You are on page 1of 2

Solution

Vehicle Routing problem, is a vital problem, there are various solution for the problem that is related to VDN
problem . There are various data structure algorithms that provide solution for the VDN. According to our study
and analysis we are to provided solution for the problem. According to given business scenario we reach out the
best algorithm tabu search

In this algorithm we use some variable , Here we are describe

Num Total number of consumers.


Vehicle_Time All vehicle total travel time
K The number of vehicles used (the number of routes).
Q_Load Loaded vehicle
T-MaxLength Route maximum length
cust_di Consumer cust_i demand.
cust_dir rth backpack at consumer cust_i demand.
R The max total number in the number of real backpacks of every consumer.
yijk When the vehicle k directly goes to consumer j from consumer cust_i, the value takes 1, else 0.
xirk rth backpack at consumer cust_i is distributed by the vehicle k, the set value takes 1, else 0.
timei The time when the vehicle reaches the consumer cust_i.
timeij The direct travel time between the consumer cust_i and cust_j.
VS An outcome to the problem.
VSinitial Initial outcome.
VScandi Candidate outcome.
VScurr The current outcome.
VSoptimal The optimal outcome.
P The number of Candidate outcomes.
A Candidate outcome set composed of VScandi.
Itra The upper limit of the total number of iteration.
Itrb The upper limit of the total number of iteration of the “optimal outcome” remaining unchanged.

Algorithm

1. Start
2. Enter the data and parameter values.
3. Generate a feasible initial outcome using a random strategy and take it as VScurr and Soptimal.
4. Repeat until the termination condition is not met, do
5. Repeat until the number of candidate outcome is less than p, do
6. Select a neighborhood in five types of neighborhood randomly by using random selection strategy.
7. Transform Scurr by the selected neighborhood to generate a new candidate outcome set A, and build the
optimal non-tabu outcome set B.
8. Goto step 11
9. Combine the previous analysis, if a feasible VScandi is better than the original VSoptimal, then set it
as VScurrand VSoptimal, otherwise the optimal non-tabu VScandi (VScandi ∈ B) is set as the new VSnow; if the
set B = ∅, release the optimal VScandi (VScandi ∈ A), and take it as the new VSnow.
10. Update the tabu list.
11. Stop.

You might also like