A framework for hybrid model predictive control in mineral processing

Model Predictive Control (MPC) is an advanced technique for process control that has seen a significant and widespread increase in its use in the process industry since its introduction. In mineral processing, in particular, several applications of conventional MPC can be found for the individual processes of crushing, grinding, flotation, thickening, agglomeration, and smelting with varying degrees of success depending on the variables involved and the control objectives. Given the complexity of the processes normally found in mineral processing, there is also great interest in the design and development of advanced control techniques which aim to deal with situations that conventional controllers are unable to do. In this aspect, Hybrid MPC enables the representation of systems, incorporating logical variables, rules, and continuous dynamics. This paper firstly presents a framework for modeling and representation of hybrid systems, and the design and development of hybrid predictive controllers. Additionally, two application examples in mineral processing are presented. Results through simulation show that the control schemes developed under this framework exhibit a better performance when compared with conventional expert or MPC controllers, while providing a highly systematized methodology for the analysis, design, and development of hybrid MPC controllers.     

Control relevant model reduction of Volterra series models

This paper presents a two-step method for control-relevant model reduction of Volterra series models. First, using the nonlinear IMC design as a basis, an explicit expression relating the closed-loop performance to the open-loop modeling error is obtained. Secondly, an optimization problem that seeks to minimize the closed-loop error subject to the restriction of a reduced-order model is posed. By showing that model reduction of kernels with different degrees can be decoupled in the problem formulation, the optimization problem is simplified into a mathematically more convenient form which can be solved with significantly less computational effort. The effectiveness of the proposed method is illustrated on a polymerization reactor example where a second-order Volterra model with 85 parameters is reduced to a Hammerstein model with 3 parameters. Despite the lower ‘open-loop’ predictive ability of the control-relevant model, the closed-loop performance of the reduced-order control system closely mimics that of the full order model.     

Process identification, uncertainty characterisation and robustness analysis of a pilot-scale distillation column

In this paper different approaches for developing robust advanced control techniques are investigated. A pilot-scale distillation column connected to an industrial distributed control system (ABB MOD 300) that in turn has been interfaced to a VAX-cluster through an Ethernet Gateway is used as a pseudo-industrial set-up to perform these studies. A novel robust multivariable, low order, high performance, model based controller was designed and implemented as a standard PID block within the distributed control system. To provide a systematic approach for designing such an advanced robust controller, several techniques such as dynamic modelling, system identification, uncertainty identification and characterisation etc., are incorporated. The problem of uncertainty characterisation is fully addressed from both theoretical and practical point of view. Both structured and highly structured uncertainty characterisation approaches are used to investigate the robust stability and performance of the control system. Several practical techniques are proposed for designing a robust model-based controller that are readily applicable in an industrial environment. The paper is accompanied by several simulations and also experimental evidences which demonstrate the effectiveness of the proposed approach. ©