Adaptive neural network control of nonlinear systems with unknown dynamics

In this study, an adaptive neural network control approach is proposed to achieve accurate and robust control of nonlinear systems with unknown dynamics, wherein the neural network is innovatively used to learn the inverse problem of system dynamics with guaranteed convergence. This study focuses on the following three contributions. First, the considered system is transformed into a multi-integrator system using an input–output linearization technique, and an extended state observation technique is used to identify the transformed states. Second, an iterative control learning algorithm is proposed to achieve the neural network training, and stability analysis is given to prove that the network’s predictions converge to ideal control inputs with guaranteed convergence. Third, an adaptive neural network controller is developed by combining the trained network and a proportional-integral controller, and the long-standing challenge of model-based methods for control determination of unknown dynamics is resolved. Simulation results of a virtual control mission and an aerospace altitude tracking mission are provided to substantiate the effectiveness of the proposed techniques and illustrate the adaptability and robustness of the proposed controller.     

Inverted decoupling and LMI-based controller design for a turboprop engine with actuator dynamics

The main objective of the turboprop engine control system is to ensure propeller absorbed power at a constant propeller speed by controlling fuel flow and blade angle. Since each input variable affects the selected output variables, there exist strong interactions between different control loops of a Two-Spool TurboProp Engine (TSTPE). Inverted decoupling is used to decouple the interactions and decompose the TSTPE into two independent single-input single-output systems. The multi-variable PI controller and two single-variable PI controllers are designed for the TSTPE with actuator dynamics based on Linear Matrix Inequality (LMI), respectively, which is derived from static output feedback and pole placement condition. The step responses show that due to the difference in the response times of the selected output variables, it is difficult to design an appropriate multi-variable PI controller. The designed single-variable PI controllers are tested on the TSTPE integrated model to illustrate the effectiveness of the proposed method, that is, the interactions are first decoupled and then the controllers are designed, and the resulting simulated responses show that compared with the controller designed without actuator dynamics, the gas-generator shaft speed and power turbine shaft speed can better track their respective commands under the action of the controller designed with actuator dynamics.     

低空风切变系统建模及其对直升机飞行安全威胁定性分析

系统地建立了含湍流的三维低空风切变模型，并综合分析了其对直升机飞行安全的威胁。建立了微下击暴流风场及大气湍流场组合的风切变模型，在不增加计算量的前提下，选取特征点发展直升机飞行动力学模型，有效捕捉了风切变的切变特性，提高了在风场中的动态响应计算精度。模型配备姿态保持功能的控制增稳系统已符合一般直升机的飞行状态，并改善机体响应。根据风速分布的特点，选取不同飞行速度、不同风场位置进行飞行仿真，定性地对比分析状态量变化与风场对应关系，并且以垂向通道为例，从动力学角度分析验证了响应的理论计算表达式。结果表明：湍流主要导致高频姿态角响应，风切变对飞行状态量变化占主导作用，且垂向风是引起威胁的主要因素，据此提出危险风场规避建议。     

基于三维有限元的火箭模态振型斜率预示方法研究

振型斜率火箭是姿控系统设计的重要参数,对飞行任务的可靠性与安全性有着至关重要的作用。依据全箭动特性的测量原理及参数定义,结合三维有限元建模方法和模型修正技术,全面考虑了速率陀螺、惯组安装位置处的局部刚度对振型和振型斜率的影响,建立了基于局部细节精细化建模的全箭模态振型斜率预示方法,并总结了3种不同的振型斜率提取方法,提高了振型斜率的预示精度。同时提供了三维模型振型一维化的方法,解决了使用三维有限元模型计算弹性运动方程式系数的难题。     

Attitude stability and periodic attitudes of rigid spacecrafts on the stationary orbits around asteroid 216 Kleopatra

In this work, equilibrium attitude configurations, attitude stability and periodic attitude families are investigated for rigid spacecrafts moving on stationary orbits around asteroid 216 Kleopatra. The polyhedral approach is adopted to formulate the equations of rotational motion. In this dynamical model, six equilibrium attitude configurations with non-zero Euler angles are identified for a spacecraft moving on each stationary orbit. Then the linearized equations of attitude motion at equilibrium attitudes are derived. Based on the linear system, the necessary conditions of stability of equilibrium attitudes are provided, and stability domains on the spacecraft’s characteristic plane are obtained. It is found that the stability domains are distributed in the first and third quadrants of the characteristic plane and the stability domain in the third quadrant is separated into two regions by an unstable belt. Subsequently, we present the linear solution around a stable equilibrium attitude point, indicating that there are three types of elemental periodic attitudes. By means of numerical approaches, three fundamental families of periodic solutions are determined in the full attitude model.     

VSV调节机构阻滞力和调节精度的归因分析

基于ANCF方法和摩擦力的NURBS表达算法，构建了考虑装配间隙、尺寸公差、关键件柔性、热影响以及接触摩擦等因素的复杂VSV调节机构的刚柔耦合动力学模型，并依据部件空间分布关系和参数化表述开发了VSV调节机构的快速自动化建模流程，大大提升了建模效率。分别对VSV调节机构的单级模拟气动实验及热态联调实验状态进行了对比仿真，结果表明，所构建的VSV调节机构模型在阻滞力仿真方面具有较高的可信度，且气动力引起的摩擦力矩是造成机构阻滞力、角度调节迟滞和调节精度下降的主要原因，气动力矩对机构阻滞力和角度迟滞影响不明显，关键件柔性对阻滞力影响较小，但却是调节精度降低的内在因素。     

特征无反射边界条件应用于二维圆柱绕流模拟

采用特征无反射边界条件和吸收层技术相结合的边界处理方法,研究了二维圆柱绕流模拟的边界反射问题。近年来在计算气动声学中提出了将大涡模拟和声学传播方程结合的混合方法求解声场,基于计算量的考虑,须在较小的计算区域内进行大涡模拟,这就需要在计算区域较小时,边界的反射尽可能小。作为大涡模拟研究的技术储备,将该方法应用于二维圆柱绕流的模拟中。数值结果表明这种方法在计算区域较小时,能起到很好的无反射效果,计算结果与前人在计算区域取得很大时的结果比较吻合。     

基于ANCF的松弛绳索动力学建模与仿真

传统绝对节点坐标方法(ANCF)绳索模型是基于梁单元建立的,其不能反映纤维绳索的不可抗压和松弛特性。考虑纤维绳索初始松弛余量,给出了绳索非线性轴向应力-应变关系。绳索在松弛状态下轴向力接近于零,在张紧状态下表现出线弹性特性。在此基础上,采用绝对节点坐标方法推导了松弛绳索动力学模型。通过静力学和动力学仿真,将传统绳索模型和松弛绳索模型进行对比,结果表明:在重力及不同集中载荷作用下,松弛绳索相比于传统绳索都具有更大的变形;分析移除集中载荷后绳索的动力学响应可知,传统绳索在振动过程中始终处于张紧状态,且绳索上各点振动同步,而松弛绳索会在张紧和松弛状态之间不断转换,绳索上各点的振动存在相位差,能够更好地反映绳索在松弛状态下的动力学特性。     

新型三阶TVD限制器性能分析

在计算流体力学(CFD)方法中,限制技术是影响计算精度和计算稳定性的重要因素,目前应用较广的经典二阶总变差衰减(TVD)限制器虽能较好地满足计算要求,但性能差异大且分辨率和耗散的性能间并未得到良好权衡。对一种新型的三阶TVD插值限制器(T-3限制器)进行了研究并将其与3种经典限制器进行对比。首先通过一维黎曼问题,得出T-3限制器兼顾较高间断分辨率和良好稳定性的特点;接着通过高超声速双锥绕流和X-33外形飞行器的数值实验,得到T-3限制器具有刻画复杂流动的能力以及较优的气动热计算性能。     

Control of a Spherical Robot: Path Following Based on Nonholonomic Kinematics and Dynamics

This paper presents the controller design for the path following of a spherical mobile robot,BHQ-1.Firstly,a desired velocity for the reference path is deduced from the kinematic model,which cannot be transformed into the classic chained form.Secondly,a necessary torque for the desired velocity is obtained based on the dynamic model.As to the kinematics,a one-dimensional function is selected to measure the two-directional tracking error,and the velocity of rolling forward is reasonably assumed to be constant;therefore the multiple-input multiple-output (MIMO) system is transformed into a single-input single-output (SISO) system.As to the dynamics,both exact dynamics and inexact dynamics with modeling error as well as bounded unknown disturbance are taken into account,based on which a proportional-derivative (PD) controller and a sliding mode controller with adaptive parameters are proposed respectively.Finally,convergence analysis and simulation results are provided to validate these controllers.