Semi-supervised gear fault diagnosis using raw vibration signal based on deep learning

In aerospace industry, gears are the most common parts of a mechanical transmission system. Gear pitting faults could cause the transmission system to crash and give rise to safety disaster. It is always a challenging problem to diagnose the gear pitting condition directly through the raw signal of vibration. In this paper, a novel method named augmented deep sparse autoencoder (ADSAE) is proposed. The method can be used to diagnose the gear pitting fault with relatively few raw vibration signal data. This method is mainly based on the theory of pitting fault diagnosis and creatively combines with both data augmentation ideology and the deep sparse autoencoder algorithm for the fault diagnosis of gear wear. The effectiveness of the proposed method is validated by experiments of six types of gear pitting conditions. The results show that the ADSAE method can effectively increase the network generalization ability and robustness with very high accuracy. This method can effectively diagnose different gear pitting conditions and show the obvious trend according to the severity of gear wear faults. The results obtained by the ADSAE method proposed in this paper are compared with those obtained by other common deep learning methods. This paper provides an important insight into the field of gear fault diagnosis based on deep learning and has a potential practical application value.     

Attitude control of a satellite with redundant thrusters

Redundant thrusters are generally used for a reliable attitude control system. Also, redundant thrusters yield a better performance if they are used appropriately. In this paper, the authors propose an efficient redundancy management algorithm to reduce the fuel consumption. The algorithm is based on a linear programming problem which is a constrained optimization problem. For the algorithm, a cost function is defined as a quantity related to the fuel consumption for a maneuver. The independent variables are the thrusters' on-times which are control input variables of a satellite dynamic model. The advantage of the proposed method is verified by numerical examples. The examples show that the proposed method consumes less fuel than an existing method for a given maneuvering command. A sub-optimal algorithm is also discussed for an onboard computation. The proposed algorithm is applied to two maneuvers: move-to-rest and rest-to-rest. This is verified by a numerical simulation.     

基于混合灵敏度下H∞控制在三滑块质量矩导弹上的应用

由于质量矩导弹是一个具有非线性和模型不确定性的多变量系统，为了进一步提高控制系统的鲁棒性，对三滑块质量矩导弹进行了鲁棒控制系统的研究。以所建立的质量矩导弹数学模型为基础，通过对模型合理的简化，得到一个耦合的非线性动力学系统，考虑到质量矩导弹的鲁棒性要求和三个滑块的协调控制问题，在混合灵敏度控制的基础上，采用H∞控制理论对导弹进行姿态控制系统的研究，仿真结果验证了这种方法的有效性。     

基于分层式强化学习的移动机器人导航控制

针对未知环境下的移动机器人导航问题,本文提出了一种基于分层式强化学习的混合式控制方法。利用栅格-拓扑相结合的环境表示及地图学习方法,通过分层式强化学习在不同控制层次的扩展设计移动机器人的反应式和慎思式导航控制,实现了全局导航和局部导航控制的协调优化。实验及测试结果证明,该控制方法能实现导航任务的全局优化,避免陷入局部极小,并对未知动态环境具有较强的适应性。     

并联六自由度对接平台运动学标定方法研究

对并联六自由度对接平台的结构参数标走问题进行了研究。通过实测运动平台的位姿变化以及对应六根丝杆的增量，提出了直接确定结构参数的标定方法，与传统标定方法相比，避免了迭代计算与收敛判断。仿真验证表明了所提方法的有效性和实用性。     

线性回归模型岭迹选择主元方法及应用研究

针对线性回归模型的多重复共线性问题，在理论上，引入方差膨胀因子，从量化的角度探测线性回归模型的多重复共线性；并采用岭迹分析法对此类模型的结构辨识方法展开研究，归纳了岭迹选择主元的若干原则，并通过实际计算证实了该方法的有效性。     

非线性动态逆神经元解耦飞行控制方法

 提出一种非线性动态逆神经元控制系统设计方法，并成功的将其应用于某战斗机非线性解耦控制问题。该方法的基本思想是采用神经元网络建立非线性被控对象的动态逆模型，将被控对象转化为伪线性系统，并用现代控制系统综合设计方法对神经元伪线性系统进行闭环优化设计。给出的战斗机非线性动态逆神经元解耦飞行控制系统的仿真结果显示出人工神经元网络作为非线性动态逆控制单元所具有的潜在能力。     

Numerical investigation of the shock interaction effect on the lateral jet controlled missile

A computational study on the supersonic flow around the lateral jet controlled missile has been performed. A three-dimensional Navier–Stokes computer code (AADL 3D) has been developed and case studies have been performed by comparing the normal force coefficient and the moment coefficient of a missile body. Different jet flow conditions including jet pressures and jet Mach numbers, and the circumferential jet positions have been incorporated into the case studies. The missile surface is divided into four regions with respect to the center of gravity, and the normal force and moment distribution at each region are compared. The results show conspicuously different normal force and moment variations according to each parameter variation. From the detailed flow field analyses, it has been verified that most of the normal force loss and the pitching moment generation are taking place at the low-pressure region behind the jet nozzle. Furthermore, it is shown that the pitching moment can be efficiently reduced by the lateral thrust obtained through higher jet Mach number rather than high jet pressure. Thus, an angle of yaw is more effective for missile control by side jet than an angle of attack.     

Optimal positioning of piezoelectric actuators on a smart fin using bio-inspired algorithms

In this paper a novel approach is developed for optimization of piezoelectric actuators in vibration suppression. A scaled model of a vertical tail of F/A-18 is developed in which piezoelectric actuators are bounded to the surface. The frequency response function (FRF) of the system is then recorded and maximization of the FRF peaks is considered as the objective function of the optimization algorithm to enhance the actuator authority on the mode, which assigns the optimal placement of the pair of piezoelectric actuators on the smart fin. Six multi-layer perceptron neural networks are employed to perform surface fitting to the discrete data generated by the finite element method (FEM). Invasive weed optimization (IWO), a novel numerical stochastic optimization algorithm, is then employed to maximize the FRF peak which in due reduces the vibration of the smart fin. Results indicated an accurate surface fitting for the FRF peak data as well as the optimal placement of the piezoelectric actuators for vibration suppression.     

非线性逆在ASTOVL控制系统中的应用

研究了先进短距起飞垂直着陆飞机（ASTOVL）的飞行／推进控制系统的控制律，基于解和掌握气动舵面和推力矢量舵面控制下的飞机数学模型，应用非线性动态逆理论，为ASTOVL飞机从巡航到悬停阶段的飞行，设计了推力矢量控制器和俯仰控制器，以合理地分配飞机上的力和力矩，使得俯仰姿态速度和法向速度的控制在推进系统性能包括线范围内是近似线性和解耦的，最后以ASTOVL飞机悬停阶段的数字仿真曲线为例，证明了所设计的两个控制器是可行的，基本上达到了设计要求。