A Non-contact Syrup Brix Measurement System Based on Transmittance Intensity Transducer

In the sugar production process, detecting the brix of syrup is crucial for the generation of sugar. Currently, sugar mills all use contact-type detection for measuring brix. Due to the viscosity of the syrup, scaling can occur on the sensor probes, affecting the accuracy of the equipment measurements. This paper proposes a non-contact syrup brix measurement system for sugar syrup, which addresses the clogging issues commonly encountered in traditional measurement methods by utilizing penetration optical transmittance signal acquisition. A transmittance intensity transducer, developed based on the Lambert-Beer Law, was employed in syrup brix measurement experiments. To enhance regression analysis, the bayesian hunter prey optimization (BHPO) algorithm was integrated with the incremental extreme learning machine (IELM) algorithm. The experimental results demonstrated a mean square error of 0.6. Furthermore, comparative analyses of measurement accuracy among the ELM, support vector machines (SVM), and partial least squares (PLS) methods revealed that the ELM exhibited superior performance in accurately measuring the brix of sugar syrup, thereby mitigating the limitations of traditional methods.