Design And Model Implementation Of Milk Quality Monitoring System Based On Multifrequency Ultrasound

Milk, as a main source of the protein ingestion for human, and the quality safety is related to human health. Therefore, research on milk quality and safety monitoring are very important. Laboratory chemical analysis is the traditional detecting method of milk quality, can provide detailed composition information of milk. In present, it is urgently need a monitoring system which can be applied in the online, non-destructive monitoring and evaluation of milk quality. This paper puts forward multifrequency ultrasonic monitoring method of milk quality evaluation based on the ultrasonic detecting technology. The method detects the ultrasonic attenuation spectrum of normal milk and abnormal milk just like the additive milk or spoiled milk by different frequency ultrasound based on the attenuation mechanism between milk and ultrasound. The online, on-site, non-destructive ultrasonic monitoring and evaluating of milk quality was accomplished using ultrasonic multi-sensor information fusion technology.The main research works in this paper is the design and model implementation of milk quality monitoring system based on multifrequency ultrasound. This system uses embedded system and peripheral circuits to complete the design of hardware system. The driving signal generator using AD9850DDS chip can meet the needs of stable output of1to30MHz sinusoidal signal with step of1MHz,0.1MHz, and0.01MHz. In this detecting system, we use three groups of1.25MHz,2.5MHz and5MHz ultrasonic transducers as detecting elements. During the detecting process, the driving signal is amplified by power unit to implement the emission of ultrasound. After the ultrasound transmitting through milk, the amplitude of emission signal is attenuated, and then the ultrasonic detecting signal is received by the receiving ultrasonic transducer which transforms ultrasonic signal to electric signal, and sent the signal to amplifier. The signal acquisition module completes the data acquisition. Analysis of ultrasonic detecting signal is done by the analysis software system of ultrasonic detecting signal on PC. The analysis software system consists of original data input, power transform, feature extraction, storage and playback modules.In an ultrasonic detecting system, a variety of factors, such as circuit, environment will affect the stability and accuracy of detecting results, the process of system design and its realization occupy a lot of time. In order to verify the principle of multifrequency ultrasonic monitoring of milk quality, we used pure water, bagged yoghurt drinks containing kinds of food additives and bagged milk as samples to finish the validation experiments.2700groups of ultrasonic attenuations by different samples are studied under different driving signals and ultrasonic transducers. Under the same detecting conditions,100groups of ultrasonic attenuations were divided into10small groups to calculate the mean value and do T test among the samples each other. The detecting results show when the compositions of detecting samples are different, the ultrasonic attenuation is different on each frequency point, and especially there are significant differences (p<0.05) in special frequency point. So the principle of multifrequency ultrasonic detecting of milk quality is right and executable. The150groups of ultrasonic attenuations by water on1.25MHz,2.5MHz,5MHz as main frequency point were analyzed, the variation coefficient were0.16%,0.11%and0.25%respectively. It shows that the system has a high reliability and stability. Therefore, this system can be used in the online, on-site, nondestructive, monitoring and evaluating of milk quality further.