Formulation of an agrochemical product using neural network driven design
Chemical Product Design and Engineering (CPD&E)
Chemical Product Design & Engineering - I (CPD&E - 1)
Keywords: agrochemical,neural,network,formulation,optimisation
Agrochemical formulations are complicated mixtures. Conventional experimental designs suffer from the large number of variables and non-linear behaviour. A coarse grained approach to optimisation uses initial experiments to find the approximate optimised region which is then studied in greater detail with further experiments. This approach can be automated using modelling.
Optimisation of a seven component formulation was studied using a combination of neural network software, genetic algorithms, simulated annealing and Monte Carlo simulation. Successive optima were generated and used to examine experimental designs around smaller and smaller regions of the design space.
Presented Wednesday 19, 12:07 to 12:26, in session Chemical Product Design & Engineering - I (CPD&E - 1).
