Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.

  

ACTUATOR SATURATION AND ALLOCATION ISSUES IN CONTROL OF FLUID FLOW NETWORKS

Author(s):

Marco Lüthi
Conference/Journal:

Diploma/Master Thesis, WS 03/04
Abstract:

Control of fluid flow networks is a nonlinear multivariable control problem. A change in one branch affects the flows in the whole network. The objective is typically to drive flow rates to desired values. This includes regulation (constant reference values) and tracking (time varying reference values). To reach the desired values one needs to use compressors or pumps to drive the flows and valves to actuate resistances in the network branches. In contrast to most other control problems, some of the inputs (resistances) do not have an upper, but a lower bound. Furthermore, due to the actuator constraints, stable regulation is only possible in a region around the operating point.
A first method to control such a network is by using feedback linearizing controllers (FLC) and control allocation. Because of the branch connections, we have an overactuated system and redundant solutions are possible. Control allocation solves this with allowing the inclusion of some additional objectives such as specifying a preferred resistance or a minimum fan or compressor activity.
To achieve the desired values a second method is to use a purely numerical optimization technic like model predictive control (MPC). By solving quadratic programs, MPC yields a fast convergence to the desired reference values and can also handle input and output constraints.
Additionally, the advantages and disadvantages of the above methods are illustrated and compared for some model networks. Keywords: nonlinear control, fluid flow networks, flow control, feedback linearization, control allocation, MPC.

Supervisors:
Prof. M. Morari, ETH Zurich
Prof. M. Krsti´c, U of California San Diego


Year:

2004
Type of Publication:

(12)Diploma/Master Thesis
Supervisor:

M. Morari

No Files for download available.
% Autogenerated BibTeX entry
@PhdThesis { Xxx:2004:IFA_2746
}
Permanent link