Development of a microcontroller-based microwave instrumentation system for determination of moisture content in oil palm fruits and ginger

In agriculture, moisture content (MC) is among the important factors that are closely correlated to the properties of agriculture products. There is a diverse range of techniques such as Near-Infrared (NIR) spectroscopy, magnetic resonance, X-ray and computed tomography that have been used for the quality assessment in agriculture products. However, the microwave aquametry technique which applies MC measurement and correlates it to the quality of the agriculture products, has been widely used due to the fast, precise, cost and energy saving as well as compliance with safety regulations. Several sensors such as, microstrip and coplanar which are based on the attenuation measurement, have been suggested previously for the MC determination but these techniques required laborious preparation. The Keysight OEC probe has been used to determine the permittivity of agriculture products, however, it requires a network analyzer which is bulky and expensive. As a solution to this limitation, the low cost microcontroller-based microwave instrumentation system for determination of MC percentage in oil palm fruits and gingers is developed in order to determine the quality of samples by relying on MC measurement. This thesis describes in detail the development of a microcontrollerbased microwave instrumentation system for the determination of MC percentage in oil palm fruits and ginger to determine the quality of samples by relying on MC measurement. This instrumentation system or also known as reflectometer operates at 2GHz includes a stripline directional coupler, two diode detectors, a PIC microcontroller, liquid crystal display (LCD), and sensors such as open ended coaxial-stub contact panel and monopole sensors. The stripline directional coupler’s design, analysis and also performance testing are accomplished by using Microwave Office Software. The Flow Code version 5.5 was used to program the PIC16F690 microcontroller for data acquisition as well as to calculate the MC based on measured reflected voltage, processing and LCD display. The permittivity measurement is carried by using the Keysight 85070B dielectric probe kit that utilize with the computer controlled HP 8720B vector network analyzer software. The COMSOL Multiphysics® software is used to visualization of the electric field distribution of both the open ended coaxial-stub contact panel and monopole sensors based on the permittivity value measured. The calibration equations relating the measured reflected voltage by using the reflectometer to the actual MC, and permittivity (dielectric constant (ԑ′) and loss factor (ԑ")) has been established. The predicted percentage of MC in samples is calculated based on the measured reflected voltages can by interchanging the y and x axes. On the other hand, for the predicted permittivity measurement, the calibration equation between the reflected voltage and MC is substituted into the relationship between permittivity and MC. The accuracies of the calibration equations were determined by comparing the predicted MC with the actual MC using microwave oven drying method on another batch of the samples while for the permittivity measurement, the accuracy is determined by comparing the predicted and actual permittivity values that obtained from the another batch of permittivity measurement by using Keysight OEC probe. All of the calibration equations shows a good agreement for input reflected voltage which utilizing the reflected voltage values to determine the MC, ԑ′ and ԑ" by utilizing the mean relative error formula. For oil palm fruits, the accuracies for MC, ԑ′ and ԑ" were within 3.8%, 4.1%, and 4.5%, respectively. While for ginger, the accuracies for MC, ԑ′ and ԑ" were within 2.9%, 2.7%, and 3.6%, respectively.

Preview Text FS 2018 31 - IR.pdf Download (665kB) | Preview

Actions (login required)

View Item