Adaptive force–environment estimator for manipulators based on adaptive wavelet neural network

This study focuses on the accurate tracking control and sensorless estimation of external force disturbances on robot manipulators. The proposed approach is based on an adaptive Wavelet Neural Network (WNN), named Adaptive Force-Environment Estimator (WNN-AFEE). Unlike disturbance observers, WNN_AFEE does not require the inverse of the Jacobian transpose for computing the force, thus, it has no computational problem near singular points. In this scheme, WNN estimates the external force disturbance to attenuate its effects on the control system performance by estimating the environment model. A Lyapunov based design is presented to determine adaptive laws for tuning WNN parameters. Another advantage of the proposed approach is that it can estimate the force even when there are some parametric uncertainties in the robot model, because an additional adaptive law is designed to estimate the robot parameters. In a theorem, the stability of the closed loop system is proved and a general condition is presented for identifying the force and robot parameters. Some suggestions are provided for improving the estimation and control performance. Then, a WNN-AFEE is designed for a planar manipulator as an example, and some simulations are performed for different conditions. WNN_AFEE results are compared attentively with the results of an adaptive force estimator and a disturbance estimator. These comparisons show the efficiency of the proposed controller in dealing with different conditions.     

底板岩层变形破坏过程中混沌性态的Lyapunov指数描述研究

基于大量实测信息 ,对承压水上开采煤层底板变形破坏过程中岩移、注水量 (渗透性 )等矿压显现的观测时序 ,进行Lyapunov指数的提取 ,并对其混沌性态进行了研究。实例分析表明 :煤层底板变形破坏过程中呈现出典型的混沌性态 ;另外 ,由于煤层底板渗透性指标的Lyapunov指数比岩移量指标大的多 ,表明用渗透性指标描述煤层底板变形破坏特征要比岩移量指标更敏感     

饱和嵌入维数确定最大Lyapunov指数的准则探讨

基于对大量砂岩试样裂隙系统的MTS全应力–应变时间序列的非线性动力学分析,构造单位阶跃函数并引入到最大Lyapunov指数LE1的判别模型中,建立了确定LE1的新准则。新准则充分考虑LE1-m关系中普遍存在的阶跃现象,更符合确定LE1的理论条件。新准则确定的LE1与传统准则的解释结果比较,偏离幅度巨大,使得对系统稳定性程度的混沌判别结果迥异,并且LE1的正负号会因此改变,导致对系统所处状态的定性评价不同。研究成果在一定程度上克服了传统准则确定LE1鲁棒性较差的缺陷,对混沌动力学基础研究具有一定的理论意义。     

基于混沌时序预测方法的冲击地压预测研究

冲击地压的预测研究通常是通过监测对冲击地压的发生、发展过程比较敏感的指标来进行的，监测指标值的大小和变化规律是进行预测的基础，监测数据在未来一定时期的峰值变化和走势变化规律对于预测过程具有重要意义。首先，通过对互信息和伪邻近点数的计算，确定观测序列的延迟时间和嵌入维数等相空间重构参数；然后，在对观测序列相空间重构的基础上，运用一阶局域近似法和基于最大Lyapunov指数法等混沌预测方法对冲击地压上作面的观测时间序列进行数学建模，并与传统的数理统计预测方法进行对比分析；最后，用实例对冲击危险区域的电磁辐射序列及顶板下沉速度序列等进行预测运算和分析，其结果表明，运用混沌理论的预测方法可达到较高的预测精度。     

应用Lyapunov指数研究混凝土阶段特征的稳定判据

根据混凝土材料的应力-应变试验研究,得出试验曲线呈明显的分段特征,曲线的离散性发生于应变软化阶段。通过对试验数据序列的混沌动力学分析得出:随着单向抗压强度的增加,混凝土材料全应力-应变过程的系统状态由较有序的定常态质变为低混沌态,进而逐渐演化到强混沌态;混凝土材料峰后应力-应变阶段曲线的线性性质增强,裂隙系统状态由强有序的定常态向弱混沌态演化;混凝土材料峰前阶段的裂隙系统基本呈现弱混沌状态。提出将LE1作为混凝土材料阶段特征的稳定判据,将LE1作为系统稳定度的表征量,并在实例计算中证明了该法确定的稳定判据符合实际情况。     

双组分层撞击流反应器浓度场混沌分析

采用平面激光诱导荧光技术测量双组分层式撞击流反应器撞击区浓度分布,利用混沌理论分析不同喷嘴间距、喷嘴直径和射流雷诺数下撞击流反应器浓度场混沌特征参数(关联维、Kolmogorov熵和最大Lyapunov指数)的变化规律。结果表明,撞击流反应器内浓度场混沌特征参数随喷嘴间距的增大呈上下波动的变化趋势,Kolmogorov熵在L=2d时达到最大。浓度场混沌特征参数随喷嘴直径增大呈上下波动的变化趋势,Kolmogorov熵在d=8 mm时整体上最大;射流雷诺数为Re=22 000时整体上混沌特征参数最佳。受二次撞击区的影响,撞击面上靠近二次撞击区各点的混沌特征参数有明显提高。     

异步电机基于MRAC的转子时间常数在线辨识算法的统一描述

应用于异步电机转子时间常数在线辨识的模型参考自适应控制器(MRAC),依据其参考模型设计的不同,有着不同的算法形式。在深入分析转子时间常数T_r偏差对电机间接磁场定向、输出电磁转矩、无功功率影响的基础上,运用李雅普诺夫稳定理论提出了T_r在线辨识的统一MRAC自适应律,通过自适应律中参数的不同配置,可以得到多种基于MRAC的T_r在线辨识算法。借助于统一自适应律理论,为多种辨识模型间的演化和对比提供了理论依据。同时,该文所提自适应律的统一描述也有助于基于MRAC的T_r辨识理论体系的完善,使得辨识模型的选择有章可循。仿真和实验结果验证了该文分析的有效性。     

Robust nonlinear adaptive backstepping excitation controller design for rejecting external disturbances in multimachine power systems

This paper proposes a new approach to design a robust adaptive backstepping excitation controller for multimachine power systems in order to reject external disturbances. The parameters which significantly affect the stability of power systems (also called stability sensitive parameters) are considered as unknown and the external disturbances are incorporated into the power system model. The proposed excitation controller is designed in such a way that it is adaptive to the unknown parameters and robust to external disturbances. The stability sensitive parameters are estimated through the adaptation laws and the convergences of these adaptation laws are obtained through the negative semi-definiteness of control Lyapunov functions (CLFs). The proposed controller not only provides robustness property against external disturbances but also overcomes the over-parameterization problem of stability sensitive parameters which usually appears in some conventional adaptive methods. Finally, the performance of the proposed controller is tested on a two-area four machine 11-bus power system by considering external disturbances under different scenarios and is compared to that of an existing nonlinear adaptive backstepping controller. Simulation results illustrate the robustness of the proposed controller over an existing one in terms of rejecting external disturbances.     

A stochastic methodology to adjust controllers based on moments Lyapunov exponents: Application to power systems

This paper presents a methodology, based on linear analysis, that uses stability indices based on the calculation of moments Lyapunov exponents to determine the optimum adjustment of controllers in linear stochastic systems. The purpose is to define what values the parameters must have so that the system, subjected to perturbations self-sustained over time, can maintain its stability during operation in permanent regime. The proposed development is applied to three test systems, including two electric power system.     

Sepic变换器直接电流控制

基于李亚普诺夫直接法和输入输出线性化控制法,提出了一种新型Sepic变换器直接电流控制策略。建立了Sepic变换器单输入单输出仿射非线性系统模型。采用李亚普诺夫直接法,建立了Sepic变换器李亚普诺夫直接法控制系统。在李亚普诺夫直接法的基础上提出了一种直接电流控制方法,通过直接控制电感电流,间接获得期望的输出电压。比较分析发现该直接电流控制法比李亚普诺夫直接法控制结构更简单,控制代价更低,易于数字实现。数值仿真和实验结果验证了所提控制策略的正确性和优越性。基于所提控制方法的系统输出无稳态误差,并对负载扰动具有较强的鲁棒性。