Title :
Chemical reaction rate parameter estimation by MAP particle filter algorithm
Author :
Kashiwaya, Shigeru
Author_Institution :
PSE Japan, Yokohama
Abstract :
A fast search algorithm for estimation of chemical reaction kinetics parameters is investigated. A novel parameter estimation technique using maximum a posteriori (MAP) particle filter is presented. Low discrepancy sequences (LDS) are applied to generate a population of particles that are regarded as a set of samples drawn from a uniformly distributed probability density in a high dimensional parameter space. The notion of simulated annealing (SA) is adopted to "rejuvenate" a system of particles around an MAP estimate that is obtained through evolution of such filter particles.
Keywords :
chemical reactions; chemistry; parameter estimation; particle filtering (numerical methods); probability; search problems; simulated annealing; MAP particle filter algorithm; chemical reaction kinetics parameters; chemical reaction rate parameter estimation; fast search algorithm; high dimensional parameter space; low discrepancy sequences; maximum a posteriori particle filter; simulated annealing; uniformly distributed probability density; Chemicals; Parameter estimation; Particle filters;
Conference_Titel :
Evolutionary Computation, 2007. CEC 2007. IEEE Congress on
Conference_Location :
Singapore
Print_ISBN :
978-1-4244-1339-3
Electronic_ISBN :
978-1-4244-1340-9
DOI :
10.1109/CEC.2007.4425059