基于声振响应法的香梨硬度无损检测

为了探索香梨硬度无损检测的方法,该研究搭建了由压电梁式传感器进行信号激励和感测的检测装置,分析了装置信号检测的稳定性,提取了香梨共振频率和声速并进行香梨硬度评估,然后将这两响应参数的评估硬度与香梨硬度的M-T穿刺法(Magness-Taylor)测量结果进行线性回归分析,以构建香梨硬度的检测模型。结果表明,检测装置的共振频率和声速信号检测稳定可靠,并且在香梨赤道部不同位置激励感测均无显著差异(P0.05);基于共振频率法和声速法的硬度评估指标可以构建3种香梨硬度检测模型,相关系数分别为0.841、0.877和0.938,其中由共振频率检测硬度和声速检测硬度联合构建的检测模型硬度检测敏感度为67.30%,最接近M-T穿刺法检测敏感度,该模型对香梨正常果和粗皮果的检测准确率较高,达到86.7%。研究结果可为声振法无损检测香梨硬度提供参考。

Non-destructive detection of Korla pear stiffness based on acoustic vibration measurement

Abstract: Korla pear is one of the characteristic fruits in Xinjiang, and it has a wide international market. The quality of pear is related to its economic value. An effective and inexpensive non-destructive detection method is needed to reliably evaluate the internal quality of Korla pear. Stiffness is one of the most important indices to indicate fruit internal quality. Traditionally, Korla pear stiffness is quantitatively measured with the method of M-T test. This method is not suitable for commercial application because it can destroy the fruit when measuring. Thus, this study presents a measuring device based on the acoustic impulse response of Korla pear for non-destructively determining its stiffness. The measuring device consists of vibration control and dynamic signal analyzer, voltage amplifier, software system of vibration measurement and analysis, and test bench installed with 2 piezo beam transducers (one operates as an actuator, and the other as a sensor). In order to simulate hammer tapping signal, a positive half-sinusoid pulse with a peak amplitude of 2.5 V was generated by the vibration control and dynamic signal analyzer. This output signal was amplified by the voltage amplifier, which offered an impulsive excitation with an 80 V peak amplitude to the actuator, which was in contact with pear. Then the sensor, which was positioned at the opposite side of the actuator and also in contact with pear, detected the response signal of fruit. Both the excitation and response signal were acquired by the signal analyzer. The resonance frequency was obtained by the FFT of the response signal with the software system. The sound propagation velocity was calculated from the distance of the 2 contact points between pear and sensors divided by the lag time between excitation and response signal, which was determined by the cross correlation analysis. The fruit stiffness could be measured by the resonance frequency or sound propagation velocity. Different stiffness detection models were acquired by the linear regression analysis of the relationships between the stiffness obtained by resonance frequency and sound propagation velocity detection method and the stiffness measured with M-T method. The sensitivities of the different stiffness detection models were tested. Also, the discriminating rates of normal pear and rough-skinned pear by different stiffness detection models were compared. The results of repeated excitation at the same point showed that the acquired response signal from the system was stable and reliable. The differences of the resonance frequency and sound propagation velocity detected from different sensing points at the pear equator were not significant (P>0.05). All the stiffness detection models showed a good correlation with high correlation coefficient. The correlation coefficient (r=0.938) of the model based on resonance frequency detection stiffness combined with sound propagation velocity detection stiffness was significantly higher than the model based on resonance frequency detection stiffness (r=0.841) or sound propagation velocity detection stiffness (r=0.877), and its sensitivity was also the highest among the 3 models, which reached 67.30% and was very similar to that of the M-T method (68.49%). When the normal pear and rough-skinned pear were detected with this model, the discriminating rate was 86.7%, which was higher than that of the model based on resonance frequency detection stiffness (78.3%) or sound propagation velocity detection stiffness (80.0%). The results have proved that the acoustic vibration measurement is feasible for non-destructive detection of Korla pear stiffness, and can provide theoretical support for the commercial application of the new detection method.