Uma contribuição para a otimização de portfólios de séries heteroscedásticas usando projeto de experimento de misturas: uma abordagem do desirability aplicada a modelos

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Aplicação do método desirability para otimização de processo com múltiplas respostas

Currently, the competition between organizations in the pursuit of consumer preference has become increasingly fierce. In addition, consumers have become increasingly demanding due to high speed with which innovations occur, leaving the companies meet and sometimes surpass those expectations In this context, there is the necessity to use methods as mathematical models capable of dealing with the optimization of multiple responses simultaneously. In this context, this study presents an application of techniques of Design of Experiment in a machining process of a NIMONIC 80 alloy, a “superalloy” that has thermal and mechanical properties that make its machining difficult and in order to do this, the Desirability Function was used. As they are determining conditions in the machining capability of the alloy, the roughness and the cutting length were considered as variable settings, and the factors that can influence them are cutting speed, feed rate, cutting depth, inserts type and lubrication. The analysis of the result pointed out how was the influence of all factors on each response and also showed the efficiency and reliability of the method

Utilização da função Desirability em um processo de fundição de aço ABNT 1030

This dissertation addresses the main theme the use of desirability tool in optimizing a process with multiple response variables. The current scenario of strong bids to conquer the consumer market makes it necessary to develop improvements for better process performance as a whole, is to cut costs, increase efficiency or effectiveness. Thus, the use of methods to assist in this process is becoming increasingly feasible. This study used the MINITAB program and the data of the doctoral thesis of Dr. Luiz Henrique Dias Alves, in order to compare the results obtained in both studies. As a result, after applying the desirability method, simulated to optimize two responses variables regarding the formation of voids related to solidification in ABNT 1030 steel casting process. Thus, it was possible to evaluate the behavior of the variables with the variation of parameters of the desirability function

Multivariate Development and Validation of a Stability-Indicating HPLC Method for the Determination of Glimepiride in Tablets

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

A utilização de modelos matemáticos para a otimização de processos

This dissertation has as main theme the discuss about how the use of mathematical models for process optimization. The current scenario of strong competition to conquer the consumer market necessitates the development of improvements to better performance of the process as a whole, is to reduce costs, increase efficiency or effectiveness. Thus, the use of methodologies to assist in this process is becoming increasingly viable. Methodologies developed in the past are being studied and improved. An example is the Desirability, the object of the present study, which was developed in the 80's and has been improved over time. To understand and study this methodology was applied to the desirability function in three instances, where it was used Design of Experiments (DOE), taken from scientific papers, using the Solver tool (Excel ®) and desirability (Minitab ®). Thus, in addition to studying the methodology, it was possible to compare the performance of tools used for optimization in different situations. From the results of this study, it was possible to validate the superiority of one of the models studied compared fairly