SUBJECT: M.S. Thesis Presentation
BY: Alexander Roeder
TIME: Monday, September 24, 2018, 3:00 p.m.
PLACE: Universität Stuttgart, ISYSISYS, Seminar
TITLE: Transient Simulation of Ammonia-Water Mixture Desorption for Absorption Heat Pumps & Optimal Control of an Electric Vehicle Cabin Air Conditioning System
COMMITTEE: Dr. Srinivas Garimella, Chair (ME)
Dr. G. Paul Neitzel (ME)
Prof. Dr. Oliver Sawodny (Universität Stuttgart, ISYS)
Dr. Ulf Säger (Daimler AG)


Part I:
This paper presents a mathematical framework to simulate the transient response of the desorber for small-capacity ammonia-water absorption heat pumps. The model is based on dynamic analysis of conservation equations and accounts for the thermal capacitances in the heat exchanger wall material and fluid volumes. The numerical solver and its implementation are presented. Improved computational performance is achieved by using advanced solvers for stiff differential-algebraic equations. The model is utilized to predict desorber performance at steady-state conditions and to simulate the transient response of the component to ramping or perturbation of input parameters from steady state. It can also be used to develop reduced-order models suitable for the design of control strategies to optimize system performance.

Part II:
This work presents a control scheme for the air conditioning system of an electric vehicle based on non-linear, constrained optimal control theory with the simultaneous goals of reducing the energy consumption of the AC system and meeting the passengers’ required cabin conditions. Cabin and air system models are derived to define the dynamics of the thermal system. Different formulations of the cabin outlet air temperature available in the literature are explore. From these models, an optimal control problem (OCP) is formulated, with system constraints explicitly considered. The OCP is transformed into a non-linear program using the multiple shooting and fourth-order Runge-Kutta methods and solved using the open-source IPOPT solver. The required derivatives are provided by the CasADi toolkit. The effectiveness of the proposed optimal control scheme in reducing the energy consumption of the electric vehicle AC system is compared to a simple feedback controller using scenarios with constant and time-varying disturbances.