Safe Active Learning and Safe Bayesian Optimization for Tuning a PI-Controller

Constrained Bayesian Optimization with Particle Swarms for Safe Adaptive Controller Tuning

Tuning controller parameters is a recurring and time-consuming problem in control. This is especially true in the field of adaptive control, where good performance is typically only achieved after significant tuning. Recently, it has been shown that constrained Bayesian optimization is a promising approach to automate the tuning process without risking system failures during the optimization pr...

Reliability based multiobjective optimization design procedure for PI controller tuning . ?

In this work, we propose an hybrid multi-objective optimization design procedure for PI controller tuning. This procedure focuses on reliability-based optimization instances, where Montecarlo methods are used to evaluate quantitatively the performance degradation of a controller, due to unexpected or unmodeled dynamics. The procedure is evaluated on a non-linear Peltier process. The presented r...

Tuning methodology for a fuzzy PI controller

A Unified IMC based PI/PID Controller Tuning Approach for Time Delay Processes

This paper proposes a new PI/PID controller tuning method within filtered Smith predictor (FSP) configuration in order to deal with various types of time delay processes including stable, unstable and integrating delay dominant and slow dynamic processes. The proposed PI/PID controller is designed based on the IMC principle and is tuned using a new constraint and without requiring any approxima...

Bayesian Optimization with Safety Constraints: Safe and Automatic Parameter Tuning in Robotics

Robotics algorithms typically depend on various parameters, the choice of which significantly affects the robot’s performance. While an initial guess for the parameters may be obtained from dynamic models of the robot, parameters are usually tuned manually on the real system to achieve the best performance. Optimization algorithms, such as Bayesian optimization, have been used to automate this ...