Abstract:Control algorithms such as model predictive control (MPC) and state estimators rely on a number of different parameters. The performance of the closed loop usually depends on the correct setting of these parameters. Tuning is often done manually by experts based on a simulation model of the system. Two problems arise with this procedure. Firstly, experts need to be skilled and still may not be able to find the optimal parametrization. Secondly, the performance of the simulation model might not be able to be carried over to the real world application due to model inaccuracies within the simulation. With this contribution, we demonstrate on an industrial milling process how Bayesian optimization can automate the tuning process and help to solve the mentioned problems. Robust parametrization is ensured by perturbing the simulation with arbitrarily distributed model plant mismatches. The objective is to minimize the expected integral reference tracking error, guaranteeing acceptable worst case behavior while maintaining real-time capability. These verbal requirements are translated into a constrained stochastic mixed-integer black-box optimization problem. A two stage min-max-type Bayesian optimization procedure is developed and compared to benchmark algorithms in a simulation study of a CNC machining center. It is showcased how the empirical performance model obtained through Bayesian optimization can be used to analyze and visualize the results. Results indicate superior performance over the case where only the nominal model is used for controller synthesis. The optimized parametrization improves the initial hand-tuned parametrization notably.

Optimization of the Parameters of a Model Predictive Control System for an Industrial Fractionator

Generalized Predictive Control For A Precise Crude Oil Blending Process

Simultaneous design and control of polyethylene process based on uncertainty Kriging model

Dynamic Modeling And Nonlinear Predictive Control Based On Partitioned Model And Nonlinear Optimization

Research on Tuning Parameters for a Model Predictive Controller Based on CSA in CSTR Process

Get the design right, the rest is automatic: Towards an adaptive MPC controller

Key Technology and Industrial Application of Model Predictive Control Engineering Software

Some notes on MPC relevant identification

A Fuzzy Approximation for FCS-MPC in Power Converters

Generalized Parameter Estimation Method for Model-Based Real Time Optimization

Robust Parametrization of a Model Predictive Controller for a CNC Machining Center Using Bayesian Optimization

Robust explicit model predictive control for hybrid linear systems with parameter uncertainties

Fast Calibration of a Robust Model Predictive Controller for Diesel Engine Airpath

Optimization of the Model Predictive Control Meta-Parameters Through Reinforcement Learning

Parametric Optimization Based MPC for Systems of Systems with Affine Coordination Constraints

Distributed Economic Model Predictive Control with Pseudo-steady State Modifier Adaptation for an Industrial Fluid Catalytic Cracking Unit

An Improved Robust Model Predictive Control Approach to Systems with Linear Fractional Transformation Perturbations

Model Predictive Control Design for Polytopic Uncertain Systems by Synthesising Multi-Step Prediction Scenarios.

Optimal Control for Refinery Synthesis Inventory Management

Predictive control optimization based PID control for temperature in an industrial surfactant reactor

A Multi-Step Robust Model Predictive Control Scheme for Polytopic Uncertain Multi-Input Systems