- Title
- An optimization approach to the auto-carrier problem
- Author
- Agbegha, Gerald Yinkefe
- Degree
- Doctor of Philosophy, Case Western Reserve University,
Operations Research, 1992.
- Advisor
- Ronald Ballou
- Pages
- 184p.
- Abstract
- A common problem in the auto-transportation industry is the distribution of cars of different characteristics to spatially distributed dealers. The objective is to minimize the total delivery cost and to obtain a more efficient management of the daily operations. This problem falls among the general class of vehicle routing problems (VRPs). What makes this problem quite different from other VRPs is the presence of a loading problem, which involves assigning the cars to the appropriate slots of a special tractor-trailer rig (called an auto carrier). As a particular type of car is assigned to a particular slot, adjustments are invariably made to the boundaries of this slot. These adjustments also affect the size and shape of adjacent slots, thus preventing the subsequent assignment of certain types of cars to some adjacent slots. In this dissertation, the problem is formulated as a highly nonlinear VRP in which two cost components – mileage cost and reloading cost – are considered. The focus of the dissertation is on developing efficient algorithms for the loading problem – a problem which must be solved many times before the overall problem is solved. The problem of determining whether a given car load can be feasibly assigned to a trailer is complex in its own right. Moreover, if a feasible load is destined for several dealers, then one must find a proper order to assignment the cars for easy delivery – that is, minimize the number of reloads. A reload occurs when a car is unloaded and then reloaded before its destination in order to unload other cars. Efficient algorithms have been designed and implemented for this purpose. For the loading problem, the auto carrier can be represented as a directed acyclic network. The level of difficulty of these algorithms depends to a large extent on the type of network involved. Loading problems involving the tree network are generally easier to solve than those involving a general network. A vehicle routing scheme incorporating the loading algorithm is also presented. Within this scheme, an attempt is made to balance the routing and reloading costs by considering alternative routes. A case study of the auto-carrying operations at the Lorain Ford plant in Ohio is presented. Two example auto-carrier problems are solved. These examples show that reasonable size auto-carrier problems can be solved in very little time on commonly found computer operating systems such as the sunOs-Unix systems.
- Keywords
- optimization auto-carrier problem
Document number: case1056564344.
Bookmark this page as
<http://rave.ohiolink.edu/etdc/view?acc_num=case1056564344>.
