Step 1 Initialization: (1) Determine values of the parameters; (2) Set a value to p, e.g.
2
the area of the region
p 1
C π
= ⎡ ⎤ − ⎢⎢ ⎥⎥
.
Step 2 p := p + 1.
Step 3 With given p, solve the model below with the modified VNS algorithm.
Minimize z (8)
subject to (2)-(5), (7)
Step 4 Check if Constraint (6) is met. If it is, go to Step 5; otherwise, go to Step 2.
Step 5 Output the solution and stop.
Fig. 1. Pseudocode of the VNS based algorithm for the EWL problem
algorithm for the EWLP model is more complicated than the one for P-center
problem. The major modifications are in the sub-algorithms of Move and
Update, which are not discussed as the limited space. The pseudocode of the
VNS based algorithm for the EWLP model is shown in Figure 1.
Step 1 of the algorithm involves setting up parameters for VNS and giving
p an initial value to start the program. If only considering the constraint
of the coverage of area, the minimum number of warehouses should be p =
the area of the region
C2π
−1. For example, China has an area of 9,600,000 km2. The
distance limit C equals 840 km. So the area of the region
C2π
= 5 . In Step 3, VNS
will optimize a modified P-center problem when the number of opened facilities
is equal to p. After that, the process of Step 4 checks whether Constraint (6)
is met. If Z is no larger than C, then the corresponding p is the final solution
of the EWLP model and the algorithm ends. Otherwise, the value of p should
be increased as in Step 2 and the processes of Step 3 and 4 will be repeated.
6