In-situ Food Contamination Detector Design And Its Quantitative Analysis Arithmetic

Food contamination detection is a growing concern in the food safety today. To effectively control degree of food contamination during food production, this paper introduces a design for a food contamination detector, which allows the in-situ examination requirement for rapid detection and easy debugging. Based on spectrophotometry, we have finished the following design.(1) Design an optical path to obtain monochromatic light output. Silicon photocell was used as detector to achieve photoelectric conversion. tungsten halogen lamp was chosen for system light sources by the range the characteristic absorption wavelength of common food contamination. The type of grating was chosen by the efficiency curves in the range of continuous spectrum radiated by light sources; the included angle between incident direction and diffraction direction was designed based on the blaze of grating.(2) Design light splitting unit to achieve the required spectral scanning. Based on sinusoidal mechanism, this design improves the light splitting unit to expand the scanning range and enhance the accuracy of output wavelength. In this design, the motor rotation angle φ is designed to have a linear relationship with the output wavelength angle λ, which simplifies the conversion of output spectral curves into wavelength vs. light intensity curves; the size of mechanical components is reduced to realize miniaturization; the limit of the output wavelength of the monochromatic light is expanded, and thus allows wide spectral scanning; the linear coefficient of the motor rotation angle φ and the output wavelength λ is reduced, which improves the accuracy of the wavelength of the output monochromatic light. In this study, we also derive the relationship between the device’s major sources of errors and cumulative error of the output wavelengths, and suggest a simple correction for these errors.(3) Adjustment light source, optical devices were adjusted to reduce stray light;Adjust the voltage of the light source so that the desirable spectra, exhibiting gentle wave peak variance and below the saturation limit of silicon photocell, will form. The system outputted the light intensities on the step number. Therefore, the step number vs. light intensity spectral lines should be converted into wavelength vs. light intensity curves,a didymium filter was employed for the task.(4) Detect single contamination, we detected single pigment to evaluate the capacity of system, because quantitative analysis was based on lambert-beer law, and the spectrum radiated by light source was in the range of visible spectrum. 4 curves of absorbance vs. gradient concentrations with different pigments were fitted by the least square method, and their linear correlation coefficients surpassed 98%. We detected standard Amaranth solution with a predetermined concentration, the deviations between the measured values and the preset, values were only 1.84%. The quantitative analysis of compound additive and microbial contamination were achieved by integrating chromogenic agents.(5) Design algorithm for multiple pigments, relation matrix was calculated by matrix least square fitting for the simple case(the absorbance curves of pigments in the mixture are different, or their characteristic absorption wavelength are separated); For the complex case of spectral curve overlap, Algorithm design removes the duplicate information to achieve quantitative analysis,the experiment results showed that predicted values calculated by this algorithm coincided with the presets.