Search

This thesis investigates the design and development of a Hybrid Electric Propulsion System (HEPS) for aircraft. The main contributions of the study are the multi-objective system sizing and the two energy optimization algorithms. First, the system sizing method is employed to design the hybrid electric propulsion system for a prototype aircraft. The sized hybrid propulsion system can ensure that no significant performance is sacrificed and the fuel economy is improved. The novel approach in this work is a new non-dominated sorting algorithm for the Non-dominated Sorting Genetic Algorithm (NSGA). The new algorithm can improve the time complexity of non-dominated sorting process. The optimized hybrid aircraft can save up to 17% fuel, achieve higher cruising speed and rate of climb. It is concluded that the optimal results are more sensitive to the variation of battery energy density than other parameters. Next, the main components of the HEPS are modelled for example. The engine model provides an insight into the inherent relationship between the throttle command and the output torque. Regarding the d-q model of motor/generator, the estimation of torque loss at steady state is achieved using the efficiency map from experiments. The application of Shepherd model leads to the straightforward parameter identification. In this research, both non-causal and causal energy management strategies for HEPS are investigated. The main novelty when studying convex optimization is the proposal of a new lossless convexification, which simplifies the formation of the convexified problem, and the proof of equality between the original problem and convexified problem. The introduced variable-battery internal energy, is proposed to convexify the battery model. The first test case verifies that the convex relaxation does not sacrifice the optimality of the solution nor does the variable change lose the original bounds. Also, the optimal control from convex optimization is demonstrated to be robust to a disturbance in power demand. Comparison with the benchmark optimization-dynamic programming, shows that convex optimization achieves a minimal objective value with much less optimization time. Most significant is that the convexification reduces the optimization computation time to a level compatible with implementation in practical application. In causal control, the main focus is to extend the original Equivalent Consumption Minimization Strategy (ECMS) with the fuzzy control. The proposed algorithm can maintain the battery State of Charge (SoC) in a desirable range, without the requirement of off-line estimation of equivalence factor. By comparing with non-causal control-dynamic programming, the test cases validates that the fuzzy based ECMS succeeds in converting the non-causal optimization, with little sacrifice of the optimality of the solution. In other words, the prior-knowledge of flight mission is not a pre- requisite, and the fuzzy based ECMS can achieve the sub-optimal control for on-line implementation. The fuzzy based ECMS is also validated to outperform the adaptive ECMS, since it can reduce the computation time of optimization and save more fuel usage. The theoretical relationship between the equivalence factor of ECMS and the co-state variable of Hamiltonian function is also demonstrated in this thesis. The convex optimization and fuzzy based ECMS are combined to complete a flight mission with several sub-tasks. Each task has different power and SoC requirements. The test case demonstrates that only the combination of non-causal and causal optimization can satisfy the various constraints and requests of the test scenario. Compared with the engine-only powered aircraft, the hybrid powered aircraft saves 18.7% on fuel consumption. Furthermore, the hybrid propulsion system has better efficiency since it integrates the high efficient electric powertrain.

This work presents a study of the energy-efficient operation of all-electric vehicles leveraging route information, such as road grade, to adjust the velocity trajectory Minimization of energy consumption is one of the main targets of research for both passenger vehicles in terms of economic benefit, and army vehicles in terms of mission success and decision making. The optimization of a speed profile is one of the tools used to achieve energy minimization and it can also help in the useful utilization of autonomy in vehicles. The optimization of speed profile is typically addressed as an Optimal Control Problem (OCP). The obstacle that disrupts the implementation of optimization is the heavy computational load that results from the number of state variables, control inputs, and discretization, i.e., the curse of dimensionality. In this work, Pontryagin's Maximum Principle (PMP) is applied to derive necessary conditions and to determine the possible discrete operating modes. The analysis shows that only five modes are required to achieve minimum energy consumption; full propulsion, cruising, coasting, full regeneration, and full braking. Then, the problem is reformulated and solved in the distance domain using Dynamic Programming to find the optimal speed profiles. Various simulation results are shown for a lightweight autonomous military vehicle. Army Programs use various drive cycles including time, speed, and grade, for testing and validating new vehicle systems and models. Among those cycles, two different drive conditions are studied: relatively flat, Convoy, and hilly terrain, Churchville B. For the Convoy cycle, the optimal speed cycle uses 21% less energy for the same trip duration or reduces the time by 14% with the same energy consumption while for the Churchville B cycle, it uses 24% less energy or provides 24% reduction in time. Furthermore, the sensitivity of energy consumption to regenerative-braking power limits and trip time is investigated. These studies provide important information that can be used in designing component size and scheduling operation to achieve the desired vehicle range. Lastly, the work provides parametric studies about the influence of the efficiency of an electric motor on performance including energy consumption and control modes.