Optimal Selection of Raw Materials for Pharmaceutical Drug Product Design and Manufacture using Mixed Integer Nonlinear Programming and Multivariate Latent Variable Regression Models

The increasing growth of automotive industry in Indonesia has not been matched by the number of local suppliers and makes the automotive industry too dependent on imported raw materials. Along with the needs of import activities, it is also required a greater logistics activities. However, with high logistics costs, the manufacturer must increase efficiency to be able to compete in the global market. This can be accomplished by planning inbound logistics activities that control the movement of materials from suppliers to the manufacture. In this research, an optimization methodology, based on Mixed Integer Nonlinear Programming (MINLP) approach is developed and solved with branch and bound algorithm. The result of this research, which obtained the total cost of optimal inbound logistics include material cost, transportation cost, and administration cost. This model can also be used as a tool for the company in making decisions about the type and the number of container also with the total of the optimal material load in each container, therefore the optimal container space utilization value can be obtained.

Download Full-text

Multiscale Latent Variable Regression

International Journal of Chemical Engineering ◽

10.1155/2010/935315 ◽

2010 ◽

Vol 2010 ◽

pp. 1-8 ◽

Cited By ~ 2

Author(s):

Mohamed N. Nounou ◽

Hazem N. Nounou

Keyword(s):

Latent Variable ◽

Prediction Accuracy ◽

Regression Models ◽

Multiple Scales ◽

Principal Component Regression ◽

Principal Component ◽

Measurement Noise ◽

Process Data ◽

Latent Variable Regression ◽

Modeling Algorithm

Multiscale wavelet-based representation of data has been shown to be a powerful tool in feature extraction from practical process data. In this paper, this characteristic of multiscale representation is utilized to improve the prediction accuracy of some of the latent variable regression models, such as Principal Component Regression (PCR) and Partial Least Squares (PLS), by developing a multiscale latent variable regression (MSLVR) modeling algorithm. The idea is to decompose the input-output data at multiple scales using wavelet and scaling functions, construct multiple latent variable regression models at multiple scales using the scaled signal approximations of the data and then using cross-validation, and select among all MSLVR models the model which best describes the process. The main advantage of the MSLVR modeling algorithm is that it inherently accounts for the presence of measurement noise in the data by the application of the low-pass filters used in multiscale decomposition, which in turn improves the model robustness to measurement noise and enhances its prediction accuracy. The advantages of the developed MSLVR modeling algorithm are demonstrated using a simulated inferential model which predicts the distillate composition from measurements of some of the trays' temperatures.

Download Full-text

A COMPARISON BETWEEN MILP AND MINLP APPROACHES TO OPTIMAL SOLUTION OF NONLINEAR DISCRETE TRANSPORTATION PROBLEM

Transport ◽

10.3846/16484142.2014.933361 ◽

2014 ◽

Vol 30 (2) ◽

pp. 135-144 ◽

Cited By ~ 8

Author(s):

Uroš Klanšek

Keyword(s):

Transportation Problem ◽

Optimal Solution ◽

Optimization Method ◽

Mixed Integer ◽

Test Problems ◽

Model Formulation ◽

Reference Test ◽

Adequate Model ◽

Integer Nonlinear Programming ◽

Selection Of

Finding an exact optimal solution of the Nonlinear Discrete Transportation Problem (NDTP) represents a challenging task in transportation science. Development of an adequate model formulation and selection of an appropriate optimization method are thus significant for attaining valuable solution of the NDTP. When nonlinearities appear within the criterion of optimization, the NDTP can be formulated directly as a Mixed-Integer Nonlinear Programming (MINLP) task or it can be linearized and converted into a Mixed-Integer Linear Programming (MILP) problem. This paper presents a comparison between MILP and MINLP approaches to exact optimal solution of the NDTP. The comparison is based on obtained results of experiments executed on a set of reference test problems. The paper discusses advantages and limitations of both optimization approaches.

Download Full-text