Algorithm: Synergistic Fibroblast Optimization (SFO)

Definition

Synergistic Fibroblast Optimization (SFO) is a stochastic population based global search algorithm which has been developed by the inspiration obtained from the intellectual behavior of fibroblast organisms present in the dermal wound healing process. Various characteristics of fibroblast include differentiation, proliferation, inflammation, migration, reorientation, alignment, ECM synthesis, collaborative, goal-oriented, interaction, regeneration, self adaptation and evolution are studied to design and develop a computational model of SFO algorithm. The goal of this SFO algorithm is to find a global optimum (minima or maxima) solution.

The algorithm and the architecture of this newly developed Nature inspired SFO algorithm has been given below for the perusal of researchers worldwide. The MATLAB code for this algorithm is also enclosed. 2 research papers presented and published in International forums have been listed out. Few more papers are yet to be published. I request those who are using the algorithm and code in their research to kindly cite the papers.

Algorithm: Synergistic Fibroblast Optimization (SFO)

Step 1: Initialize the population of fibroblast cells fi, i = 1,2,…,n; with randomly generated position (xi), velocity (vi) and collagen deposition (ecm) in the n-dimensional problem space. The parameter such as cell speed (s) and diffusion coefficient (ρ)values are defined.

Step 2: Repeat

Evaluate the individual fibroblast using fitness function F (fi) in n variables for n times.

Step 3: The reorientation of cell can be performed to find optimal (maxima or minima) solution in the evolutionary space.

Step 4: Update the velocity (vi) and position (xi) of a cell using the following equations:

(3)

where

t = current time;

τ = time lag;

vi= velocity of ith cell

ρ = 0.5

(4)

where

µ/min, L = cell length;

Step 5: Remodeling of collagen deposition (ci) is upgraded in the extracellular matrix (ecm).

Until the predetermined conditions / maximum iterations is met.

Architecture of SFO algorithm

MATLAB Source code

clc;

clear all;

close all;

% Initialization

datestr(clock);

d = rem(now,1);

t = datestr(d);

cells = [1 2 3 4 5 6 7 8 9 10]; % Discrete unit of cells

position = rand(1,11); % Random generation position

arrayposition = position;

%disp('initial 1');

% disp(arrayposition);

velocity = rand(1,11);

arrayvelocity = velocity; % Random generation of velocity

%disp('initial 2');

% disp(arrayvelocity);

ecm = []; % Extracellular matrix – candidate solution are given as input values here

speed = 15;

L = 10;

rho = 0.5; % Parameter initialization

kro = 1000;

r = 0;

% Fitness evaluation

for outer = 1:1000

optima = randsample(ecm,1); % Iteration count

% Initial selection of optimal value (maxima or minima)

for inner = 1:10 % Cells count

n1 = randsample(ecm,1); % Random selection of a collagen

n = randsample(cells,1); % Randomly chosen by cells

a = find(cells == n); % Access the index of an element in ecm

p1 = arrayposition(a);

v1 = arrayvelocity(a);

fx = (n * n); % Sphere benchmark function

% Reorientation

if(optima <= n1)

cbest = n1;

% Velocity and Position updating

datestr(clock);

d = rem(now,1);

torque = datestr(d);

temp = (torque - t);

lag = round(temp);

s = (speed / kro * L);

vel = v1 + (1 - rho) * n + rho * (n * (torque - t)/lag);

v = vel(1);

t = round(v);

%disp('final 1');

%disp(t);

arrayvelocity(a) = t; % Remodelling

pos = p1 + (s * v1);

t1 = round(pos);

%disp('final 2');

arrayposition(a) = t1; % Remodelling

%disp(t1);

r = sprintf('%.2f',cbest);

else

cbest = optima;

r = sprintf('%.2f',cbest);

end % if statement terminates here

end

end

disp('Global optimal solution');

disp(r);

Publications:

1.  Dr. P. Subashini , Ms. T.T.Dhivyaprabha, Dr. M. Krishnaveni , 2016, ‘Synergistic Fibroblast Optimization‘, Proceedings of Joint International Conference on Artificial Intelligence and Evolutionary Computations in Engineering Systems (ICAIECES-2016) and Power, Circuit and Information Technologies (ICPCIT-2016), SRM UNIVERSITY, CHENNAI, 19-21ST MAY 2016.

1. Dr. M. Krishnaveni, Dr. P. Subashini , Ms. T.T.Dhivyaprabha,‘A New Optimization Approach – SFO for Denoising Digital Images’, IEEE International Conference on Computational Systems & Information Technology for sustainable Solution (CSITSS 2016), R V College, Bengaluru, 6th-8th Oct 2016, pg. 34-39, ISBN 978-1-5090-1022-6/16.

Page 4 of 4