It is claimed in previous research that inertia weight is used during the implementation PSO and many research work is conducted where the value is chosen as: w(initial) = 0.9 and decrease to w(final) =0.2.
Can anybody suggest the formula that how we are decreasing the value?
Hi,
Look at this
http://home.agh.edu.pl/~slukasik/pub/031_Kowalski_Lukasik_IS2014_2.pdf
equations no. 11, 16 and 18.
All the best,
Piotr
To clearly understand your problem, You can read Fernandez martinez work's about Generalized Particle swarm optimization and the other PSO variants
Hi,
You can find in this article:
http://www.softcomputing.net/nabic11_7.pdf
Best Regards
Mehdi
It will depend on Local search or global search and your problem space. Good Luck!
Hi
The appropriate values for initial and final inertia weights mainly depend on the problem to be solved. In a trial and error process, try different combinations of initial and final weights and see which set of values lead to the best results in your own problem.
Kind Regards
Ahmad
When I wrote a PSO, I implemented decreasing inertia in two ways:
Decreasing inertia could happen in any manner you prefer; these implementations were very easy to code and benefited the PSO algorithm by initially searching the space very quickly, testing many values, then calming down and converging to the best found places.
I hope it will be very helpful to you
https://www.researchgate.net/publication/296636431_Codes_in_MATLAB_for_Particle_Swarm_Optimization?ev=prf_pub
Code Codes in MATLAB for Particle Swarm Optimization
https://www.researchgate.net/publication/297245624_Particle_Swarm_Optimization_Algorithm_and_its_Codes_in_MATLAB
Research Particle Swarm Optimization: Algorithm and its Codes in MATLAB
https://www.researchgate.net/publication/280244004_A_comparative_study_of_metaheuristic_optimization_approaches_for_directional_overcurrent_relays_coordination
Article A comparative study of metaheuristic optimization approaches...
Hari Mohan Pandey I encourage you to look at Section 4.1 of the full text: Mach. Learn. Knowl. Extr. 2018, 1(1), 157-191; https://doi.org/10.3390/make1010010. Section 4.1.1-4.1.4 talk about Random Selection (RS), Linear Time Varying (LTV), Nonlinear Time Varying (NLTV) and Fuzzy Adaptive (FA) respectively. Section 4.2 talks about the constriction factor. The article is open-access and available at: Article Particle Swarm Optimization: A survey of historical and rece...
Hope this helps.
The best way is to conduct an experiment consideing the problema at hand.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Artificial Intelligence