Abstract: Abstract: Muskmelons, a typical fruit with excellent flavor produced in northwest of China, are always harvested from late July to mid-August. At time of harvest, it floods the local market. Transportation and circulation of the muskmelon outside the area are common and trucks are the most usual and important tools for the product transportation and distribution. However, the fruits are easily injured during the transportation because they contain high percentage of water in their fresh forms and are continuing to ripen during transportation resulting from their physiological function through post-harvest respiration. Injure of fruits caused by vibration during long distance transportation manifest typically itself in two aspects. One lies in acceleration due to vibration caused by repeated low-stress fatigue damage, the other is the vibration at resonant frequency, which is a special case in the random vibration. Resonant frequency is mainly affected by the damping force, followed with the object shape, mass and firmness and widely known to be harmful in most mechanical systems but few reports have been published about its harmfulness in fruits transportation. In addition, how to measure the resonant frequency and to further improve the accuracy of random vibration analysis of fruits transportation have not been reported. This study was to investigate the resonant frequency and to improve the accuracy of random vibration analysis in the vibration process through simulating actual transportation situation by the use of both window function and transferring analysis on the platform of DASP software, which enhanced useful information in signals and weaken those useless ones. The window functions essentially referred to limit the scope of signal and observation time, and data must be windowed in order to minimize spectrum leakage. To maximize the credibility of the data processing results and minimize errors due to spectrum leakage, Hanning window of DASP software was used in the test with the number of analysis points greater than 8 192 and overlapping coefficient 31/32. Resonant frequency of sample, which can be detected repeatedly through Hanning window functions, was the one when the maximum acceleration transferring rate was achieved during the whole sweep frequency process with a value of 35.38 Hz. Meanwhile, resonant frequencies of different ripeness groups of muskmelons with firmness of (7-8.2)×105, (5.4-6.8)×105 and (4.1-5.2)×105 Pa were achieved respectively with the value of 41.8-45.5, 36.2-38.87 and 33.25-36.8 Hz, showing significant positive correlation between resonant frequency and firmness of muskmelons with a correlation coefficient 0.9132. The study of the impact of both resonant frequency vibration and low frequency vibration on the qualities, such as firmness, soluble solid content (SSC) and damage volume rate (DVR), was carried out as well as the prediction model during storage. The low frequencies of 4 and 6 Hz were adopted in the test according to survey of truck transportation profiles combining with characteristics of muskmelons at the speed of 50-70 km/h in northwest of China. Methods were respectively used to determine the firmness, SSC and DVR. The results showed that the qualities of muskmelons always deteriorated during storage after vibration with prolonging vibration time under any vibration frequency. As both the firmness and SSC decreased, the DVR sharply increased. In addition, the DVR caused by vibration of resonant frequency was obviously greater than that of low frequency in the condition of long lasting time. Multiple linear regression models using storage time and quality parameters obtained showed better correlation (R2﹥0.95). This study demonstrated the feasibility of detecting the resonant frequency during fruits transportation vibration using window functions and of predicting shelf life of muskmelons postharvest, and also provides basic foundation for appropriate package avoiding resonant frequency in real transportation conditions.