Miroslav R. Mataušek

On the modified Smith predictor for controlling a process with an integrator and long dead-time

A dead-time compensator for controlling higher-order processes with integral action and long dead-time is proposed. Tuning formulas are derived. If the velocity gain and the dead-time are estimated experimentally, only one parameter, the time constant defining the speed of the closed-loop setpoint response, has to be tuned manually. The same setpoint response is obtained as in the modified Smith predictor, while the load disturbance rejection is considerably faster.

Optimization of PID Controller Based on Maximization of the Proportional Gain Under Constraints on Robustness and Sensitivity to Measurement Noise

This technical note presents a new, simple and effective, four-parameters proportional-integral-derivative (PID) optimization method. The set of adjustable parameters is defined by the proportional gain k, integral gain ki, damping ratio of the controller zeros (DRCZ), and desired value of the sensitivity to measurement noise Mn. Given Mn and desired value of the maximum sensitivity Ms, for both maximization of k and maximization of ki, only three nonlinear algebraic equations need to be solved for a few values of DRCZ. Contrary to the method based on maximization of ki, in the method based on maximization of k the improvement of performance is obtained by decreasing DRCZ from 1 to the value corresponding to the minimum of the integrated absolute error (IAE). Moreover, this is achieved without deteriorating robustness to the model uncertainties, for a large class of stable processes. Compared to the recently proposed PID optimization methods, for the same Ms and Mn, lower values of IAE and M p are obtained by using the method presented here.

PID Controller Tuning Based on the Classification of Stable, Integrating and Unstable Processes in a Parameter Plane

Classification of processes and tuning of the PID controllers is initiated by Ziegler and Nichols (1942). This methodology, proposed seventy years ago, is still actual and inspirational. Process dynamics characterization is defined in both the time and frequency domains by the two parameters. In the time domain, these parameters are the velocity gain Kv and dead-time L of an Integrator Plus Dead-Time (IPDT) model GZN(s)=Kvexp(-Ls)/s, defined by the reaction curve obtained from the open-loop step response of a process. In the frequency domain these parameters are the ultimate gain ku and ultimate frequency ωu, obtained from oscillations of the process in the loop with the proportional controller k=ku. The relationship between parameters in the time and frequency domains is determined by Ziegler and Nichols as

A fast closed-loop process dynamics characterization.

Stable, integrating and unstable processes, including dead-time, are analyzed in the loop with a known PI/PID controller. The ultimate gain and frequency of an unknown process G(p)(s), and the angle of tangent to the Nyquist curve G(p)(iω) at the ultimate frequency, are determined from the estimated Laplace transform of the set-point step response of amplitude r0. Gain G(p)(0) is determined from the measurements of the control variable and known r0. These estimates define a control relevant model G(m)(s), making possible the use of the previously determined and memorized look-up tables to obtain PID controller guaranteeing desired maximum sensitivity and desired sensitivity to measurement noise. Simulation and experimental results, from a laboratory thermal plant, are used to demonstrate the effectiveness and merits of the proposed method.

A New Predictive PI Controller with Additonal Filtering

Abstract A new predictive PI controller is proposed and applied to the benchmark MIMO PID 2012. Estimate of disturbance, obtained from the Disturbance Observer (DO), is introduced in the loop to obtain the offset free control. First-Order Plus Dead-time (FOPDT) model of stable, integrating and unstable plants is used to design DO, by applying inverse modeling technique. Tuning rules are proposed, and analyzed by simulation of stable, integrating and unstable processes. The high performance of the proposed predictive PI controller with additional filtering is demonstrated by simulation and on the benchmark MIMO PID 2012 plant, with the open-loop dynamics approximated by FOPDT models for the pressure and the water level.

Development and realization of bucket wheel excavator knowledge-based neuro-fuzzy control system

Development of a new control system, which significantly increases excavating capacity, as well as availability, and reliability of the bucket wheel excavator, is presented in this paper. Reference of slewing speed and controller parameters are adapted by predicting cutting resistance of materials to be excavated. The predictive-adaptive higher-level control system is realized as a neuro-fuzzy controller. The fuzzy rules for defining cutting resistance and its correlation with operating conditions, as well as the structure of the new control system are based on expert knowledge and numerous experiments.

Decentralized Automatic Generation Control Based on a Stochastic Inclusion Principle

Abstract An inclusion principle is formulated for stochastic systems, including the LQG optimal design. The obtained results are applied to decentralized overlapping control of large-scale interconnected systems. The efficiency of the proposed methodology is illustrated on a stochastic model of automatic generation control (AGC) for interconnected electric power systems. Two types of decentralized sub-optimal dynamic controllers consisting of state estimators and feedback gains are proposed. An extensive analysis of both steady-state and transient regimes under a variety of operating conditions shows the efficiency of the proposed AGC scheme.

A New Algorithm for Determining Continuous or Discrete Low Order Linear State Variable Models from Measured Data

Abstract A procedure, based on a particular state-space representation suitable for identifying industrial distributed parameter processes and on the generalized equation error approach, is proposed. The only structural parameter of the model is its order n. Increasing n in multiple k of P (n=kp, k=1,2,..) until the output error becomes sufficiently small, k(p 2 +pm) unknown parameters and kp initial conditions of a continuous state variable model are simultaneously determined, p and m being the numbers of the system outputs and inputs, respectively. This continuous model can be easily transformed into the corresponding discrete model. The results obtained using the real power plant input/output data, where the finite order assumption is not fulfilled, are presented and discussed. A comparison between continuous and discrete identification of ΜΙΜΟ systems is also given.