Control Optimisation of a Domestic Heating System

The paper presents various control techniques for optimization and energy efficiency for domestic heating systems. A description of the heating system model, its behaviour and underpinning processes are presented. The model of the heating system, which is inherently nonlinear, is formulated either as a linear time varying (LTV) system or a bilinear time varying (BTV) system or bilinear time invariant (BTI) system. For each model representation various formulations of generalised predictive control (GPC) are investigated. The original GPC, in its non-incremental and modified forms are implemented and tested for these systems. Depending on the model structure, i.e whether the model is LTV, BTV or BTI, the GPC is formulated as a linear self tuning control (LSTC) GPC, bilinear self tuning control (BSTC) GPC and fixed bilinear (FB) GPC, respectively. It is shown that depending on the adopted model representation various degrees of improvement are possible when using different formulations for GPC.