NMPC Design and Embedded Application for Overhead Crane: Case Study

This paper presents a nonlinear model predictive control. Within this proposal, the issue of choosing the horizon of prediction and control is described. The choice of objective function and the choice of equality and inequality constraints are also discussed. As part of the solution of the presented optimal control problem, we use the direct transcription method, which allows the introduction of constraints for individual states already at the level of constraints for the design variable and thus greatly simplifies implementation. The implementation on a single board computer is presented. The sources used are chosen with the aim of subsequent application of the entire solution within industrial practice. The functionality of the concept is demonstrated and verified using a processor in the loop simulation. The final discussion outlines other possible modifications to enable the use of the MPC concept in industry, on systems that require medium-small sampling periods.