一种基于多项式函数的VSSNLMS算法

为了改善现有变步长NLMS自适应算法存在的步长调节参数较多、不易调节、对噪声较为敏感的缺点,综合已有的变步长算法提出一种基于多项式函数的变步长归一化LMS(VSSNLMS)算法。理论分析和仿真结果表明,新算法在低信噪比条件下具有较快的收敛速度和较小的稳态误差,而且参数调节简单,适用于不同的输入信号带宽。新算法相比其他同类算法实时计算量小,具有很好的工程实用价值。     

永磁直线同步电机超短反馈滑模控制研究

永磁直线同步电机矢量控制的速度环常采用PI控制。传统PI控制器在速度追踪时存在起动过程超调大、受到负载扰动时调节时间长等不足之处。为提高伺服性能,滑模控制被应用到伺服系统中,但其缺点是存在抖振问题。因此,提出一种基于内分泌激素调节的滑模控制方法,可以有效降低速度的超调量,缩短受到外部扰动时的调节时间。在MATLAB中搭建了系统的仿真模型,仿真结果表明其对速度和推力的控制效果显著,优于传统的PI控制和常规的滑模控制。     

基于MRAS和磁链补偿的无速度传感器永磁直线同步电机直接推力控制

永磁直线同步电机(PMLSM)直接推力控制系统中的传统机械传感器在恶劣工况下难以准确获取控制系统反馈信息。将模型参考自适应算法应用到PMLSM,设计了基于模型参考自适应系统(MRAS)的无速度传感器直接推力控制系统,依据辨识得到的磁链位置重新构建了一种磁链观测器,对磁链进行补偿,减小了直接推力控制推力响应的波动。通过仿真,证明了基于MRAS和磁链补偿的无速度传感器PMLSM直接推力控制系统能够准确地辨识初级的速度和位置信息,得到了较好的动静态性能。     

涡扇发动机极点配置圆的多变量PI控制设计

为了实现对涡扇发动机高精准稳态控制,在静态输出反馈渐近稳定条件的基础上,考虑控制系统的鲁棒稳定裕度,推导了闭环极点配置圆的条件,提出了 1 种多变量控制系统闭环极点配置圆的线性矩阵不等式 LMI 设计方法。在双转子涡扇发动机上进行了仿真验证,双回路控制仿真表明:控制系统对于高压转子转速和涡轮落压比回路具有伺服跟踪和抗回路耦合干扰性能;单回路控制仿真表明:在不考虑执行机构动态直接进行控制器设计时,其相角裕度将减小 30°左右,导致系统的动态性能和稳定性变差。     

永磁直线同步电机滑模控制系统

设计了一种基于滑模变结构的永磁直线同步电机(PMLSM)矢量控制系统。从直线电机的基本工作原理出发,通过坐标变换,建立PMLSM在两相同步旋转正交坐标系上的数学模型。基于滑模变结构和李雅普诺夫稳定性理论设计转速调节器,组成PMLSM滑模控制系统。为了验证该控制系统的有效性,在MATLAB/Simulink平台下搭建系统的模型并进行仿真。仿真表明,该控制系统具有很强的鲁棒性。     

表贴式永磁直线同步电机控制系统参数自动整定方法

针对表贴式永磁直线同步电机矢量控制系统参数设置需反复测量和调试的问题,提出一种自动整定的方法。测量电流控制器输出波形并计算相应的沃尔什一阶系数,反复迭代反电动势常数和定子电感,直至沃尔什一阶系数收敛到零,自动辨识表贴式永磁直线同步电机的电气参数;随后让电机自由运行辨识出机械参数;最后计算出速度和电流控制器的PI参数。样机测试结果验证了提出方法的有效性。     

月球探测器动力下降段最优轨迹参数化方法

为保证月球探测器进入姿态调整段时具有充分的高度与速度余量,本文提出一种基于控制变量参数化的月球探测器动力下降段最优轨迹求解方法.在三维探测器软着陆动力学模型基础上,将月球探测器软着陆制导律设计等效为燃料最优约束下的探测器俯仰角控制问题,利用控制变量参数化(Control Variables Parameterizati...     

基于环路滤波器的低复杂度LMS算法设计

研究利用多天线结构对测控信号内存在的干扰进行对消和信号增强的技术。首先介绍传统空域干扰对消算法LMS，以及LMS的诸多改进算法，分析当前这些LMS算法在强干扰存在时收敛稳定性和收敛速度之间的折中关系以及参数调节的难度，分析了定步长和变步长LMS算法的优劣。针对变步长LMS由于动态环境以及大功率干扰的存在引起的收敛速度变慢和步长计算复杂的问题，提出了一种基于二阶环路滤波器累计误差的低复杂度LMS算法LF-LMS，该算法弱化单次误差的不稳定性并提高累计误差的稳定性，利用大步长和可变误差以实现快速收敛并缩小进入稳定状态后的误差波动，且计算量较小，可以有效应用于DSP工程实现，并通过仿真验证了算法在收敛速度和动态环境下的鲁棒性。     

Guaranteed performance based adaptive attitude tracking of spacecraft with control constraints

The guaranteed performance control problem of spacecraft attitude tracking with control constraint, disturbance and time-varying inertia parameters is investigated. A new saturation function is designed to satisfy different magnitude constraints by introducing a piecewise smooth asymmetric Gauss error function. Based on the mean-value theorem, the constrained problem is transformed into an unconstrained control design subject to an unknown bounded coefficient matrix. To satisfy the constraints by performance functions, a tracking error constrained control is developed based on a hyperbolic arc-tangent asymmetric barrier Lyapunov function (BLF). In the backstepping framework, an adaptive robust control law is proposed by employing a smooth robust term simultaneously counteracting the parametric and non-parametric uncertainties, where the unknown coefficient matrix resulting from the control constraint is compensated by a Nussbaum function matrix. Rigorous stability analysis indicates that the proposed control law realizes the asymptotically tracking of spacecraft attitude and that the tracking error remains in a prescribed set which implies the achievement of the guaranteed transient performance. Numerical simulations validate the proposed theoretical results.     

Influence of land surface parameters on the spatio-seasonal land surface temperature regime in rural West Bengal,India

The magnitude and causes of changes in the land surface temperature of rural areas have not been extensively studied. The thermal band of Landsat imagery is taken to extract winter, summer, and monsoon season land surface temperature (LST) and relate it to surface parameters over a 30-year period. From the extracted parameters constructed a prospective surface temperature (PST) model using Multivariate Adaptive Regression Splines. The Chandrabhaga river basin in West Bengal of the lateritic Rarh Tract at the Chota Nagpur Plateau fringe was chosen as the study area because it is far from urban influences, to avoid the well-known heat island effect. Over the study period, summer and winter average LST increased linearly by 0.085?°C/y and 0.016?°C/y respectively. These results were validated with air temperature (RMSE?=?x and y, respectively). Over time more of the area is in the higher temperature zones, e.g., in April 2011, 4% area exceeded >32°, whereas in 2015 this proportion reached 52%. PST models of all the seasons were moderate to highly correlate (0.57–0.87) with actual LST, showing the value of this model. It also revealed the relative importance of the regional factors. Based on this information factor management is a scientific step to restrict or minimize the temperature rise effect.