Feedforward control of the transverse strip profile in hot-dip galvanizing lines

Hot-dip galvanizing is a standard technology to produce coated steel strips. The primary objective of the galvanizing process is to establish a homogeneous zinc layer with a defined thickness. One condition to achieve this objective is a uniform transverse distance between the strip and the gas wiping dies, which blow off excessive liquid zinc. Therefore, a flat strip profile at the gas wiping dies is required. However, strips processed in such plants often exhibit residual curvatures which entail unknown flatness defects of the strip. Such flatness defects cause non-uniform air gaps and hence an inhomogeneous zinc coating thickness. Modern hot-dip galvanizing lines often use electromagnets to control the transverse strip profile near the gas wiping dies. Typically, the control algorithms ensure a flat strip profile at the electromagnets because the sensors for the transverse strip displacement are also located at this position and it is unfeasible to mount displacement sensors directly at the gas wiping dies. This brings along that in general a flatness defect remains at the gas wiping dies, which in turn entails a suboptimal coating.In this paper, a model-based method for a feedforward control of the strip profile at the position of the gas wiping dies is developed. This method is based on a plate model of the axially moving strip that takes into account the flatness defects in the strip. First, an estimator of the flatness defects is developed and validated for various test strips and settings of the plant. Using the validated mathematical model, a simulation study is performed to compare the state-of-the-art control approach (flat strip profile at the electromagnets) with the optimization-based feedforward controller (flat strip profile at the gas wiping dies) proposed in this paper. Moreover, the influence of the distance between the gas wiping dies and the electromagnets is investigated in detail.     

Boole函数的线性可分性(Ⅰ)——n-维超立方体的基本理论

Boole函数的线性可分性是前向人工神经网络理论中的一个比较困难的问题之一。目前仅对变量数n≤7的某些问题给予讨论。当n≥8时,尚无判别Boole函数线性可分的一般准则,更无线性可分Boole函数的计数公式。基于此,本文详细地研究了与Boole函数线性可分性有关的n-维超立方体的基本理论,包括n-维超立方体的基本性质、超立方体中的平行线、子超立方体的计数等,并给出了构造n-维超立方体图的一种新方法。     

提高前馈神经网络推广能力的若干实际方法

提高前馈神经网络的推广能力是深受关注的问题，本文根据我们最近提出了一个网络有效推广的准则，从提高网络特征提取能力，分类能力和修改神经元激活函数等几方面给出了若干实际方案，我们在任意手写数字识别问题上的实验结果证实了这些方法的有效性。     

Tracking trajectories of the cart-pendulum system

The problem of tracking a periodic trajectory of the well-known cart-pendulum system is solved. After a change of coordinates and a change of feedback, the equations of this system are nonlinear but feedforward. This property is extensively used to carry out for this system the design of uniformly asymptotically stabilizing time-varying state feedbacks by using the forwarding approach.     

THE EFFECT OF OPERATING VARIABLES ON THE DYNAMICS OF CATALYTIC CRACKING PROCESSES

The dynamic Characteristics of a Fluidized Catalytic Cracking Process are examined over a wide range of operating conditions. A novel Order of Magnitude Approach is introduced to successfully provide physical insight into the cause and effect relationship between operating conditions and dynamic characteristics. It is shown that the original five-dimensional dynamic model is characterized by three fast time constants and two slow ones that dominate the dynamic responses. The two most important time constants are expressed as explicit functions of the operating conditions. These formulas correctly indicate in which parameter regions the open loop response is oscillatory (underdamped) or nonoscillatory (overdamped). Extensive process variables, that are either flow or capacity, related are defined in order to provide an approximate physical meaning for the dynamic modes of the system. It is shown that the two slow modes of the process are related to the enthalpy content of the regenerator and the sensible heat content of the catalyst phase in both the reactor and regenerator.     

A robust two degree-of-freedom controller for systems with both model and measurement uncertainty

This paper presents a design method for robust two degree-of-freedom (DOF) controllers that optimize the control performance with respect to both model uncertainty and signal measurement uncertainty. In many situations, non-causal feedforward is a welcome control addition when closed loop feedback bandwidth limitations exist due to plant dynamics such as: delays, non-minimum phase zeros, poorly placed zeros and poles (Xie, Alleyne, Greer, and Deneault (2013); Xie (2013), etc. However, feedforward control is sensitive to both model uncertainty and signal measurement uncertainty. The latter is particularly true when the feedforward is responding to pre-measured disturbance signals. The combined sensitivity will deteriorate the feedforward controller performance if care is not taken in design. In this paper a two DOF design is introduced which optimizes the performance based on a given estimate of uncertainties. The controller design uses H_∞ tools to balance the controlled system bandwidth with increased sensitivity to signal measurement uncertainties. A successful case study on an experimental header height control system for a combine harvester is shown as an example of the approach.     

基于嵌入式的空间光通信系统前馈补偿系统设计

为满足光通信系统对于目标数据的快速捕获需求,缩小各级通信设备间的动态跟踪误差,设计基于嵌入式技术的空间光通信系统的前馈补偿系统;以并串转换电路作为唯一的电能输出装置,在光通信信号发射模块、波分复用器件、嵌入式光放大器三个通信设备元件的共同调节作用下,确定PWM功率放大器的实时执行能力,达到对前馈型角编码器的调度与应用,实现空间光通信系统前馈补偿结构的搭建;配置Linux嵌入内核的移植协议,多次创建空间光环境下的通信数据根文件,以此为存储标准,计算补偿调制的基础传信率,实现对前馈脉冲波的调制与处理,完成嵌入式通信系统的前馈补偿原理研究;实验结果表明,应用前馈补偿系统后,在光通信参量等于5.0×1011 T、7.0×1011 T、9.0×1011 T的情况下,目标数据的捕获速率均超过7.0 T/mm,满足缩小通信设备间动态跟踪误差的实际应用需求。     

Boundedness and Convergence of Online Gradient Method with Penalty for Linear Output Feedforward Neural Networks

This paper investigates an online gradient method with penalty for training feedforward neural networks with linear output. A usual penalty is considered, which is a term proportional to the norm of the weights. The main contribution of this paper is to theoretically prove the boundedness of the weights in the network training process. This boundedness is then used to prove an almost sure convergence of the algorithm to the zero set of the gradient of the error function.     

Feedforward model predictive control

This paper examines the role played by feedforward in model predictive control (MPC). We contrast feedforward with preview action. The latter is standard in model predictive control, whereas feedforward has been rarely, if ever, used in contemporary formulations of MPC. We argue that feedforward can significantly improve performance in the presence of measurement noise and certain types of model uncertainty.