System and method for determining optimal solution in a swarm of solutions using swarm intelligence

An optimization system for determining an optimal solution to an optimization problem in a swarm of candidate solutions, the optimization system comprising:

• a memory storing program instructions;
• a processor configured to execute program instructions stored in the memory;
• and an optimization engine in communication with the processor and configured to:
• a) generate a population of particles in a search-space for an optimization problem based on a first set of rules, wherein each particle is representative of a candidate solution;
• b) evaluate a best particle from the generated population of random particles based on a second set of rules, wherein the best particle is representative of an optimal solution;
• c) convert each particle from the generated population into a datatype compatible for further processing by one or more multivariate measurement techniques and an imitation technique, wherein the datatype is a matrix and generation of the matrix associated with a particle of the generated population comprises:
• evaluating a number of variables in the particle of the generated population;
• evaluating dimensions of the matrix based on the number of variables, wherein number of rows of the matrix is equivalent to the number of variables;
• generating a constant value for first column based on the dimensions of the matrix, wherein the constant values are sequential integers;
• and selecting the variables of the particle as a second column of the matrix;
• d) categorize the generated population of particles into similar and non-similar particle groups based on a similarity with the best particle by applying one or more multivariate measurement techniques on the converted compatible datatype;
• e) update each particle of the non-similar particle group with respect to the best particle by applying the imitation technique;
• f) update the generated population of particles with at least one of:
• one or more updated particles of the non-similar particle group and one or more randomly generated new particles, wherein any one or more of the updated particles is replaced with respective randomly generated new particles if a degree of similarity of the any one or more of the updated particles in relation to the best particle is less than the respective randomly generated new particles;
• g) evaluate a new best particle from updated population of particles based on the second set of rules;
• h) determine if at least one condition from a set of target conditions has been achieved and select the particle with highest solving value in last updated population of particles as final best particle;
• and
• i) repeat steps c,d,e,f,g and h until at least one of the target conditions are achieved.

A computer program product comprising:

• a non-transitory computer-readable medium having computer-readable program code stored thereon, the computer-readable program code comprising instructions that, when executed by a processor, cause the processor to:
• a) generate a population of particles in a search-space for an optimization problem based on a first set of rules, wherein each particle is representative of a candidate solution;
• b) evaluate a best particle from the generated population of random particles based on a second set of rules, wherein the best particle is representative of an optimal solution;
• c) convert each particle from the generated population into a datatype compatible for further processing by one or more multivariate measurement techniques and an imitation technique, wherein the datatype is a matrix and generation of the matrix associated with a particle of the generated population comprises:
• evaluating a number of variables in the particle of the generated population;
• evaluating dimensions of the matrix based on the number of variables, wherein number of rows of the matrix is equivalent to the number of variables;
• generating a constant value for first column based on the dimensions of the matrix, wherein the constant values are sequential integers;
• and selecting the variables of the particle as a second column of the matrix;
• d) categorize the generated population of particles into similar and non-similar particle groups based on a similarity with the best particle by applying one or more multivariate measurement techniques on the converted compatible datatype;
• e) update each particle of the non-similar particle group with respect to the best particle by applying an imitation technique;
• f) update the generated population of particles with at least one of:
• one or more updated particles of the non-similar particle group and one or more randomly generated new particles, wherein any one or more of the updated particles is replaced with respective randomly generated new particles if a degree of similarity of the any one or more of the updated particles in relation to the best particle is less than the respective randomly generated new particles;
• g) evaluate a new best particle from updated population of particles based on the second set of rules;
• h) determine if at least one condition from a set of target conditions has been achieved and select the particle with the highest solving value in last updated population of particles as final best particle;
• and
• i) repeat steps c,d,e,f,g and h until at least one of the target conditions are achieved.