概率近似正确的强化学习算法解决连续状态空间控制问题

在线学习时长是强化学习算法的一个重要指标.传统在线强化学习算法如Q学习、状态–动作–奖励–状态–动作(state-action-reward-state-action,SARSA)等算法不能从理论分析角度给出定量的在线学习时长上界.本文引入概率近似正确(probably approximately correct,PAC)原理,为连续时间确定性系统设计基于数据的在线强化学习算法.这类算法有效记录在线数据,同时考虑强化学习算法对状态空间探索的需求,能够在有限在线学习时间内输出近似最优的控制.我们提出算法的两种实现方式,分别使用状态离散化和kd树(k-dimensional树)技术,存储数据和计算在线策略.最后我们将提出的两个算法应用在双连杆机械臂运动控制上,观察算法的效果并进行比较.     

Intelligent adaptive observer-based optimal control of overhead transmission line de-icing robot manipulator

In cold season, wet snow ice accretion on overhead transmission lines increases wind load effects which in turn increases line tension. This increased line tension causes undesirable effects in power systems. This paper discusses the design of an observer-based boundary sliding mode control (BSMC) for 3 DOF overhead transmission line de-icing robot manipulator (OTDIRM). A robust radial basis functional neural network (RBFNN) observer-based neural network (NN) controller is developed for the motion control of OTDIRM, which is a combination of BSMC, NN approximation and adaptation law. The RBFNN-based adaptive observer is designed to estimate the positions and velocities. The weights of both NN observer and NN approximator are tuned off-line using particle swarm optimization. Using Lyapunov analysis the closed loop tracking error was verified for a 3 DOF OTDIRM. Finally, the robustness of the proposed neural network-based adaptive observer boundary sliding mode control (NNAOBSMC) was verified against the input disturbances and uncertainties.     

Dynamic neural network method-based improved PSO and BR algorithms for transient probabilistic analysis of flexible mechanism

To improve the computing efficiency and precision of transient probabilistic analysis of flexible mechanism, dynamic neural network method (DNNM)-based improved particle swarm optimization (PSO)/Bayesian regularization (BR) (called as PSO/BR-DNNM) is proposed based on the developed DNNM with the integration of extremum response surface method (ERSM) and artificial neural network (ANN). The mathematical model of DNNM is established based on ANN on the foundation of investigating ERSM. Aiming at the high nonlinearity and strong coupling characteristics of limit state function of flexible mechanism, accurate weights and thresholds of PSO/BR-DNNM function are discussed by searching initial weights and thresholds based on the improved PSO and training final weights and thresholds by the BR-based training performance function. The probabilistic analysis of two-link flexible robot manipulator (TFRM) was investigated with the proposed method. Reliability degree, distribution characteristics and major factors (section sizes of link-2) of TFRM are obtained, which provides a useful reference for a more effective TFRM design. Through the comparison of three methods (Monte Carlo method, DNNM, PSO/BR-DNNM), it is demonstrated that PSO/BR-DNNM reshapes the probability of flexible mechanism probabilistic analysis and improves the computing efficiency while keeping acceptable computational precision. Moreover, the proposed method offers a useful insight for reliability-based design optimization of flexible mechanism and thereby also enriches the theory and method of mechanical reliability design.     

Boundary Control of a Flexible Robotic Manipulator With Output Constraints

In this study, we consider a boundary control problem of a flexible manipulator with input disturbances and output constraints, achieving pre‐set performance attributes on position tracking error and the deflection error at the end of the beam. The dynamics of the system are represented by partial differential equations (PDEs). With the Lyapunov's direct method, a boundary controller with disturbance observer is designed to regulate the angular position and suppress elastic vibration simultaneously. The proposed control scheme allows the errors to converge to an arbitrarily small residual set, with convergence rate larger than a pre‐specified value. Numerical simulations demonstrate the effectiveness of the proposed scheme.     

End‐Point Regulation and Vibration Suppression of a Flexible Robotic Manipulator

In this paper, end‐point regulation and vibration suppression are investigated for a flexible robotic manipulator subject to the external disturbances. The dynamics of the flexible robotic manipulator is represented by one partial differential equation (PDE) and five ordinary differential equations (ODEs). Based on the Lyapunov's direct method, boundary control is developed to drive the manipulator to the desired set‐point and simultaneously suppress the vibrations of the flexible manipulator. Considering the unknown spatiotemporally varying disturbance, uniform boundedness of the closed‐loop system is achieved. The control performance of the closed‐loop system is guaranteed by suitably choosing the design parameters. Simulations are provided to illustrate the effectiveness of the proposed control.     

基于Kinect的机械臂增强现实示教研究

机械臂广泛地应用在工业自动化生产以及人们日常生活中,机械臂而示教学习作为一种重要的运动规划方式,是机械臂研究领域的一项重要研究内容。鉴于此,设计了一种基于Kinect的虚拟机械臂增强现实示教系统,通过Kinect传感器采集人手位置数据,利用机械臂逆运动学模型驱动虚拟机械臂跟踪人手运动,实现对机械臂的虚拟示教;利用增强现实技术,将虚拟机械臂叠加到真实环境中,实现对机械臂的增强现实示教。实验结果表明该系统能够将虚拟机械臂准确地叠加到真实环境中,并实时跟踪人手运动,系统具有良好的交互性。     

基于压电陶瓷的柔性机械臂主动振动控制实验研究

柔性机构因其工作效率高,能量消耗低和结构简单等优点被广泛应用于机器人领域.但柔性机构由于刚度较低,易产生弹性振动,使机器人的定位精度和运动精度降低.为了验证在前期的工作中提出的柔性机械臂主动振动最优控制位置分析的正确性,搭建了一个基于Labview的测试系统.在柔性臂上不同位置粘贴压电片来实现对柔性臂不同位置的主动振动控制,通过测试系统得到最优控制位置处和最优控制位置附近的柔性机械臂振动的传感电压,根据控制前后传感电压的变化分析柔性臂抑制的强弱.根据实验结果显示,柔性臂在前三阶振动下,分别在其最优控制位置上获得最高的抑振率,从而验证了最优控制位置的有效性.     

一种程控标准电流源设计

设计一款应用于指针表的图像采集系统的程控标准电流源,从功能要求、硬件结构设计及电路设计等几个方面进行了论述.设计的程控标准电流源遵循与计算机匹配的串口通讯规约,可对计算机命令做出正确的执行和回答,可在程序控制下,输出标准电源,电源的精度要求比被检定仪表精度高,符合图像采集系统要求.     

多状态监控自动窗的设计

针对现在的热点物联网(IOT)和具有"蓝海产业"之称的智能家居,提出了一套家庭窗户智能管理方案,能对室内外多种状态进行监控、分析、显示和对窗户进行自动控制,并集成了GSM模块,实现远程报警,也能通过手机方便的查询窗户及室内外状态和进行远程控制。     

低空飞艇地面监控系统设计与实现

以某低空实验飞艇项目为背景,基于LabWindows/CVI设计并实现了低空飞艇地面监控系统的上层软件,对底层串口通信程序、飞艇遥控操纵杆程序以及飞艇实时状态数据库连接等几个关键问题进行了详细设计。该地面监控系统经过长期运行表明该监控软件运行稳定,功能完备实用,具有较高的工程应用价值。