A New Computational Technique for the Generation of Optimised Aircraft Trajectories

Abstract

A new computational technique based on Pseudospectral Discretisation (PSD) and adaptive bisection

Quantifying trade-offs for the spatial allocation of onshore wind generation capacity – A case study for Germany

“…First we apply a bi-objective Pareto optimization in which the cumulated costs for one criterion are subject to optimization while we apply constraints to the cumulated costs for the second criterion, which cover the entire range of values from zero to the maximum cumulated cost for the second criterion, described by Soroudi (Soroudi 2017) and Chircop et al (Chircop and Zammit-Mangion 2013) as the equidistant Ɛ-constraint method. For the description and implementation of the bi-objective Pareto optimization into GAMS, we refer to these two references and another application of the Ɛ-constraint method performed in a paper by Neumann and Brown (Neumann and Brown 2021).…”

“…Consequently, two Pareto frontiers result from the optimization: an upper Pareto frontier marking the nondominated solutions for the maximum cumulated cost, and a lower Pareto frontier for the minimum cumulated cost. The two Pareto frontiers create an envelope around all feasible solutions, the feasible region for the allocation (Chircop and Zammit-Mangion 2013). The feasible region entails all the solutions for two criteria covering the expansion of wind power production from zero to full expansion potential in a paired comparison (Figure 1).…”

A two-phase Pareto front method for solving the bi-objective personnel task rescheduling problem

“…The latter method is also called the a posteriori determination of a preferred solution (Marler and Arora, 2004). As computing the entire set of Pareto optimal points is computationally very expensive, this should be avoided and only be done when it is of crucial importance (Chircop and Zammit-Mangion, 2013). Therefore, we approximate the Pareto front by constructing a restricted pallet P F c of the complete set of non-dominated points P F t .…”

Towards Temporally Uncertain Explainable AI Planning

“…To solve this problem, a bi-objective optimisation framework such as the ϵ-constraint method [4] can be used to generate the Pareto optimal front minimising both risk and cost. Within this framework, one of the objectives (i.e.…”