Igor Belai

Noise attenuation motivated controller design. Part II: Position control

This paper treats a noise attenuation motivated position controller design outlined in [1]. This modular approach to a filtered PD and a disturbance observer based filtered PID (FPD and DO-FPID) control design, including an experimental evaluation of an optimal filter degree choice, is extended to a constrained control grounded in the invariant set approach [2]-[4]. Loop performance is evaluated by recently introduced measures for deviations from monotonic and two-pulse shapes of transients typical for control of plants with dominant 2nd order dynamics. The analysis shows that a simplified disturbance observer (DO) based constrained filtered PID controller (DO-CFPID) derived for a double integrator plant model gives an interesting performance also for constrained integral systems with a stable mode. It remains simple and it offers an excellent performance also from the point of view of the noise attenuation versus speed of transients.

Experimental evaluation of a DO-FPID controller with different filtering properties

Abstract The paper brings experimental results and evaluation to modular disturbance observer based filtered PD and PID controller design based on theoretical results of papers Huba (2013b,c). The approach is focused on achieving the possibly best filtering properties by keeping nearly constant dynamics of the setpoint responses. The developed controller applied to a positional DC motor control is evaluated for different values of the tuning parameter T D and different filter orders n by using time and shape related performance measures.

Noise attenuation motivated controller design. Part I: Speed control

The paper deals with an experimental evaluation of a modular design of the disturbance observer (DO) based filtered PI control with scalable filtering properties (DO-FPI). Based on simple one-parameter tuning methods it compares outcomes of its application to the speed servo control using incremental position sensor towards the alternative solution based on the two-degree-of-freedom (2DOF) PI control. Under the first tuning scenario, noise filters added to the DO and the P-controller paths may take arbitrary order by simultaneously keeping fixed chosen disturbance response dynamics. This enables a simple noise attenuation adjustment and a significant performance improvement both in the setpoint response, as well as in the noise attenuation. Performance improvements towards the 2DOF PI control reaching to double-figures were confirmed also under the second tuning scenario based on keeping an equal noise attenuation corresponding to a constant impact of a torque ripple.

The positional servo drive with the feedforward control and the noise attenuation

The paper deals with the design and with the comparison of properties of positional servo drive with the three control structures: 1. A filtered disturbance observer based PID control (DO-FPID). 2. The cascaded control with proportional positioning controller and PI speed controller (P-PI). 3. The cascaded control with proportional positioning controller, the PI speed controller and with filtration of actual speed (P-FPI). It shows that whereas in terms of IAE (Integral of absolute error) all solutions may yield an equivalent closed loop dynamics, the controlls with the feedback filters guarantee a much better quantization noise attenuation.

Optimal tuning for a double integrator plus dead time

This paper deals with an optimal disturbance observer (DO) based filtered PID (DO-FPID) controller design based on the performance portrait method solved for a PD controller + double integrator + dead time loop configuration. By virtue of an equivalence of a dead time and n-th order filters used in DO-FPID control, the derived results may be used in a broad range of tasks originating in a control of plants with a dominant 2nd order dynamics, which will be illustrated by a noise attenuation motivated design for a positional DC motor control.

Adaptive DO-FPI controller for the first-order plant

The paper discusses an application of the model reference adaptive control (MRAC) in automatic tuning of the disturbance observer (DO) based filtered PI control with scalable filtering properties (DO-FPI). The control structure with the DO-FPI controller is characterized by a low ripple of the plant input signal. A MRAC algorithm has been designed for the control loop with the first order plant. The proposed algorithm is not sensitive to the amplitude of the setpoint. The settling time of parameter adaptation is almost independent of the control loop dynamics. The proposed solution has been verified on a speed AC drive. The effectiveness of the proposed method is shown through the experimental results.

The PROFIBUS Protocol Observation

Abstract PROFIBUS is a very complex protocol created in 3 layers of the reference communication model. The paper shows experimental way to explain the behaviour of devices on bus under normal traffic and diagnostic events. The paper deals with two communication experiments performed on the same devices on the bus. For measurements on PROFIBUS DP physical layer the professional tester was used. It provides an overview of the signal conditions on the bus and detects the signal quality of each device on the bus. A universal tool for PROFIBUS protocol analysis can perform online bus traffic monitoring. It is able to detect and analyse the PROFIBUS frames for education and diagnosis purposes.

Limits of a simplified controller design based on integral plus dead time models

This article presents design and evaluation of filtered proportional–integral controllers and filtered Smith predictor–inspired constrained dead time compensators. Both are based on the integral plus dead time and on the first-order time delayed plant models. They are compared as for tuning simplicity, robustness and noise attenuation. Such a comparison, which presents a robustness test regarding the importance of the internal plant feedback approximation, may be carried out by performance measures built on deviations of the input and output transient responses from their ideal shapes. When combined with integral of absolute error measures of both solution types with the disturbance responses set as nearly equivalent, we can see that the filtered Smith predictor setpoint responses may be significantly faster than the filtered proportional–integral controller responses, more robust and, using higher-order filters, also sufficiently smooth. Furthermore, tuning of the possibly higher-order filters for filtered Smith predictor is simpler. Its overall design is more transparent and straightforward with respect to the control constraints, where the filtered Smith predictor requires some additional anti-windup measures.

A servo-drive with an integral disturbance observer filter. Part 2: Position control

The paper brings second part of an analysis and synthesis of servo-drives with integral low-pass disturbance observer (DO) filters. The loop established by extending a stabilizing PD-controller by a binomial noise attenuation filter and by a disturbance reconstruction and compensation is devoted to the position control. The dominant loop dynamics represented by an inertial mass is approximated by a second-order integrator. The torque generator dynamics, together with the processing and communication delays are approximated by a dead-time. The solution proposed is compared with a DO based solution using binomial filters both in the disturbance reconstruction and in the stabilizing feedback (DO-FPID) and with a traditional cascaded P-PI structure without additional filters. The output position is considered to be measured by an incremental encoder. Impact of the filter parameters and of the sampling period on the resulting control performance are being explored. The solution using combination of both basic types of low-pass filters shows a significantly improved loop performance.The paper analyses a speed servo-drive with integral low-pass filters applied in the disturbance observer (DO) and in a stabilizing loop with a P-controller. The dominant loop dynamics represented by an inertial mass is approximated by a first-order model. The torque generator dynamics, together with the processing and communication delays are approximated by a dead-time. Two compared approaches focus on reconstruction and compensation of external load disturbances by integral low-pass filters implemented by discrete-time finite-impulse-response (FIR) filters. One of them uses as a disturbance-observer input the generated control signal. The second one has the measured output position at its input. This is considered to be measured by an incremental encoder. Impact of the filter parameters and of the sampling period on the resulting control performance are being explored and compared with a DO using binomial filter and the traditional IP control without an additional filtration.

Extended Matlab based interactive model of a speed servo drive

An extension and modification of a recently developed interactive MATLAB based simulation tool is presented that may be used for verifying properties of several speed control loops for electrical drives with a DC motor and speed estimation based on output differentiation of an incremental position sensor. The compared loops include two degree of freedom (2DOF) PI control, its filtered version denoted as 2DOF FIP control and two solutions with a disturbance observer (DO) using either low pass filter in the velocity P controller loop (DO-FPI), or a DO with a disturbance feedforward and a disturbance reference model (DORM-PI). As a default option the tool provides parameter tuning of all speed controllers for a nearly equal integral of absolute error (IAE) of disturbance (load) steps. Differences in the noise impact are highlighted by relative total variance of the input and output signals. The interactive tool enables users an experimental way of discovering properties of particular loops. It exhibits high importance of appropriate filter design and shows that the newly added solution with a disturbance feedforward and a disturbance reference model may yield for both the nominal and uncertain situations the highest performance to noise impact (torque ripple) ratio.