A Mathematical Programming Model for Rural Transportation Network Design Considering Budget Constraint and Environmental Impacts

نوع: Type: thesis

مقطع: Segment: masters

عنوان: Title: A Mathematical Programming Model for Rural Transportation Network Design Considering Budget Constraint and Environmental Impacts

ارائه دهنده: Provider: Mojtaba Behsabz

اساتید راهنما: Supervisors: Dr. Mohsen Babaei

اساتید مشاور: Advisory Professors:

اساتید ممتحن یا داور: Examining professors or referees: Dr. Javad Taherinezhad and Dr. Jalal Akbari

زمان و تاریخ ارائه: Time and date of presentation: 2023/10/14 at 18:00

مکان ارائه: Place of presentation: Faculty of Technology and Engineering, Civil Seminar 2

چکیده: Abstract: Transportation development is considered one of the examples and levers of countries development. In addition to the fact that transportation itself is divided into four subcategories: road, rail, sea and air, it also has a relatively wide interdisciplinary scope and includes various disciplines. The way of prioritization and selection among the defined projects (Either building a new road or increasing the capacity of existing roads) to be added to the existing transportation routes has always been one of the concerns and challenges of the decision makers in the field of transportation engineering (especially transportation planning). The need to comply with macro policies and also the existence of limitations such as budget limitations are added to this issue. In this regard, various approaches have been used so far. One of the most important approaches to the development of rural roads is the approach based on the problem of network design. In most cases, the network design problem is presented in the form of a bi-level mathematical model. At the upper-level of the problem, optimization of projects and at the lower-level of the problem, traffic assignment (choosing the route of road users) affected by the projects selected at the upper-level of the problem is done. Of course, in this approach, according to the need, other approaches can also be used in combination. In the current research, the approach of the network design problem, considering indicators such as total travel time (Separately for passenger and cargo), construction cost (in the form of budget restrictions), fuel consumption (Separately for gasoline and diesel), and also environmental areas (four ecological zones), has been used as an important indicator in dealing with climate change. Network design problems are considered in the category of problems with very high complexity of solution. For example, the increase in the number of projects causes an exponential growth in the number of combinations of projects (scenarios). This is one of the most important challenges that casts a shadow on network design issues. Therefore, the solution algorithm is the most important thing that requires investigation in the design of transportation networks (especially large networks with a relatively large number of projects). In fact, the proposed algorithm should have the ability to produce relatively suitable answers in a reasonable time. Based on this, in the current research, a probabilistic heuristic algorithm based on the marginal (individual) effect of projects has been proposed. The assumption of simplifying the solution process in the algorithm is that the effect of the selected projects in a scenario, if calculated separately and then aggregated, with the time when they are calculated cumulatively, the results will be close to each other. For example, generating solutions based on the marginal minimization of the total network travel time for individual projects can produce solutions close to or in the neighborhood of optimality. In this research, MATLAB software was used for coding (programming) the solution algorithm and PTV-Visum software was used to implement the network and check the efficiency of the generated scenarios. The ability of the component object model has also been used to establish the interaction between the two software. The component object model is one of the capabilities of Visum software, which provides access to data and defined functions in other programs. Since the proposed heuristic algorithm is considered a random algorithm, its performance has been evaluated for five executions and with 20 generated solutions. The results of the algorithm review about the non-dominated scenarios in the Sioux falls network (with 10 projects) as a benchmark network and the country's road network (with 10 projects) as a study sample show the effectiveness of the proposed algorithm. It should be noted that the non-dominated scenario is a scenario in which there is no better solution with a budget level lower than that. The frequency of achieving the optimal solution in the case of Sioux falls network with 29 non-dominated scenarios was 82.1%; Meanwhile, this value has been increased to 85.9% in the case of the country's road network with 27 non-dominated scenarios. After proving the effectiveness of the algorithm in producing optimal solutions, the determination of the best scenario in the country's road network with 100 candidate projects and assuming a budget limit of 50% of the total cost of the projects was also done. The results indicate that the cost required to implement the best scenario in the country's road network is equivalent to 40% of the total cost. This means that investing more than this amount is considered a waste of resources; Because by increasing investment, better results are not obtained. Therefore, assuming a budget limit of 50% of the total cost, the budget level of 40% of the total cost is introduced as the optimal budget level (for building the best scenario)

فایل: ّFile: