Gideon Villar Leandro

Evolução Diferencial Otimizado por Lógica Nebulosa

This work applies changes in the differential evolution algorithm via fuzzy logic, in order to reduce the computational effort in the identification process of a multivariable system. This work is carried out in two stages. In the first stage, the differential evolution algorithm is simulated several times to obtain the best values: cross rate, mutation rate and population size. This step is also determined the best model of representation of multivariable system that is used in the second stage. In the second stage, the modifications based on fuzzy logic are applied. These modifications adjust the values of cross rate and mutation rate, in real time. The results show that the modifications implemented improve the efficiency of the conventional differential evolution algorithm. Keywords—differencial evoluntion; multivariable system; fuzzy logic; turning parameters online

Analysis of capacitive measurements at low frequencies for moisture content determination in soybeans

This study aims to evaluate the capacitive measurement method at low frequencies for determining the moisture content in soybeans. Furthermore, a technique for measuring the moisture content of soybeans at different temperatures is proposed. A practical cylindrical capacitor was designed for the experiments. Capacitive measurements at frequencies of 1 kHz, 10 kHz and 100 kHz were performed using an LCR meter. Two varieties of soybeans were used in the tests, and the differences were evaluated. The errors of the proposed method were analyzed, and the correlation between moisture content and capacitance was verified. The influence of temperature on the measurements was also quantified. Polynomial models are proposed to compensate for temperature variations. It was found that reasonable precision can be achieved at 100 kHz using the temperature compensation provided by the designed models.