摘要
利用层次分析法确定农产品综合运输水平效能的综合评价指标体系,依据此指标体系建立单目标农产品综合运输水平效能的综合评价模型,并利用优化后的神经网络对模型进行训练求解。训练结果与传统神经网络训练结果和线性规划软件计算结果相比,所获得的综合运输最佳方案相同,而优化后的神经网络比传统神经网络训练时间短,迭代次数少,拟合误差小。算例表明:此模型具有良好的泛化能力,可对农产品运输水平效能做出有效的综合评价。
Analytic hierarchy process was used to determine the agricultural products comprehensive evaluation index system. According to this index system, a single objective and comprehensive evaluation model was established for agricultural products. The results were gotten by using the optimized neural network model for training. The comparison of training results, traditional neural network training results and linear programming optimization algorithm showed almost the same. The optimized neural network showed some traits such as short training time, few number of iterations, fitting high degree. Examples showed that this model has good generalization ability, and makes effective comprehensive evaluation to the transportation performance of agricultural products.
出处
《沈阳农业大学学报》
CAS
CSCD
北大核心
2015年第5期634-640,共7页
Journal of Shenyang Agricultural University
基金
山东省自然科学基金项目(ZR2010GL008)
关键词
农产品运输评价指标
综合运输水平评价
优化的神经网络
农产品
变尺度混沌
evaluation scale of agricultural products transportation
comprehensive evaluation of integrated transport level
optimized model of neural network
agricaltural products
variable-metric and chaos
