Paper__Feedback_Linearization.pdf (944 kB)
Download fileA Feedback Linearization Approach for Coordinated Traffic Flow Management in Highway Systems
This paper aims to develop a traffic control approach that provides an
optimal real-time solution for large-scale highway networks. The traffic
flow model’s nonlinearity is the main reason for the complex optimization
problems that consequently require high computational effort. Feedback
linearization is a well-known approach in nonlinear systems control. It
can provide an exact linear representation of the original nonlinear
system. Thus, it facilitates further steps in the controller design. This
research combines the feedback linearization method and linear MPC to
simultaneously guarantee optimal performance and real-time feasibility.
While developing feedback linearization for METANET model, we
discovered a pattern that describes the expansion of the control signals
(ramp metering and variable speed limits) and disturbances effects
through the highway networks. Utilizing this pattern, we design a generic
feedback linearization control law for METANET model. The control law
provides the key connection between the linearized and original model.
Finally, we complete the methodology by employing a linear MPC to
regulate the linearized model. The performance of the designed method
is evaluated by conducting comprehensive simulation studies, including a
large-scale network. The simulation results are promising. Eventually,
comparing the developed methodology with an equivalent nonlinear MPC
verifies a substantial improvement in computational costs.
History
Email Address of Submitting Author
kimiac@ethz.chORCID of Submitting Author
0000-0001-7527-9251Submitting Author's Institution
ETH ZurichSubmitting Author's Country
- Switzerland