Assessing the Freshness of Meat by Using Quantum-Behaved Particle Swarm Optimization and Support Vector Machine

Abstract

To improve the performance of meat freshness identification systems, we present a new identification method based on quantum-behaved particle swarm optimization (QPSO) and the support vector machine (SVM). Fresh pork, beef, mutton, and shrimp samples were stored in a hypobaric chamber for several days, and the conventional indices of meat freshness, including total volatile basic nitrogen content, aerobic plate count, pH value, and sensory scores, were determined to achieve the identification of sample freshness. However, the experiments showed that it was difficult to obtain an ideal freshness assessment by any single physicochemical or sensory property. Therefore, SVM was introduced to use these data to build a freshness model. Furthermore, QPSO was proposed to seek the optimal parameter combination of SVM. The experimental results indicated that the hybrid SVM model with QPSO could be used to predict meat freshness with 100 % classification accuracy.