Automatic Receding Horizon Optimal Control of the Natural Ventilation Process in Cattle Barns

Abstract

The design and simulation results of an automatic receding horizon optimal controller (RHOC), that controls the ventilation and the associated climate in cattle barns through variable valve openings, are presented. The design is based on so called comfort parameters for cattle in barns and a dynamic model, originally developed using a finite-element programming package, that describes the natural ventilation and the associated climate in cattle barns. In addition to the model the RHOC uses the so called 'lazy man weather prediction' to control the ventilation and the associated climate in the barn. The 'lazy man weather prediction', which assumes the future weather to stay equal to the most recent measurements, is to be preferred over other weather predictions. The choice of the optimisation horizon length and the implementation of the RHOC are discussed and illustrated. The simulation results illustrate the high potential of an RHOC to control the ventilation and the associated climate in cattle barns. When compared to optimal open loop control over the full horizon the loss of perfonnance of the closed loop RHOC, caused by the receding horizon, is limited to about 5 percent. In practice closed loop control is necessary to counteract disturbances and modelling errors.