Optimizing Logistic Problem in Automotive Corporations by Linear Programming
Optimizing Logistic Problem in Automotive Corporations by Linear Programming
Optimizing Logistic Problem in Automotive Corporations by Linear Programming
I.
INTRODUCTION
We face to multi-objective optimization in real world decision making and in our model we minimize transfer costs according to the priority of sending cars [2,3]. To solve this problem we use Lexicographical method; we divide the whole country in definite number of zones by taking a glance on our present trajectory network of whole country and also the geographical place of the target cities; Our purpose is omitting irrational answers of model, making the process of problem solving faster, and also minimizing the scope of problem. It is necessary to mention that solving the problem by increasing the number of zones makes the process easier but it lasts longer. After all these done, by sending up cars to the zones with the highest priorities which is calculated by summing up the priorities of the cars which are assumed to be send to that zone, we allocate full trucks to that zone to transfer our soled cars. Car priorities are as a direct function of delay time of delivering the cars and also special
49
Taking into account tactical decisions, in addition to the conventional strategic one, we can give more operational and practical insights and provide innovative resolutions to logistics problems After presenting some preliminary statements, we provide a general formulation and propose a solution procedure. Finally, the validity of the proposed method is shown through numerical experiments.
In addition, suppose that we have n trucks with the capacity c. We also define three zero-one functions [1]:
Fig 1. Solution image for the problem considered III. PROBLEM DEFINITION
(2)
The proposed logistic model includes some objectives: using maximum capacity of trucks for transferring sold cars and also minimizing the length of routes which a truck traverse in, especially when a truck runs through several cities. The direct relation between these items and the automotive corporations logistic cost is assumed [6]. First, the country will be divided into definite zones according to the latest map of roads which connects different cities to each other in a country. By increasing the number of zones the answer of the model will be more precise, but it lasts more time and the number of operations will increase rapidly. It is obvious that by supposing the whole country as a united zone, we will face the most general model [11]. The priority of loading cars to trucks depends on time and managerial view points. Cars which are waiting in the queue delivering for more time have more priority; Besides it is possible to send some cars sooner by managerial commands according to social and political conditions. Priority of the cars not loaded to trucks today will increase the next day and the probability of sending them tomorrow will be added. By supposing this matter, the average and the maximum time of delivering cars to customers will increase after a long time [7-9]. IV. PROBLEM SOLVING APPROACH Our solving approach for solving the implied logistic problem is using Lexicographical method. The most preferable consignments must be delivered at first. The goal is to maximize the sum of loaded cars to trucks which are going to be delivered in different cities and according to logistic limitations in an automotive corporation, the number of allocated trucks and the cars supposed to be delivered to each region will be specified in this stage. In next stage minimizing the length of routes by trucks will occur [8]. V. MATHEMATICAL MODEL For the model systematically, some sets are assumed;
In addition assume that function length(r) denotes the length of route r and priority(i) is the priority of car i. Now consider the formula of the priority of regions for allocating trucks. We calculate the average priority of cars which must be delivered in each region:
priority(k) =
sum of priorities of cars delivered in region k (4) sum of number of cars delivered in region k
.
[region( j, k)*t arg et(i, j)* priority(i)] priority(k) = [region( j, k)*t arg et(i, j)]
j i j i
(5)
For initializing the parameters we suppose the number of not allocated trucks to all the regions names rest which is n, the total number of trucks, at first. We choose the region with the highest priority to allocate trucks to it. The number of trucks which should be assigned to region k is as follows:
region k ' s car assign(k ) = min , rest trucks (6) truck capacity
Consider that we use the absolute value of the number of regions car to trucks capacity in order to send full trucks not partially filled trucks to their destination. Usually a truck can carry 6 cars, so its capacity is 6. The most advantage of this formula is the complete usage of our truck capacities. According to our functions,
50
Constraint (i) implies that each of cars can be carried just by one truck; Constraint (ii) is the constraint of trucks capacity, for example 6 cars; Constraint (iii) says that car i will be carried by truck t when truck t goes across the city in which car I must be delivered. VI. CONCLUSION In this article we solved a problem that most of automotive corporations face with it in real world. A comprehensive model based on linear programming is proposed in this research. The model is formulating in mathematical programming whit objective function of minimizing the whole path which sold cars are carried on, subject to various operational constraints. Therefore, we present a Linear Programming Model which can offer an answer by means of most mathematical softwares very fast and in a rational time. For future studies we suggest Dynamic Programming with more objectives. REFERENCES
[1] [2] [3] [4] Balas, E., An Additive Algorithm for Solving Linear Programs with Zero-One Variables, Operations Research 13, 1965. Ballou H. Roland, Business Logistics management, prentice Hall Inc, 1999 L. D. Burns, R. W. Hall, Distribution strategies that minimize transportation and inventory costs, Operation Research. 33, 1985. Cannon, T.L. and K.L. Hoffman, Large-Scale 0-1 Linear Programming on Distributed Workstations, Annals of Operations Research, 22(3), 1990. G. B. Dantzig., R. H. Ramser, The truck dispatching problem, Management Science 6, 1959. E. V. Denardo, U. G. Rothblum and A. J. Swersey. A transportation problem in which costs depend on the order of arrival. Management Science, 34(6), 774783, 1988. B. Golden, and A. Assad, Vehicle routing: methods and studies, North-Holland, Amsterdam, 1998. [8] M. Iori, J.J. Salazar Gonzalez, D. Vigo, An exact approach for vehicle routing problem with two dimensional loading constraints, Transportation Science, 2005. D. Naddef,, G. Rinaldi, , Branch-and-Cut algorithms for the VRP. in the vehicle routing problem, SIAM Monographs on Discrete Mathematics and Applications, Philadelphia, 2002. Orchard-Hays, W., History of Mathematical Programming Systems In Design and Implementation of Optimization Software, Harvey J. Greenberg, ed., Sijthoff & Noordhoff (Alphen aan den Rijn, The Netherlands, 1978. D. Pisinger, S. Ropke, A general heuristics for vehicle routing problems, Technical report, Datalogisk Institute Kobenhavn Universitet (DIKU), Denmark, 2005. Wu Qingyi, Logistics management. Chinese Material Press: Beijing, 2003. [13] J. L. Ringuest and D. B. Rinks. Interactive solutions for the linear multiobjective transportation problem. European Journal of Operations Research, 32,1987.
(8)
[5] [6]
[7]
(9)
[8]
[9]
[10]
[11]
[12] [13]
[14]
P. Toth, D. Vigo, The vehicle routing problem, Monographs on Discrete Mathematics and Applications. SIAM, Philadelphia, 2002.
51