Journal of Geo-information Science >
A Method for Multi-constraint Location Decision of Distribution Center Based on Refined Ant Colony Algorithm and GIS
Received date: 2014-05-05
Request revised date: 2014-10-30
Online published: 2015-02-10
Copyright
Location decision of any logistics distribution center meets multiple constraints, such as the specific spatial environment, the single assignment constraint, the capacity of warehouses and the minimum cost of capital. This paper proposed a model based on refined ant colony algorithm and GIS tools to solve Single Source Capacitated Facility Location Problem (SSCFLP). Firstly, a location selection model was established, which met the target of minimizing the total cost. Secondly, by combining ant colony algorithm and local search, the refined bi-level ant colony optimization to solve the SSCFLP problem was proposed. The solving process was divided into two layers: the layer of choosing facilities and the layer of assigning demands. These two layers were associated with each other. In each iteration, the ants would generate solutions by selecting new sets of facility locations from the candidate sites according to the capacity constraint, and establish the assignment of each customer to a selected facility location using pseudorandom search. The iteration-best solution was optimized and memorized using local search. Then the global optimal solution could be attained through conducting multiple iterations. Finally, a location decision case of the car logistics distribution center in Binhai district was constructed. Site selection space was constructed based on GIS tools, considering demands, candidate sites and shipping cost, and other spatial factors, such as land use, hydrology and terrain. The experimental results revealed that the method was efficient and could find reasonable scheme for determining location and allocation. It had certain academic significance to other similar problems.
Key words: refined bi-level ant colony optimization; SSCFLP; GIS; capacity constraint
ZHAO Renhui , YANG Lina , SHAO Jing . A Method for Multi-constraint Location Decision of Distribution Center Based on Refined Ant Colony Algorithm and GIS[J]. Journal of Geo-information Science, 2015 , 17(2) : 172 -177 . DOI: 10.3724/SP.J.1047.2015.00172
Fig. 1 Acquisition procedure of candidate sites图1 设施候选点的获取流程 |
Fig. 2 Schematic diagram of site connectivity lines图2 选址连通边线示意 |
Tab. 1 Main parameters of bi-level ant colony optimization表1 双层蚁群算法主要参数取值 |
参数名称 | 取值集合 | 最优值 | |
---|---|---|---|
设施选择层 | (启发信息的影响力) | [0.3、0.5、1、3、5、7] | 1 |
(伪随机参数) | [0.05、0.1、0.3、0.5、0.7] | 0.1 | |
(信息素挥发速度) | [0.05、0.1、0.3、0.5、0.7] | 0.1 | |
需求指派层 | (启发信息的影响力) | [0.3、0.5、1、3、5、7] | 3 |
(伪随机参数) | [0.05、0.1、0.3、0.5、0.7] | 0.1 | |
(信息素挥发速度) | [0.05、0.1、0.3、0.5、0.7] | 0.1 |
Tab. 2 Location decision result of bi-level ant colony optimization表2 B-ACO的选址结果 |
设施点 | 指派给该设施点的需求点 | 容量承载(万辆) |
---|---|---|
74号 | 4、6、8、9、11、12、18 | 5.994 |
79号 | 0、3、7、10、14、16、20 | 5.958 |
90号 | 1、2、5、13、15、19、22、23、24、25、27、29、30、32、33、35 | 9.512 |
102号 | 17、21、26、28、31、34、36、37、38 | 5.971 |
注: 90号为已有设施点,最大可承载容量为10万辆 |
Fig. 3 Location and assignment diagrams of car logistics distribution centers in Binhai district图3 滨海新区汽车配送中心的空间分布及需求指派示意图 |
Tab. 3 Comparison of algorithms performance表3 算法性能对比 |
算法 | 最小成本(×104元) | 平均成本(×104元) | 标准差(×104元) | 平均计算耗时(S) |
---|---|---|---|---|
RB-ACO | 2420.99 | 2421.97 | 0.52 | 6.92 |
HACO-LS | 2421.12 | 2423.85 | 1.36 | 1850.40 |
The authors have declared that no competing interests exist.
[1] |
|
[2] |
|
[3] |
|
[4] |
|
[5] |
|
[6] |
|
[7] |
|
[8] |
|
[9] |
|
[10] |
|
[11] |
|
[12] |
|
[13] |
|
[14] |
|
[15] |
|
[16] |
|
[17] |
|
[18] |
|
[19] |
|
[20] |
|
[21] |
|
/
〈 | 〉 |