基于模型预测控制的部分分散控制系统设计(英文)

An expansion procedure to design partially decentralized controllers via model predictive control is proposed in this paper. Partially decentralized control is a control structure that lies between a fully decentralized structure and a fully centralized one, and has the advantage of achieving comparable performance as a fully centralized controller but with simpler structure. The proposed method follows the expansion method proposed in a previous paper where internal model control (IMC) was used to design controllers for non-square subsystems. The method requires computing the pseudo-inverse of a non-square matrix via pseudo-inverse factors. Instead, the proposed method uses dynamic matrix control (DMC) to design PID controllers for non-square subsystems without using additional factors. The effectiveness of the proposed method is demonstrated on several chemical examples. Simulation results show that the proposed method is simple and can achieve better performance.     

Simulated nonlinear control studies of five-effect evaporator models

This paper is devoted to the simulated nonlinear control studies of two dynamic models of an industrial five-effect evaporator of an alumina refinery. The simulated control studies were carried out to ascertain the performance and robustness of nonlinear control techniques on the five-effect evaporator prior to its implementation on-site. The nonlinear control structure used in the studies was the multi-input multi-output (MIMO) globally linearizing control (GLC) structure of Kravaris and Soroush (1990). The I/O linearizing controller was implemented in a cascade arrangement with the industrial coded velocity form of proportional-integral (PI) controllers in the simulation. The design parameters were chosen such that the desired decoupling control of the ill-conditioned evaporator models was achieved, and such that cascade arrangement of the nonlinear controller was possible. Simulated results indicates that the MIMO GLC structure provides superior servo and regulatory control to multi-loop single-input single-output (SISO) PI controllers that are currently being used to regulate the five-effect evaporator on-site.     

Simplified design of proportional-integral-derivative (PID) controller to give a time domain specification for high order processes

An efficient simplified method is proposed for the time domain design of industrial proportional-integral-derivative (PID) controllers and lead-lag compensators for high order single input single output (SISO) systems. The proposed analytical method requires no trial error steps for a lead-lag compensator design in the time domain by using the root-locus method. A practical PID controller design method was obtained based on the corresponding lead-lag compensator to give a required time-domain specification. Simulation studies were carried out to illustrate the control performance of the controllers by the proposed method. The proposed PID controller and lead-lag compensator directly satisfied time domain control specifications such as damping ratio, maximum overshoot, settling time and steady sate error without trial and error steps. The suggested algorithm can easily be integrated with a toolbox in commercial software such as Matlab.     

The design of experiments,training and implementation of nonlinear controllers based on neural networks

In the area of nonlinear predictive control, several control schemes using artificial neural networks have been proposed. In this work, the issues relating to the information contents of the data used to train the neural network components of these nonlinear predictive control schemes are considered. This raises questions about the design of experiments. A class of feedback-feedforward nonlinear controller based on the model predictive structure (also known as Internal Model Control, IMC, structure) is investigated. The implementation and performance of these neural network based controllers, together with comparisons to other nonlinear and linear controllers, are illustrated on two nonlinear continuous-stirred-tank-reactor simulations.     

An overview of discrete stochastic controllers: Generalized PID algorithms with dead-time compensation

Digital computer control systems have enabled use of a wide variety of different algorithms compared with the classical analog proportional-integral-derivative controllers. Design techniques have been presented for many of these in diverse areas of the literature. Some algorithms are designed to give minimum time responses, or to give specified closed loop responses, or to satisfy classical guidelines for stability. By contrast, the design of discrete optimal stochastic controllers is not based on considerations such as stability, frequency response, or dead-time compensation properties, but rather on optimizing some criterion of performance such as the output error variance. Therefore, it is of interest to investigate how these controllers behave with respect to some of the more traditional ways of evaluating or designing controllers. In this paper, both intuitive and theoretical arguments are presented to illustrate the very desirable properties of these controllers and their relationship to many other algorithms. In particular, they are shown to contain proportional-integral-derivative terms and to be related to other well-known controllers, to have optimal compensation for process dead-time and optimal filtering properties, and to have very desirable stability and frequency response characteristics. In this sense, the paper provides a unifying overview of univariate stochastic controllers.     

Control structure comparison for three-product Petlyuk column

The focus of this paper is to investigate different control structures (single-loop PI control) for a dividing wall (Petlyuk) column for separating ethanol, n-propanol and n-butanol. Four control structures are studied. All the results are simulations based on Aspen Plus. Control structure 1 (CS1) is stabilizing control structure with only temperature controllers. CS2, CS3 and CS4, containing also composition controllers, are introduced to reduce the steady state composition deviations. CS2 adds a distillate composition controller (CCDB) on top of CS1. CS3 is much more complicated with three temperature-composition cascade controllers and in addition a selector to the reboiler duty to control the maximum controller output of light impurity composition control in side stream and bottom impurity control in the prefractionator. CS4 adds another high selector to control the light impurity in the sidestream. Surprisingly, when considering the dynamic and even steady state performance of the proposed control structures, CS1 proves to be the best control structure to handle feed disturbances inserted into the three-product Petlyuk column.     

参数不确定非线性系统切换控制

针对一类含有不确定参数的非线性系统,基于其参数的不确定范围,设计多个滑模变结构控制器。在此基础上,基于一给定的指标切换函数构造切换控制器。在确保系统Lyapunov稳定性的前提下,被控对象的控制器按照预先设定好的切换条件,在多个控制器之间相互切换,从而极大地改善系统的瞬态响应。以工业机器人手臂为研究对象,针对机器人手臂运动方程构造滑模变结构控制器,并设计以输出误差为自变量的指标切换函数,基于此切换函数构造切换控制器,使机械手的控制器在多个控制器之间进行切换。针对不同的参数变化范围,研究切换控制的有效性。多个仿真实例表明切换控制能极大地改善控制品质。     

Application of the method of inequalities to the multivariable control of binary distillation columns

The method of inequalities is used to design controllers for dual quality control of four binary distillation columns whose top and bottom compositions are controlled by manipulating reflux rate and steam. Two types of multivariable proportional-plus-integral controllers are designed and the performance of the system with these controllers is compared with that of non-interacting control. The results of implementing the simpler controller, which is represented by a diagonal matrix, on a computer-controlled packed distillation column are also presented.     

Simple mass balance controllers for continuous sedimentation

The rational use of water in the mineral processing industry has become an important issue due to the geographical location of many plants. The increase of capacity in many copper concentrators has lead to an increased effort for recovering the maximum amount of water in the solid–liquid separation process. Thickeners work continuously to produce a concentrated underflow and a water overflow free from particulate matter. The behavior of many processes can be represented by a set of intensive and extensive variables. In this case, practice has shown that standard feedback control based on intensive variables has not been very easy to tune and effective in providing consistent operations. In many plants, thickeners operate with poor standards, with high dosages of flocculants, overflows with high fine particles contents and highly variable underflows. This work presents a novel nonlinear PI controller which is able to stabilize thickener operation using a simple control structure. An internationally accepted model and calibration using plant data is used to illustrate the design methodology and the level of performance attained by the controllers. The analysis of the results points out the improved performance by using extensive variables. In addition some guidelines concerning controllers tuning are also provided.     

Design of decentralized proportional-plus-integral controllers for multivariable systems

This paper presents a simple method for the design of decentralize proportional-plus-integral controllers for multivariable linear systems where the controllers act directly on the available systems outputs. The proposed design scheme, which includes local as well as hierarchical types of control systems, involves the solution of one nonlinear Riccati matrix equation and one linear Liapunov matrix equation. The practicality and good performance of control systems developed, using this approach, is demonstrated on the mathematical model of a plate absorption column.     

工业过程多变量系统的辅助常规控制设计

利用PID控制器对多变量系统进行控制,主要采取分散常规控制或解耦控制的策略。对于存在耦合的系统,分散常规控制无法消除各变量间的相互扰动,而直接设计动态解耦矩阵有时候会存在困难,静态解耦的效果却并不明显。针对已设计好分散常规控制器的方系统,以解耦为目标和原则,设计多变量系统中分散常规控制方案剩余配对的控制器作为辅助控制器,以频域近似的方法整定其PID参数,实现多输入多输出的控制结构,降低了各变量的相互耦合作用,提高了控制性能。通过对Shell公司的典型控制问题的辅助控制器设计与仿真实验,取得了比原有分散常规控制与静态解耦控制更好的控制效果,验证了控制策略的可行性与有效性。     

Performance analysis of three advanced controllers for polymerization batch reactor: An experimental investigation

The performances of three advanced non-linear controllers are analyzed for the optimal set point tracking of styrene free radical polymerization (FRP) in batch reactors. The three controllers are the artificial neural network-based MPC (NN-MPC), the artificial fuzzy logic controller (FLC) as well as the generic model controller (GMC). A recently developed hybrid model (Hosen et al., 2011a. Asia-Pac. J. Chem. Eng. 6(2), 274) is utilized in the control study to design and tune the proposed controllers. The optimal minimum temperature profiles are determined using the Hamiltonian maximum principle. Different types of disturbances are introduced and applied to examine the stability of controller performance. The experimental studies revealed that the performance of the NN-MPC is superior to that of FLC and GMC.     

Computation of multiloop controllers having desired closed-loop responses

For n-by-n multivariable processes, multiloop controllers have n degrees of freedom and hence the n diagonal elements of closed-loop transfer functions can be designed to have desired closed-loop responses. Multiloop controllers having desired closed-loop responses can be considered as an extension of the single-input single-output internal model control and they can be used as reference controllers. However, computations of such multiloop controllers have not been well developed. The Newton-Raphson method and the iterative sequential loop closing method can be used, but they can suffer from a divergence problem for some processes. Here, the continuation method is applied to obtain multiloop control systems with desired closed-loop responses for a robust computation. The multiloop controllers with desired closed-loop responses can be used to obtain dynamic interaction measures and design multiloop PID controllers.     

Feedback control design for an anaerobic digestion process

An approach for the design of linear feedback controllers for anaerobic digestion systems is presented. The effluent chemical oxigen demand (COD) concentration and the dilution rate are taken respectively as the regulated and the manipulated variables. The control design is based on simple step‐response models of the process endowed with an input delay to account for dead‐times induced by measurement devices. The resulting feedback controller has a traditional proportional‐integral (PI) control structure, so it can be easily implemented with conventional control technologies. Since the concentration of volatile fatty acids can be easily and quickly measured as compared with COD concentration, it is used as a secondary measurement that is incorporated into the feedback loop scheme to enhance the robustness of the control scheme with respect of influent disturbances. The performance of the proposed control scheme is illustrated via numerical simulations and experimental work. © 2002 Society of Chemical Industry     

Bounded robust control of constrained multivariable nonlinear processes

This work focuses on control of multi-input multi-output (MIMO) nonlinear processes with uncertain dynamics and actuator constraints. A Lyapunov-based nonlinear controller design approach that accounts explicitly and simultaneously for process nonlinearities, plant-model mismatch, and input constraints, is proposed. Under the assumption that all process states are accessible for measurement, the approach leads to the explicit synthesis of bounded robust multivariable nonlinear state feedback controllers with well-characterized stability and performance properties. The controllers enforce stability and robust asymptotic reference-input tracking in the constrained uncertain closed-loop system and provide, at the same time, an explicit characterization of the region of guaranteed closed-loop stability. When full state measurements are not available, a combination of the state feedback controllers with high-gain state observes and appropriate saturation filters, is employed to synthesize bounded robust multivariable output feedback controllers that require only measurements of the outputs for practical implementation. The resulting output feedback design is shown to inherit the same closed-loop stability and performance properties of the state feedback controllers and, in addition, recover the closed-loop stability region obtained under state feedback, provided that the observer gain is sufficiently large. The developed state and output feedback controllers are applied successfully to non-isothermal chemical reactor examples with uncertainty, input constraints, and incomplete state measurements. Finally, we conclude the paper with a discussion that attempts to put in perspective the proposed Lyapunov-based control approach with respect to the nonlinear model predictive control (MPC) approach and discuss the implications of our results for the practical implementation of MPC, in control of uncertain nonlinear processes with input constraints.     

改进Smith预估控制器的解析设计

针对化工过程常见的积分和不稳定时滞对象,基于改进的史密斯预测控制提出了两自由度控制方案。首先根据鲁棒控制理论H2最优性能指标设计设定值跟踪控制器,然后在分析稳定性和抗扰性的基础上设计了扰动抑制控制器,设定值跟踪控制器和扰动抑制控制器可通过性能参数独立调节和优化。同时针对存在的乘性不确定性过程分析了系统的鲁棒稳定性。最后通过仿真实例验证了该控制方案的有效性。     

Assessing the Performance of Model Predictive Controllers

Performance assessment of model predictive controllers is a problem of significant industrial relevance. Model predictive controllers belong to a class of linear time‐varying controllers, which compute the future control actions by minimizing a constrained, time‐varying objective function. In this work we propose a performance statistic that takes into account the time‐varying and constrained nature of model predictive control. The proposed measure compares the achieved objective function with its design value, online. Analytical expressions are derived to calculate the expected value of the design objective function under closed loop conditions. Simulation and industrial case studies are used to illustrate the applicability of the proposed metric.     

Optimization and Control of Pressure Swing Adsorption Processes Under Uncertainty

The real‐time periodic performance of a pressure swing adsorption (PSA) system strongly depends on the choice of key decision variables and operational considerations such as processing steps and column pressure temporal profiles, making its design and operation a challenging task. This work presents a detailed optimization‐based approach for simultaneously incorporating PSA design, operational, and control aspects under the effect of time variant and invariant disturbances. It is applied to a two‐bed, six‐step PSA system represented by a rigorous mathematical model, where the key optimization objective is to maximize the expected H₂ recovery while achieving a closed loop product H₂ purity of 99.99%, for separating 70% H₂, 30% CH₄ feed. The benefits over sequential design and control approach are shown in terms of closed‐loop recovery improvement of more than 3%, while the incorporation of explicit/multiparametric model predictive controllers improves the closed loop performance. © 2012 American Institute of Chemical Engineers AIChE J, 59: 120–131, 2013     

蒸馏塔分散PID控制器整定研究

结合一种双输入双输出 (TITO)分散PID控制器的设计方法 ,并将其应用于乙醇—水的蒸馏分离过程控制中。该法根据相应的收敛性算法 ,通过一系列继电闭环实验辨识预期临界点 (DCP) ,利用T L整定公式获取分散PID控制器参数。通过仿真对比试验 ,表明所设计的控制系统具有满意的动态性能     

Controller adjustment for improved nominal performance and robustness—I. A frequency domain approach

A frequency domain method, which makes it possible to adjust multivariable controllers with respect to both nominal performance and robustness, is presented. The basic idea in the approach is that the designer assigns objectives such as steady-state tracking, maximum resonance peaks, bandwidth, minimum stability margin, steady-state sensitivity and maximum sensitivity to modelling errors. For a given control structure the parameters are found which minimize an objective function consisting of a weighted sum of deviations between desired and obtained values of these objectives.