Model Based Control of Single-Phase Marine Cooling Systems

Publikation: ForskningPh.d.-afhandling

Abstrakt

This thesis is concerned with the problem of designing model-based control for a class of single-phase marine cooling systems. While this type of cooling system has been in existence for several decades, it is only recently that energy efficiency has become a focus point in the design and operation these systems. Traditionally, control for this type of cooling system has been limited to open-loop control of pumps combined with a couple of local PID controllers for bypass valves to keep critical temperatures within design limits. This research considers improvements in a retrofit framework to the control strategy and design for this particular class of marine cooling systems. The project has been carried out under the Danish Industrial PhD programme and has been financed by Maersk Maritime Technology together with the Danish Ministry of Science, Technology and Innovation. The main contributions in this thesis are on the subjects of modeling and control of a single-phase marine cooling system and experimental model validation.A great deal of attention is on the derivation and experimental validation of a model covering the relationships between pressure, flow and temperature in the cooling system.The proposed model is derived with the intention of being scalable and of low complexity, while capturing important dynamics to make it suitable for model-based control design and simulation. Based on experimental data compiled from a retrofitted test installation on board the container vessel ”Maersk Senang”, it is shown that the part of the proposed model relating to the thermodynamics is dynamically accurate and with relatively small steady state deviations. The same is shown for a linear version of the part of the model governing the hydraulics of the cooling system.
On the subject of control, the main focus in this work is on the development of
a nonlinear robust control design. The design is based on principles from feedback. linearization to compensate for nonlinearities as well as transport delays by including a delay estimate in the feedback law. To deal with the uncertainties that emerged from the feedback linearization, an H∞-control design is applied to the resulting linear system. Disturbance rejection capabilities and robustness of performance for this control design methodology is compared to a baseline design derived from classical control theory. This shows promising results for the nonlinear robust design as disturbance rejection
overall is improved, while robustness of performance is similar to the baseline design, even when considering significant model uncertainties. This improvement to the control is expected to result in a significant reduction of the annual energy consumption of the single-phase marine cooling system. For the specific configuration and control used for the test installation on board ”Maersk Senang”, it is estimated that energy savings above 53% are achievable.
Luk

Detaljer

This thesis is concerned with the problem of designing model-based control for a class of single-phase marine cooling systems. While this type of cooling system has been in existence for several decades, it is only recently that energy efficiency has become a focus point in the design and operation these systems. Traditionally, control for this type of cooling system has been limited to open-loop control of pumps combined with a couple of local PID controllers for bypass valves to keep critical temperatures within design limits. This research considers improvements in a retrofit framework to the control strategy and design for this particular class of marine cooling systems. The project has been carried out under the Danish Industrial PhD programme and has been financed by Maersk Maritime Technology together with the Danish Ministry of Science, Technology and Innovation. The main contributions in this thesis are on the subjects of modeling and control of a single-phase marine cooling system and experimental model validation.A great deal of attention is on the derivation and experimental validation of a model covering the relationships between pressure, flow and temperature in the cooling system.The proposed model is derived with the intention of being scalable and of low complexity, while capturing important dynamics to make it suitable for model-based control design and simulation. Based on experimental data compiled from a retrofitted test installation on board the container vessel ”Maersk Senang”, it is shown that the part of the proposed model relating to the thermodynamics is dynamically accurate and with relatively small steady state deviations. The same is shown for a linear version of the part of the model governing the hydraulics of the cooling system.
On the subject of control, the main focus in this work is on the development of
a nonlinear robust control design. The design is based on principles from feedback. linearization to compensate for nonlinearities as well as transport delays by including a delay estimate in the feedback law. To deal with the uncertainties that emerged from the feedback linearization, an H∞-control design is applied to the resulting linear system. Disturbance rejection capabilities and robustness of performance for this control design methodology is compared to a baseline design derived from classical control theory. This shows promising results for the nonlinear robust design as disturbance rejection
overall is improved, while robustness of performance is similar to the baseline design, even when considering significant model uncertainties. This improvement to the control is expected to result in a significant reduction of the annual energy consumption of the single-phase marine cooling system. For the specific configuration and control used for the test installation on board ”Maersk Senang”, it is estimated that energy savings above 53% are achievable.
OriginalsprogEngelsk
Antal sider206
ISBN (trykt)978-87-7152-008-8
StatusUdgivet - 2014

Presse/medie-elementer

Download-statistik

Ingen data tilgængelig
ID: 180343454