Boundary geometric control of a linear Stefan problem

This paper addresses the geometric control of the position of a liquid–solid interface in a melting process of a material known as Stefan problem. The system model is hybrid, i.e. the dynamical behavior of the liquid-phase temperature is modeled by a heat equation while the motion of the moving boundary is described by an ordinary differential equation. The control is applied at one boundary as a heat flux and the second moving boundary represents the liquid–solid interface whose position is the controlled variable. The control objective is to ensure a desired position of the liquid–solid interface. The control law is designed using the concept of characteristic index, from geometric control theory, directly issued from the hybrid model without any reduction of the partial differential equation. It is shown by use of Lyapunov stability test that the control law yields an exponentially stable closed-loop system. The performance of the developed control law is evaluated through simulation by considering zinc melting.     

Modeling of extrusion coating for applications in electronics manufacturing

This paper focuses on the development of a model of the extrusion coating process that can be used to generate an automated (as opposed to manual) procedure for determining the control setpoints for a desired deposition thickness and range of deposition materials. This model includes the dynamics of the bead formation process and the bead variation during the coating. The “bead” refers to the deposition material that collects in front and under the die head as the substrate is moved beneath it. Understanding and modeling the dynamics of the bead process as a function of the control‐variable settings is crucial in obtaining an automated procedure for setpoint determination in extrusion coating.     

基于自适应递归模糊神经网络的污水处理控制

针对污水处理过程中具有的非线性、大时变等特征,提出了一种基于自适应递归模糊神经网络(recurrent fuzzy neural network,RFNN)的污水处理控制方法.该方法利用自适应RFNN识别器建立污水处理过程的非线性动态模型,建立的模型可以为RFNN控制器提供污水处理过程中的状态变量信息,保证了控制器根据系统响应调整操作变量的精确性;并且RFNN辨识器及RFNN控制器基于自适应学习率进行学习,确保了递归模糊神经网络的收敛精度和速度,并通过构造李雅普诺夫函数证明了此算法的收敛性;最后,基于基准仿真模型(benchmark simulation model 1,BSM1)平台进行仿真实验.结果表明,与PID、模型预测控制及前馈神经网络相比,该方法对污水处理中溶解氧浓度和硝态氮浓度的跟踪控制精度具有明显的提升.     

Automatic control of the simulated moving bed process for C8 aromatics separation using asymptotically exact input/output-linearization

An approach to automatic control of the simulated moving bed process (SMB) applied to the separation of C₈ aromatics is presented. The principle of asymptotically exact input/output-linearization is used. The controller is based on a nonlinear state estimator using the true moving bed model (TMB). The estimator receives measurement data from four spectroscopic measurement cells. The problem of moving measurement positions with respect to the TMB model is addressed. An exactly linearizing feedback of the estimated states is designed using the nonlinear TMB model equations. The performance of the controller is shown in simulations using a detailed SMB model as a representative of the real process.     

Control of coating properties of LDPE through melt strength measurements

Control of melt strength is proposed to indirectly regulate the neck-in in a twin screw extrusion process producing pellets of low density polyethylene. The melt strength is measured by passing the polymer strand leaving the extruder, through a system of pulleys connected to a balance. Two types of controllers were used, a conventional proportional integral (PI) and an model predictive controller (MPC). Both the simulation and experimental results indicated that the MPC outperformed the PI controller. The reason is that the model relating the manipulated variable (screw speed) to the controlled variable (melt strength) is of high order. MPC is especially suitable for controlling these types of systems.     

Erratum to “A stochastic controller for a scalar linear system with additive Cauchy noise” [Automatica 50(1) (2014) 114–127]

A stochastic control scheme is developed for scalar, discrete-time, and linear-dynamic systems driven by Cauchy distributed process and measurement noises. When addressing the optimal control problem for such systems, the standard quadratic cost criteria cannot be used. In this study we introduce a new objective function that is functionally similar to the Cauchy probability density function. The performance index, defined as the expectation of this objective function with respect to the Cauchy densities, exists. The dynamic programming solution to the fixed and finite horizon optimal control problem that uses this performance index appears to be intractable. Therefore, a moving horizon optimal model predictive control problem is implemented, for which the conditional expected value of the objective function and its gradients can be computed in closed   form and without assumptions such as certainty equivalence. Numerical results are shown for this m$m$-step model predictive optimal controller and compared to a similar, Linear–Exponential–Gaussian model predictive controller. An essential difference between the Cauchy and Gaussian controllers when applied to a system with Cauchy noises is that, while the Gaussian controller is linear and reacts strongly to all noise pulses, the Cauchy controller can differentiate between measurement and process noise pulses by ignoring the former while responding to the latter. This property of the Cauchy controller occurs when an impulsive measurement noise is more likely than an impulsive process noise. The Cauchy and Gaussian controllers react similarly when applied to a system with Gaussian noises, demonstrating the robustness of the proposed control scheme.     

A stochastic controller for a scalar linear system with additive Cauchy noise

A stochastic control scheme is developed for scalar, discrete-time, and linear-dynamic systems driven by Cauchy distributed process and measurement noises. When addressing the optimal control problem for such systems, the standard quadratic cost criteria cannot be used. In this study we introduce a new objective function that is functionally similar to the Cauchy probability density function. The performance index, defined as the expectation of this objective function with respect to the Cauchy densities, exists. The dynamic programming solution to the fixed and finite horizon optimal control problem that uses this performance index appears to be intractable. Therefore, a moving horizon optimal model predictive control problem is implemented, for which the conditional expected value of the objective function and its gradients can be computed in closed   form and without assumptions such as certainty equivalence. Numerical results are shown for this m$m$-step model predictive optimal controller and compared to a similar, Linear-Exponential-Gaussian model predictive controller. An essential difference between the Cauchy and Gaussian controllers when applied to a system with Cauchy noises is that, while the Gaussian controller is linear and reacts strongly to all noise pulses, the Cauchy controller can differentiate between measurement and process noise pulses by ignoring the former while responding to the latter. This property of the Cauchy controller occurs when an impulsive measurement noise is more likely than an impulsive process noise. The Cauchy and Gaussian controllers react similarly when applied to a system with Gaussian noises, demonstrating the robustness of the proposed control scheme.     

A digital quality control system for an industrial dry process rotary cement kiln

A multivariate autoregressive moving average (ARMA) model for an industrial dry process rotary cement kiln is identified, from real process data, using the maximum likelihood method. The model obtained is then used in computing a controller for quality control of clinker production. It is shown that it is relevant to compute a minimum variance controller subject to restrictions both in the controller structure and the variances of the control signals. The resulting controller is finally implemented on the cement kiln process, together with a target adaptive controller for automatic adjustment of the clinker quality setpoint, in order to save energy.     

Latent variable iterative learning model predictive control for multivariable control of batch processes

A latent variable iterative learning model predictive control (LV-ILMPC) method is presented for trajectory tracking in batch processes. Different from the iterative learning model predictive control (ILMPC) model built from the original variable space, LV-ILMPC develops a latent variable model based on dynamic partial least squares (DyPLS) to capture the dominant features of each batch. In each latent variable space, we use a state–space model to describe the dynamic characteristics of the internal model, and an LV-ILMPC controller is designed. Each LV-ILMPC controller tracks the set points of the current batch projection in the corresponding latent variable space, and the optimal control law is determined and the persistent process disturbances is rejected along both time and batch horizons. The proposed LV-ILMPC formulation is based on general LV-MPC and incorporates an iterative learning function into LV-MPC. In addition, the real physical input that drives the process can be reconstructed from the latent variable space. Therefore, this algorithm is particularly suitable for multiple-input, multiple-output (MIMO) systems with strong coupling and serious collinearity. Three studies are used to illustrate the effectiveness of the proposed LV-ILMPC .     

Discrete optimal stochastic controller and its forms

An approach to obtain the optimal controller From an optimal criterion is presented. The controller is very general in the sense that it is multivariable, it compensates Tor process delay and it allows a penalty on the input variable variance. The controller can be obtained in two forms: the state-space model form with the steady-state Kalman filter that separates the control algorithm in distinctive steps to provide an insight into the control problem, while the transfer function form is very convenient for implementation.     

考虑运动受限的履带式移动机器人轨迹跟踪控制

针对履带式移动机器人的轨迹跟踪控制问题进行研究,首先,建立了履带式移动机器人的运动学模型和跟踪误差模型;其次,设计了转速有限时间控制和线速度滑模控制的轨迹跟踪控制律,并给出了考虑运动受限作用下的控制律修正表达式;最后,基于MATLAB对所提控制律进行仿真,对比分析了不考虑运动受限情况下跟踪控制效果;结果表明,设计的跟踪控制律能够实现履带式移动机器人对圆轨迹的有效跟踪,且考虑运动受限作用的控制律更加符合实际;文章研究分析了运动受限作用对于移动机器人轨迹跟踪控制的影响,分析结果对其他移动机器人的运动控制研究具有参考价值。     

Bioreactor profile control by a nonlinear auto regressive moving average neuro and two degree of freedom PID controllers

This paper presents the use of nonlinear auto regressive moving average (NARMA) neuro controller for temperature control and two degree of freedom PID (2DOF-PID) for pH and dissolved oxygen (DO) of a biochemical reactor in comparison with the industry standard anti-windup PID (AWU-PID) controllers. The process model of yeast fermentation described in terms of temperature, pH and dissolved oxygen has been used in this study. Nonlinear auto regressive moving average (NARMA) neuro controller used for temperature control has been trained by Levenberg–Marquardt training algorithm. The 2DOF-PID controllers used for pH and dissolved oxygen have been tuned by MATLAB's auto tune feature along with manual tuning. Random training data with input varying from 0 to 100 l/h have been obtained by using NARMA graphical interface. The data samples used for training, validation and testing are 20,000, 10,000 and 10,000 respectively. Random profiles have been used for simulation. The NARMA neuro controller and the 2DOF-PID controllers have shown improvement in rise time, residual error and overshoot. The proposed controllers have been implemented on TMS320 Digital Signal Processing board using code composure studio. Arduino Mega board has been used for input/output interface.     

基于动态目标位置的智能车辆动态避障控制研究

为了真实地模拟驾驶员在动态环境中避让动态障碍物的行为方式,提出了动态目标位置概念,并采用三次样条曲线作为动态避障的路径拟合曲线。以模糊逻辑为控制策略,以T-S模糊模型为控制结构,以自适应神经网络为隶属度函数的参数调整手段,设计出一种智能车辆横向运动控制器,并通过计算机仿真实现。结果表明,基于动态目标位置概念的控制器设计具有较好的控制性能,较为理想地模拟实际交通环境中车辆动态避障的特性。     

Detection of model-plant mismatch in closed-loop control system

The performance of model-based control systems depends a lot on the process model quality, hence the process model-plant mismatch is an important factor degrading the control performance. In this paper, a new methodology based on a process model evaluation index is proposed for detecting process model mismatch in closed-loop control systems. The proposed index is the ratio between the variance of the disturbance innovation and that of the model quality variable. The disturbance innovations are estimated from the routine operation data by an orthogonal projection method. The model quality variable can be obtained using the closed-loop data and the disturbance model estimated by adaptive Least absolute shrinkage and selection operator (Lasso) method. When the order of the disturbance model is less than 2 or the process time delay is large enough, no external perturbations are required. Besides, the proposed index is independent of the controller tuning and insensitive to the changes in disturbance model, which indicates that the proposed method can isolate the process model-plant mismatch from other factors affecting the overall control performance. Three systems with proportional integral (PI) controller, linear quadratic (LQ) controller and unconstrained model predictive control (MPC) respectively are presented as examples to verify the effectiveness of the proposed technique. Besides, Tennessee Eastman process shows the proposed method is able to detect process model mismatch of nonlinear systems.     

起重机升降运动系统自适应Backstepping控制研究

通过分析起重机升降运动系统各组成部分的关联关系,建立了起重机升降运动系统数学模型,从而将电动机的运动状态和负载的运动状态有效地统一起来;以该模型为基础,将负载运动位置作为控制变量,采用自适应Backstepping控制策略设计了一种起重机升降运动系统负载位置控制器。仿真结果表明,该控制器具有较快的响应速度和较小的跟踪误差,能够有效抑制负载突变和参数摄动的影响,具有良好的鲁棒性。     

一种基于BP组件的学习控制系统及其在倒立摆控制上的应用

给出了一个组件化方法设计学习控制系统的一个实例.学习控制系统建立在两个BP组件——BP模型和BP控制器的基础上,通过双通道反向学习的方法在控制过程中进行自我调整,适应控制对象的变化以及模型和控制器本身的不同条件.首先介绍了BP组件的接口和功能规范.然后建立基于BP组件的学习控制系统的组件化框架.最后给出一个基于BP组件的学习控制系统在倒立摆控制上的应用.     

Application of fuzzy-based Taguchi method to the optimization of extrusion of magnesium alloy bicycle carriers

By using fuzzy-base Taguchi method, this study investigates the optimal process parameters that maximize multiple performance characteristics index (MPCI) for hot extrusion of AZ31 and AZ61 magnesium alloy bicycle carriers. The larger-the-better quality characteristics of flattening strength and T-slot fracture strength as well as the smaller-the-better quality characteristic of extrusion load is considered in the MPCI. Through MPCI inference model, a manufacture method with less extrusion load and better mechanical properties under hot extrusion can be obtained. The signal-to-noise (S/N) ratios of the three quality characteristics??flattening strength, T-slot fracture strength and extrusion load??are calculated for the products based on experimental results. And the S/N ratio serves as the input variable to fuzzy control unit, and MPCI is a single output variable. The obtained MPCI is used to analyze optimal process parameters. This study finds combination of process parameters that optimizes MPCI, and conducts confirmatory experiments to prove the accuracy of optimal combination of process parameters thus selected. Finally, mechanical properties of AZ31, AZ61 magnesium alloy and A6061 aluminum alloy bicycle carriers are tested to identify differences among these three materials.     

Robust forecasts and run-to-run control for processes with linear drifts

Processes experiencing linear drift over time are usually forecasted using a double exponentially weighted moving average (d-EWMA) filter. d-EWMA has incorrectly been claimed as an optimal filter for the integrated moving average (2,2) (IMA(2,2)) process, which is a stochastic equivalent of a process with linear drifts (ramps). It is shown that the optimal filter for such a process has a different structure but can be put in a similar form with same effective tuning parameters. The problem of batch-to-batch process gain variation (with known bounds) has been addressed by using a robust run-to-run control algorithm. This algorithm solves a minimax problem that determines the next run input adjustment by minimizing the worst-case predicted error. The conditions for equivalence of the minimax controller to a nominal model inverse based controller for a simple SISO system based on the type of model used and the nature of bounds have been investigated. An important implication of the equivalence result for nonlinear systems is pointed out. The proposed robust run-to-run controller formulation is tested on a number of examples including a chemical mechanical polishing (CMP) process model.     

非完整机械控制系统的模型参考变结构控制

在参数扰动和外部干扰情况下,对非完整机械控制系统设计了变结构模型参考跟踪控制器,基于适当的矩阵分解,非线性控制理论中的输入－输出解耦概念及变结构控制理论,为解决干扰非完整机械控制系统的跟踪问题提出了三级控制设计过程,最后通过一非完整机械系统的例子（在给定平面上运动的垂直轮）的计算机仿真说明了提出方法的优越性。     

Variable-Gain Controllers for Nonlinear Systems Using the T–S Fuzzy Model

This correspondence proposes two novel control schemes with variable state-feedback gain to stabilize a Takagi-Sugeno (T-S) fuzzy system. The T-S fuzzy model is expressed as a linear plant with nonlinear disturbance terms in both schemes. In controller I, the T-S fuzzy model is expressed as a linear plant around a nominal plant arbitrarily selected from the set of linear subsystems that the T-S fuzzy model consists of. The variable gain then becomes a function of a gain parameter that is computed to neutralize the effect of disturbance term, which is, in essence, the deviation of the actual system dynamics from the nominal plant as the system traverses a specific trajectory. This controller is shown to stabilize the T-S fuzzy model. In controller II, individual linear subsystems are locally stabilized. Fuzzy blending of individual control actions is shown to make the T-S fuzzy system Lyapunov stable. Although applicability of both control schemes depends on the norm bound of unmatched state disturbance, this constraint is relaxed further in controller II. The efficacy of controllers I and II has been tested on two nonlinear systems