对BP人工神经网络在农产品的灰色仿真评估上应用的探讨

Disscussion of ANN apply the gray simulated evaluation to the primary products

基于已有的农产品商品性状综合评价体系,提出了人工神经网络(ANN)农产品商品性状评价模型,进行了其网络在番茄商品加工性状评价的具体应用,并将其评价结果与已有的评判结果进行比较。结果表明,7个品系中只有品系YH02-7、YH02-11、YH02-8、YH02-5与专家排序相差一个位次,其他均相同。说明建立的模型实用、可操作性强,评价结果可靠,具有较好的应用前景,为全面评价农产品商品性状提出了新的思路和方法。

Base on the known synthetic evaluation system applied to evaluate the trade traits of primary products, the artificial neural networks （ANN） model for the trade traits of primary products was proposed, applied to evaluating of the processing characteristics in the trade traits of the fruits of tomato, and compared with the evaluation by the known evaluation results. The results showed that only the line YH02-7, YH02-11, YH02-8, YH02-5 in 7 lines compared with the order by the expert were different from a sequence order, the other lines were the same order. So the Artificial Neural Network Model was more practical and operable. The evaluation results by the model were reliable. Thus, it has good future for applications. The new thinking and method are provided for overall evaluation to the trade traits of primary products.