A Dimensionless Parameter Approach based on Singular Value Decomposition and Evolutionary Algorithm for Prediction of Carbamazepine Particles Size

Abstract

The particle size control of drug is one of the most important factors affecting the efficiency of the nano-drug production in confined liquid impinging jets. In the present research, for this investigation the confined liquid impinging jet was used to produce nanoparticles of Carbamazepine. The effects of several parameters such as concentration, solution and anti-solvent flow rate and solvent type were investigated. So far no analytical and acceptable model has been provided to predict the Carbamazepine particle size in confined liquid impinging jets. In this study the variables affecting the size of the particle became dimensionless using the dimensional analysis then by solving the equation with singular value decomposition method, a simple dimensionless relation was obtained for this process. Moreover, using the genetic algorithm the coefficients of dimensionless parameters were optimally extracted to minimize the error between the model and the laboratory outputs. The determination coefficient of the equation obtained by singular value decomposition method and the improved equation using genetic algorithm were obtained as 0.5291 and 0.5697, respectively. For such a complex experimental system, the accuracy of the obtained equations in spite of their simplicity is acceptable. The obtained results were compared with the results of the neural network model. The results showed that despite the higher precision of the obtained relations by the neural network, the relations obtained by singular value decomposition can be used as a simple method using the dimensionless parameters with acceptable acuracy to predict the particle size in confined liquid impinging jets.