井冈山大学学报(自然科学版)
井岡山大學學報(自然科學版)
정강산대학학보(자연과학판)
JOURNAL OF JINGGANGSHAN UNIVERSITY(SCIENCE AND TECHNOLOGY)
2014年
5期
29-32
,共4页
温玉锋%陈志铨%汤鹏杰%赖章丽
溫玉鋒%陳誌銓%湯鵬傑%賴章麗
온옥봉%진지전%탕붕걸%뢰장려
食品%比热容%支持向量回归%粒子群算法%预测
食品%比熱容%支持嚮量迴歸%粒子群算法%預測
식품%비열용%지지향량회귀%입자군산법%예측
food%specific heat capacity%support vector regression%particle swarm optimization%prediction
利用粒子群优化算法和支持向量回归方法建立不同食品的比热容与其水、蛋白质、碳水化合物和脂肪等含量间的预测模型,且在相同的训练样本和测试样本条件下,该预测模型的食品比热容预测精度高于反向传播神经网络模型,具有更强的泛化能力。结果表明:该预测模型能有效地预测食品比热容。
利用粒子群優化算法和支持嚮量迴歸方法建立不同食品的比熱容與其水、蛋白質、碳水化閤物和脂肪等含量間的預測模型,且在相同的訓練樣本和測試樣本條件下,該預測模型的食品比熱容預測精度高于反嚮傳播神經網絡模型,具有更彊的汎化能力。結果錶明:該預測模型能有效地預測食品比熱容。
이용입자군우화산법화지지향량회귀방법건립불동식품적비열용여기수、단백질、탄수화합물화지방등함량간적예측모형,차재상동적훈련양본화측시양본조건하,해예측모형적식품비열용예측정도고우반향전파신경망락모형,구유경강적범화능력。결과표명:해예측모형능유효지예측식품비열용。
The dependence model of specific heat capacity on the contents of water, protein, carbohydrate and fat for different foods was established using the particle swarm optimization algorithm and support vector regression approach. Furthermore, the prediction precision of the dependence model is higher than that of back propagation neural network for the same training and test samples. Its generalization ability is also stronger than that of back propagation neural network. The experiment and analysis shows that the dependence model can be used to effectively estimating the specific heat capacity of food.