Research On Evaluation Technology Of Mutton Eating Quality

With the improvement of our living standards and the change of consumption structure, The mutton consumption of our country was rising year by year. In 2009 the output reached 389 tons, first in the world, and the imports nearly 60,000 tons. Although our country's the great power of mutton production and consumption, but not processing. The development of mutton processing was slow and the mutton products were good and bad intermingled. 90 percent of the mutton products without classification cause the phenomenon of the special quality not prevalent. thus it is an inexorable trend of the development of mutton industry to establish a rapid, nondestructive and online mutton eating quality evaluation method. This paper researchs the eating quality characteristics, and develop a set of methods which could fastly quantitatively analyze and forecast mutton eating quality by sensory evaluation, TPA method and near infrared spectroscopy, then to establish and perfect the mutton eating quality evaluation system. For the intelligent classification of mutton provided scientific basis. Specific research as follows:(1) The indicators of mutton eating quality was screened by M value method, principal component analysis and correlation analysis coupled with sensory evaluation by means of 21 pairs of mutton m. longissimus thoracis and hindquarters. It was concluded that the indicators of mutton eating quality comprised appearance, texture and flavor, and the appearance comprise moisture; the texture comprised hardness tenderness and juiciness; the flavor comprised flavor intensity and the smell of mutton flavor. The result of correlation analysis shows that the moisture has a strong relationship with juicy(R=0.62), hardness and meat flavor both have a strong relationship with tenderness(R=-0.72,-0.57). So the mutton eating quality can be evaluated by smell of mutton flavor,tenderness and juicy.(2) In order to use TPA instead of sensory evaluation and reduce sensory evaluation of man-made factors, the correlation analysis of sensory evaluation indicators and the results of TPA analysis. Found that the smell of mutton flavor had a general correlation with hardness and flexibility (TPA) (R = 0.42, 0.42); Tender and juicy had a significant relation with flexibility (TPA) (R =0.64, 0.69). The model of sensory indicators were established by regression analysis. Three models were significant (P < 0.05), and explained variance was 54.57%, 43.66%, 73.41%, respectively.(3) For intelligent forecasting smell of mutton flavor, tenderness and juicy, the model of eating quality was established by near infrared spectrum, cross examination. It was concluded that the validation of the model of smell of mutton flavor, tenderness and juicy were carried out, the R~2 was 0.60, 0.67, 0.73 respectivelly and significantly (p < 0.05).(4) The classifier was established by fisher linear discrimination with 52 mutton samples and then validated. It is concluded that the correct recognition rate of first level mutton reached 90.91%; the second level reached 80%; the third level reached 87.5%. The validstion was made with the data from TPA and NIR. It is concluded the NIR method is superior to the TPA methods both on accuracy and application.