vector that allows for lower sample support and reduced computational complexity. The enhanced FRACTA algorithm, denoted as FRACTA.E, is applied to the KASSPER I challenge datacube which possesses dense ground target clusters that are known to have a significant deleterious effect on standard adaptive matched filtering (AMF) processors. It is shown that the FRACTA.E algorithm outperforms and is considerably more computationally efficient than both the original FRACTA algorithm and the standard sliding window processing (SWP) approach. Furthermore, using the KASSPER I datacube, the FRACTA.E algorithm is shown to have the same detection performance as the clairvoyant algorithm where the exact range-dependent clutter covariance matrices are known.     

一种基于特征空间的自适应天线旁瓣相消算法

把常规自适应天线旁瓣相消算法和特征空间技术相结合,提出了一种新的自适应天线旁瓣相消算法。该算法把常规自适应天线旁瓣相消算法的权矢量向由干扰特征矢量组成的干扰子空间投影,避免了由小特征值对应特征矢量组成的噪声子空间对权矢量的影响,与常规自适应天线旁瓣相消算法相比,该算法具有更好的干扰对消性能,其输出干扰对消比和波束方向图都能在很少的快拍下收敛。计算机仿真结果证实了这种算法的有效性。     

直升机结构响应自适应控制研究

研究了基于自适应滤波技术、采用伺服惯性力发生器（SIFG）的直升机结构响应主动控制;比较了被动动力吸振器和SIFG的减振效果。研究表明,SIFG在直升机前飞速度和旋翼转速及前飞速度改变的情况下均能保持良好的减振效果。     

基于自适应粒子群优化算法的无人机三维航迹规划

针对传统的粒子群优化算法容易陷入局部最优解的问题,提出了一种自适应粒子群优化算法,在迭代寻优过程中自适应地调节惯性权重和2个学习因子的数值。建立了无人机在山区环境执行勘察任务的航迹规划环境模型,分析了无人机自身约束条件。设计了自适应粒子群优化算法的适应度函数和航迹规划算法流程。分别采用自适应粒子群优化算法和传统粒子群优化算法开展了无人机三维航迹规划仿真实验。仿真结果对比表明,所提出的自适应粒子群优化算法比传统粒子群优化算法具有更高的全局搜索能力和搜索精度。     

一种变步长LMS自适应滤波算法研究

为了解决最小均方(LMS)算法的稳定性以及收敛速度和稳态误差之间的矛盾,提出了一种复数型自适应变步长的LMS算法,它的权系数的调整取决于误差曲面在新权值点上的梯度。分析了新算法的收敛特性以及参数选择对算法性能的影响。算法具有较快的收敛速度、鲁棒稳定性且运算小易于实现的特点。计算机仿真的结果与理论分析相一致,证实了算法的收敛性能优于标准的LMS算法并且具有较好的实用性。     

基于自适应遗传算法的DGPS整周模糊度快速解算

 鉴于遗传算法(GA)所具有的全局搜索特性,也为了更快速准确地解算差分全球定位系统(DGPS)整周模糊度,将自适应遗传算法(AGA)引入DGPS整周模糊度的搜索中。首先根据全球定位系统(GPS)载波相位双差方程求解出双差整周模糊度的浮点解,并以基线长度作为约束条件确定整周模糊度的搜索范围;然后利用白化滤波的方法对整周模糊度进行降相关处理,降低整周模糊度各分量之间的相关性;最后将自适应遗传算法应用在整周模糊度的解算过程中,搜索整周模糊度的最优解。仿真计算结果表明,与LAMBDA算法和简单遗传算法相比,自适应遗传算法能够快速地求解整周模糊度,也具有较好的可靠性和鲁棒性。     

Adaptive array transient sidelobe levels and remedies

The transient sidelobe level of a sidelobe canceler (SLC) is a function of the external noise environment, the number of adaptive auxiliary antennas, the adaptive algorithm used, auxiliary antenna gain margins, and the number of samples used to calculate the adaptive weights. An analytical result for the adaptive sidelobe level is formulated for the case when the adaptive algorithm is the open-loop, sampled matrix inversion (SMI) algorithm. The result is independent of whether concurrent or nonconcurrent data processing is used in the SMI algorithm's implementation. It is shown that the transient sidelobe level is eigenvalue dependent and increases proportionally to the gain margin of the auxiliary antenna elements with respect to the quiescent main antenna sidelobe level. Techniques that reduce this transient sidelobe level are discussed, and it is theoretically shown that injection independent noise into the auxiliary channels significantly reduces the transient sidelobe level. It is demonstrated that using this same technique reduces the SMI noise power residue settling time     

Doppler Compensation & Single Pulse Imaging using Adaptive Pulse Compression

The effects of target Doppler are addressed in relation to adaptive receive processing for radar pulse compression. To correct for Doppler-induced filter mismatch over a single pulse, the Doppler-compensated adaptive pulse compression (DC-APC) algorithm is presented whereby the respective Doppler shifts for large target returns are jointly estimated with the illuminated range profile and subsequently incorporated into the original APC adaptive receive filter formulation. As a result, the Doppler-mismatch-induced range sidelobes can be suppressed thereby regaining a significant portion of the sensitivity improvement that is possible when applying adaptive pulse compression (APC) without the existence of significant Doppler mismatch. In contrast, instead of compensating for Doppler mismatch, the single pulse imaging (SPI) algorithm generalizes the APC formulation for a bank of Doppler-shifted matched filters thereby producing a sidelobe-suppressed range-Doppler image from the return signal of a single radar pulse which is applicable for targets with substantial variation in Doppler. Both techniques are based on the recently proposed APC algorithm and its generalization, the multistatic adaptive pulse compression (MAPC) algorithm, which have been shown to be effective for the suppression of pulse compression range sidelobes thus dramatically increasing the sensitivity of pulse compression radar.     

优化抗差自适应滤波器在深组合导航中的应用

为了提高惯性/卫星深组合导航系统的滤波性能,在抗差自适应滤波算法的 基础上,研究了一种优化抗差自适应滤波算法。该算法通过比较实际预测残差协方差矩 阵和理论协方差阵的差值来生成自适应因子,从而优化抗差自适应滤波。将所研究的算 法应用于惯性/卫星深组合导航系统, 在高动态环境下进行仿真验证, 并与常规卡尔曼 滤波、抗差自适应滤波进行比较。结果表明,优化算法能有效地控制观测异常和动态模 型异常对状态参数估值的影响,所得组合导航位置误差和速度误差明显减小,提高了组 合导航系统的滤波精度。     

基于预警卫星系统的临近空间飞行器跟踪简

 系统地研究了如何对临近空间飞行器进行有效实时跟踪的问题,并提出了一种基于约束总体最小二乘与自适应交互式多模型（CTLS-AIMM）滤波相结合的实时跟踪滤波算法。首先考虑到临近空间飞行器的特点,选择使用红外预警卫星系统探测目标飞行器,并使用约束总体最小二乘算法（CTLS）对目标进行粗定位;然后在粗定位信息基础上,使用自适应交互式多模型滤波算法（AIMM）对目标飞行器进行实时跟踪。在AIMM中,根据临近空间飞行器机动特性,合理选择目标模型集,并使用迭代最小二乘算法对模型参数进行自适应调整。通过仿真,验证了该跟踪滤波算法的可行性。     

基于预警卫星系统的临近空间飞行器跟踪

系统地研究了如何对临近空间飞行器进行有效实时跟踪的问题,并提出了一种基于约束总体最小二乘与自适应交互式多模型（CTLS-AIMM）滤波相结合的实时跟踪滤波算法。首先考虑到临近空间飞行器的特点,选择使用红外预警卫星系统探测目标飞行器,并使用约束总体最小二乘算法（CTLS）对目标进行粗定位;然后在粗定位信息基础上,使用自适应交互式多模型滤波算法（AIMM）对目标飞行器进行实时跟踪。在AIMM中,根据临近空间飞行器机动特性,合理选择目标模型集,并使用迭代最小二乘算法对模型参数进行自适应调整。通过仿真,验证了该跟踪滤波算法的可行性。     

抗差自适应分步滤波算法在PPP/INS组合导航中的应用

针对常规抗差自适应滤波算法在PPP/INS组合导航应用中存在难以准确识别和分离观测粗差及运动异常对定位结果影响的问题，基于分类因子自适应滤波原理,提出了一种抗差自适应分步滤波算法。该算法首先执行第一步滤波,对状态模型异常信息进行隔离，仅对观测粗差进行诊断和抗差处理；然后在第一步滤波的基础上，执行第二步滤波，对状态模型异常进行诊断和自适应处理。算法分析表明，抗差自适应分步滤波算法可以准确地识别和分离观测粗差和运动异常扰动。实验结果表明，抗差自适应分步滤波算法能够进一步增强滤波算法抵抗观测粗差和运动异常扰动对滤波结果的影响，提高PPP/INS组合导航系统定位结果的稳定性和可靠性。     

An Improved Algorithm for Adaptive Processing

The Widrow-Hoff least mean square (LMS) algorithm based on the method of steepest descent is conditionally stable. A modified algorithm is given which is unconditionally stable, capable of better performance when used in adaptive filter processing, and provides a more realistic means for simulating the Applebaum-Howells adaptive loop.     

Phase-only LMS and perturbation adaptive algorithms

The author defines a phase-only gradient-based adaptive algorithm analogous to the least mean square (LMS) algorithm. Two phase-only perturbation algorithms are then defined. It is shown that the gradient for the perturbation algorithms can be obtained by cross-correlating binary perturbation sequences, that are applied to the adaptive phases, with the resulting instantaneous output power or voltage. It is also shown that a single set of phase-only adaptive weights can be used to simultaneously null interference in multiple output beams. Simulation results are presented for all of the new algorithms. The phase-only perturbation techniques eliminate the need for element level receivers and support low cost retrofitting of adaptive nulling on phased arrays by using conventional beamsteering circuits to apply the adaptive weights     

Adaptive Interpolated Time-Delay Estimation

An adaptive delay-estimation (ADE) algorithm is proposed for the continuous tracking of time-delay. The method uses an adaptive delay line which is interpolated by a first-order filter. Two delay-line interpolating filters are considered, each having a single coefficient which is estimated in real time. The first implements linear interpolation, and the second interpolates using a first-order allpass filter. Since the ADE algorithm is derived from recursive Gauss-Newton optimization, it can be viewed as a recursive maximum likelihood (RML) algorithm for time-delay estimation.