aalto1 untyped-item.component.html
Model predictive control methods for mining applications
Loading...
URL
Journal Title
Journal ISSN
Volume Title
Sähkötekniikan korkeakoulu |
Master's thesis
Electronic archive copy is available via Aalto Thesis Database.
Authors
Date
Department
Major/Subject
Mcode
ELEC3025
Language
en
Pages
56
Series
Abstract
Model predictive control (MPC) is an advanced control technique for constrained multiple input and multiple output (MIMO) systems. The present work focuses on implementing MPC within the mining industry, particularly in froth flotation processes. Froth flotation is a process used to separate valuable minerals from ore. The process features coupled interactions between its inputs and outputs. Traditional control methods, such as proportional-integral-derivative (PID) control, cannot handle the complexity of applications such as managing constraints. MPC predicts the outputs of the system using a model of its dynamics. It employs optimization methods to calculate the optimal sequence of controls for the plant. MPC follows the principle of receding horizon control and solves an optimization problem at each time step. The choice of optimization method can significantly impact the computational effort required when implementing MPC.
The objective of the thesis is pursued through three research aims. First, this study defines MPC, its principles, and its components. Second, this research enumerates the steps to implement optimization algorithms for quadratic programming and develops a model that describes the dynamics of the MIMO system. Third, this study discusses the results of implementing the active set and interior point methods for quadratic programming.
The results of this study show that the algorithm implemented using the active set method showcased better performance compared to the interior point method. Additionally, the flotation process model is influential when implementing MPC; a precise description of the system dynamics plays a significant role in the efficiency and effectiveness of MPC. The tuning of MPC is straightforward. When combined with Digital Twin technology, it can predict future events based on actual data, offering valuable insights into system performance at specific moments. This enables users to evaluate potential outcomes and devise strategies for resolution.