An optimization approach for green tourist trip design

In this paper, the Multi-Objective Multi-Modal Green Tourist Trip Design Problem (MO-MM-GTTDP) as the multi-modal variant of the orienteering problem is investigated. For this problem, a mixed-integer linear model is formulated, containing three objectives: (1) maximizing the total score of the trip, (2) minimizing the total cost of the trip, and, (3) minimizing the total CO 2 emission produced by transportations in the trip. Various transportation modes are considered for the tourist to choose to move between points of interest (POIs). The tourist choice may be affected by the transportation time and cost. Moreover, choosing the transportation mode will have an impact on the amount of trip pollutants. The cost of visiting POIs, as well as the cost of transportation between POIs, is considered as the total cost of the tour. The proposed problem is solved for two types of tourists: (1) tourists who give the highest priority to the total score, then the total cost, and at last, the total emission; (2) tourists who would like to compromise between objective functions and select a solution which best suits their situation. For the former type, the problem is solved using lexicographic method and one best solution is proposed. For the latter type, a -constraint method is implemented which provides Pareto optimal solutions. In addition, a Multi-Objective Variable Neighborhood Search (MOVNS) algorithm is designed to solve instances of this problem. The performance of this method is evaluated by comparing its results with the -constraint and the lexicographic methods. New instances of the problem are generated based on the OP existed benchmark instances. Moreover, the MOVNS is also compared against two other metaheuristic versions of the solution approach using multiple metrics. The conclusion is the high quality of the proposed MOVNS algorithm solutions in practically acceptable computation time (few seconds). Finally, a small case study based on real data on several POIs in the city of Tehran is generated and used to demonstrate the performance of the proposed model and algorithm in practice. For this case study, by using the multi-attribute decision-making method of TOPSIS, the obtained non-dominated solutions are ranked, and the best ones are presented.