Reward-Density Heuristic for Dynamic Multi-Vehicle Routing: Performance and Computational Efficiency

Obstacle-Aware Multi-Target Routing for Campus Logistics Using an Improved Mayfly Optimization Algorithm

Autonomous mobile robots are increasingly considered for campus delivery and service logistics, where route efficiency can reduce unnecessary travel under spatial constraints. This study develops an obstacle-aware routing framework that combines a 1 m occupancy grid, A* shortest-path computation, and an Improved Mayfly Optimization Algorithm (IMOA). The A* stage constructs a pairwise distance matrix using orthogonal costs of 1, diagonal costs of 2, an octile heuristic, and a no-corner-cutting rule; IMOA then optimizes the closed visiting order through random-key decoding, elite 2-opt, and stagnation handling. Validation comprises ten independent benchmark instances, the fixed G40 scenario, and a campus-derived G-real application. Under a common budget of 50,000 full-tour evaluations and 30 independent runs, a Friedman test detected overall differences across the ten instances (χ2(7) = 66.2488, p = 8.434 × 10−12). After Holm correction, IMOA significantly outperformed GA, PSO, GWO, ACO, and MOA, showed no significant difference from MS2OPT, and had a worse average rank than the deterministic LKH reference, which achieved the best overall rank. On G-real, IMOA obtained a median distance of 8178.37 m, compared with 8223.99 m for MS2OPT; this difference was not significant, while LKH achieved the lowest deterministic cost of 8076.46 m. A three-instance exploratory ablation ranked IMOA first and consistently identified elite 2-opt as the principal observed improvement source; component-level inference remains exploratory because only three instances were available. These findings establish routing-efficiency gains under the evaluated protocol. Such gains may support more resource-efficient campus logistics, but energy consumption and carbon emissions were not evaluated.

A novel ant colony system for vehicle routing problem with time windows: leveraging multiple heuristic strategies for improved solutions

From Feasibility to Multi-Criteria Optimization in Service Team Transport Scheduling: A Declarative and Metaheuristic Perspective

This paper investigates the scalability of a multi-criteria optimization for the Service Team Transport Scheduling (STTS) problem, minimizing total travel time, maximum vehicle worktime, and total vehicle engagement time. The mathematical model of the problem is provided. A simplified problem representation and two graph-based representations along with problem properties are formulated. Using these properties, we propose two complementary frameworks: a declarative Constraint Optimization Problem (COP) solved via Gurobi, and a metaheuristic Tabu Search (TS) incorporating a Constraint Satisfaction Problem (CSP) and an auxiliary greedy method for obtaining initial feasible solutions. Additionally, the population-based NSGA-II algorithm is implemented as an evolutionary baseline. The solving methods are evaluated under a strict 300-second operational limit using 280 realistic instances. The results reveal clear tractability thresholds. The COP solver obtains high-quality Pareto fronts but experiences a combinatorial collapse beyond 8 customer locations, hitting a 0% success rate for instances with size 15 or more. Conversely, the proposed hybrid CSP-TS framework demonstrate high robustness, maintaining a 90–100% operational success rate up to 50 locations. Specifically, the expansive large-neighborhood TS variant excels at medium scales (14–40 locations), while the compact small-neighborhood TS variant achieves the highest Hypervolume Indicator (HVI) at large scales (45 or more locations). Pure population-based approaches like NSGA-II are consistently outperformed by the specialized local search neighborhood structures. These quantitative findings define scale-aware algorithmic boundaries essential for real-time decision support systems.

Optimizing Multi-Depot Vehicle Routing: An ABC-GA Hybrid Algorithm

The Multi Depot Vehicle Routing Problem (MDVRP) is a generalized form of the Vehicle Routing Problem (VRP) and Travelling Salesman Problem (TSP). It is considered one of the NP-hard optimization problems. The MDVRP is a logistics problem that involves finding the most efficient route to transport goods between multiple different pickup and delivery locations. In this study a hybrid metaheuristic algorithm that integrates artificial bee colony and genetic algorithms is developed to solve MDVRP efficiently. The main objective of this study is to find the optimal route from multiple depots to serve a set of customers dispersed in different geographical locations. Initially nearest neighbor algorithm is used to assign customer to their nearest depot and randomly generated initial solution. These solutions are subsequently optimized using the proposed hybrid ABC-GA algorithm. The ABC algorithm serves as the main optimization framework, within which GA operators are embedded in the employed and onlooker bee phases. A probabilistic selection mechanism dynamically applies ABC neighbor search, GA operators, and 2-opt local search to improve solution quality. Finally, the results of the ABC-GA approach are compared against the outcomes of the conventional GA and ABC algorithm to assess performance improvements. Experimental results demonstrate that the hybrid approach consistently produces lower routing costs, achieving superior solution quality at the expense of increased computational time.

Genetic programming-assisted automatic heuristics design for dynamic vehicle routing with drones considering time window

Multi-Constraint Three-Dimensional Bin Packing Optimization for Mixed Vehicle Types: A Heuristic Approach

Aiming at the multi-constraint three-dimensional bin packing problem for mixed vehicle types in urban logistics, where traditional exact algorithms are limited by NP-hard computational complexity and practical engineering constraints, this study proposes a progressive optimization framework for vehicle loading and fleet allocation optimization. First, a heuristic loading algorithm based on the extreme point method and greedy strategy is developed to maximize single-vehicle loading efficiency by balancing space and weight utilization. Second, an NSGA-II based evolutionary framework with sequential encoding is constructed to minimize fleet size while improving loading balance for single-vehicle-type optimization. Third, a three-stage hybrid algorithm integrating greedy packing, enumerative search, and tail vehicle replacement is designed to optimize mixed-vehicle fleet composition and minimize total transportation cost. Experimental results demonstrate that the proposed heuristic achieves high composite loading performance across vehicle types, and the evolutionary framework significantly reduces fleet size compared with theoretical lower bounds. Under mixed-fleet optimization, the model identifies cost-effective vehicle configurations that outperform single-type dispatching strategies. Sensitivity analysis reveals that cargo composition, particularly the number of fragile items, is the most critical factor affecting system performance, while validation on 16 vehicle types confirms the robustness and practical generalizability of the method. This study verifies the effectiveness and stability of heuristic-evolutionary hybrid optimization methods, providing a reliable decision-making reference for logistics enterprises in vehicle selection, cargo allocation, and transportation planning.