Scribd
Upload
95 views10 pages

Large Neighborhood Search (LNS) : Metaheuristics

Large neighborhood search (LNS) is a metaheuristic for solving combinatorial optimization problems. It starts with an initial solution and iteratively destroys and repairs part of the solution to find an improved one. Adaptive LNS selects destroy and repair operators based on their historical performance weights, which are updated each iteration based on whether the new solution improved. Operators that lead to better solutions see their weights and selection probabilities increase over time. The algorithm stops after a maximum number of iterations and outputs the best found solution.

Uploaded by

Achmad P. Rifai
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
95 views10 pages

Large Neighborhood Search (LNS) : Metaheuristics

Large neighborhood search (LNS) is a metaheuristic for solving combinatorial optimization problems. It starts with an initial solution and iteratively destroys and repairs part of the solution to find an improved one. Adaptive LNS selects destroy and repair operators based on their historical performance weights, which are updated each iteration based on whether the new solution improved. Operators that lead to better solutions see their weights and selection probabilities increase over time. The algorithm stops after a maximum number of iterations and outputs the best found solution.

Uploaded by

Achmad P. Rifai
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 10

Large neighborhood

search (LNS)

METAHEURISTICS
Achmad P. Rifai
 = Number of
Pseudocode of LNS
1 Input: iterations
2 Generate an initial solution = degree of
3 Calculate the fitness values of the solution destruction
4 While
5 Perform destroy operation
6 Perform repair operation
7 Perform adjustment to the new solution
8 Calculate the fitness values of the solution
If < then ,
99 End
End while
while
10
10 Output: and

Best solution (solution with Minimum objective value


minimum/maximum objectives value)
Example 3
1. Set the parameters
 

4 1
5
depot
6
2
2. Create an initial solution
All nodes are only visited 7
once
7 4 5 1 6 3 2 8 8 3
Example 3
5. Perform destroy operation
 

If
Random removal
Else
Longest distance removal
6. Perform repair operation and adjustment
If
Random insertion
Else
Greedy algorithm
Example 3
8. Calculate the fitness values of the solution
 

If <

Else

9. Update
10. Record the best-known solution and minimum obtained
objective values, show the results
Adaptive LNS
 •
Previously, we randomly select the operator for destroy and
repair operations
• In ALNS the operators are selected based on their historical
performance

• Each operators are given weights and


• In the beginning the weights are set uniform, and
 = Number of
Pseudocode of ALNS
1 Input: iterations
2 Generate an initial solution = degree of
3 Calculate the fitness values of the solution destruction
4 Set initial weights and = weights of
5 While destroy methods
6 Select destroy and repair method based on its weight = weights of
7 Perform destroy operation repair methods
8 Perform repair operation
9 Perform adjustment to the new solution
10 Calculate the fitness values of the solution
If < then ,
11 Update the weights based on previous performance
12 End while
13 Output: and
Adaptive LNS

• The probability of an operator to be selected is based on its


weight
− −
 
𝑃(𝑐 )=𝑤 /𝑐 ∑ 𝑤 ∀ 𝑐 ∈ 𝐷𝑀
𝑑
𝑑 ∈ 𝐷𝑀

• The selection uses roulette wheel mechanism


Adaptive LNS
• Updating the weights based on its historical performance
 𝑤 − −
𝑑 , 𝑡 +1 =𝛼 (𝑤 𝑑 , 𝑡 )+(1 − 𝛼 ) 𝛽

 𝑤 − −
𝑑 , 𝑡 +1 =𝛼 (𝑤 𝑑 , 𝑡 )+(1 − 𝛼 ) 𝛽

  Z 1 if 𝑓 ( 𝑋𝑛𝑒𝑤 )≤ 𝑓 ( 𝑋 )
𝛽=
{ Z 2 if 𝑓 ( 𝑋𝑛𝑒𝑤 )> 𝑓 ( 𝑋 )
 =decay parameter, controls the sensitivity or weight to change. The higher its
value, the more reluctant the weights to change.
= score to change the weights based on the performance of operator in previous
iteration
References
• Gendreau, M., & Potvin, J. Y. (2010). Handbook of metaheuristics. Springer, Boston,
MA.
• Pisinger, D., & Ropke, S. (2010). Large neighborhood search. In Handbook of
metaheuristics (pp. 399-419). Springer, Boston, MA.
• Ropke, S., & Pisinger, D. (2006). An adaptive large neighborhood search heuristic
for the pickup and delivery problem with time windows. Transportation science,
40(4), 455-472.
• Rifai, A. P., Nguyen, H. T., & Dawal, S. Z. M. (2016). Multi-objective adaptive large
neighborhood search for distributed reentrant permutation flow shop scheduling.
Applied Soft Computing, 40, 42-57.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy