The Berth Allocation and Scheduling Problem (BASP) is a critical optimization challenge in maritime logistics, aiming to assign arriving vessels to berths efficiently, while adhering to practical constraints. Exploiting the connection of BASP with the Heterogeneous Vehicle Routing Problem with Time Windows (HVRPTW), we propose a mixed integer linear programming (MILP) formulation for a variant of BASP which is of utmost importance in real-world scenarios: the Dynamic Discrete Berth Allocation and Scheduling Problem with Time Windows (DDBASPTW). Consequently, inspired by the wealth of constructive and improvement heuristics for VRP, we design, implement and experimentally evaluate three constructive heuristics, Nearest Neighbour (NN), Insertion (INS), a quick-and-dirty variant of Insertion (qd-INS), as well as two improvement heuristics, Swap and Reinsert, taking into consideration both the online and the offline scenario with respect to vessel arrivals. Finally, we propose, implement and experimentally evaluate, custom-tailored variants for DDBASPTW of a single-solution metaheuristic, the Adaptive Large Neighborhood Search (ALNS), and of two population-based metaheuristics, the Genetic Algorithm (GA) and the Cuckoo Search Algorithm (CSA), which are aimed to solve the offline version of the problem. An extensive experimental evaluation compares these techniques against a generic state-of-the-art MILP solver. Results demonstrate that certain variants of INS not only are extremely fast and deliver competitive solutions, achieving a practical trade-off between execution times and quality of solutions. The improvement heuristics further refine the initial solutions, especially for weaker constructive approaches, offering a lightweight yet effective enhancement mechanism. The metaheuristics consistently yield high-quality solutions with significantly lower computational times compared to the exact MILP solver, making them well-suited for use in real-time or large-scale operational environments.

A key task in multi-objective optimization is to compute the Pareto frontier (a.k.a. Pareto subset) P of a given d-dimensional objective space F; that is, a maximal subset P ⊆ F such that every element in P is non-dominated (i.e., it is better in at least one criterion, against any other point) within F. This process, called dominance-filtering, often involves handling objective spaces derived from either the union or the Minkowski sum of two given partial objective spaces which are Pareto sets themselves, and constitutes a major bottleneck in several multi-objective optimization techniques. In this work, we introduce three new data structures, ND^{+}-trees, QND^{+}-trees and TND^{+}-trees, which are designed for efficiently indexing non-dominated objective vectors and performing dominance-checks. We also devise three new algorithms that efficiently filter out dominated objective vectors from the union or the Minkowski sum of two Pareto sets. An extensive experimental evaluation on both synthetically generated and real-world data sets reveals that our new algorithms outperform state-of-art techniques for dominance-filtering of unions and Minkowski sums of Pareto sets, and scale well w.r.t. the number of d ≥ 3 criteria and the sets' sizes.