CAJ | 학술논문

区域物流能力对区域经济的增长有着强劲的推动作用,为准确地预测区域物流能力大小及其变化趋势,将灰色系统理论中的GM预测模型与径向基函数神经网络有效地结合起来,综合了灰色系统贫乏数据建模的优点和神经网络特有的高度非线性映射能力,构建了基于灰色径向基函数网络的非线性组合预测模型.然后,以四川省1997-2005年的物流能力量化值为基础,对2006-2008年的四川省物流能力值进行了短期预测,并分析了预测结果的合理性.研究表明,该组合模型优于任何单一灰色预测模型,能很好地反映区域物流能力的变化规律,在小样本、贫信息的条件下,仍然能得到合理精准的预测结果,具有实用性.
구역물류능력대구역경제적증장유착강경적추동작용,위준학지예측구역물류능력대소급기변화추세,장회색계통이론중적GM예측모형여경향기함수신경망락유효지결합기래,종합료회색계통빈핍수거건모적우점화신경망락특유적고도비선성영사능력,구건료기우회색경향기함수망락적비선성조합예측모형.연후,이사천성1997-2005년적물류능역량화치위기출,대2006-2008년적사천성물류능력치진행료단기예측,병분석료예측결과적합이성.연구표명,해조합모형우우임하단일회색예측모형,능흔호지반영구역물류능력적변화규률,재소양본、빈신식적조건하,잉연능득도합리정준적예측결과,구유실용성.
Regional logistics capability (RLC) is a powerful impetus to the growth of the regional economy. In order to forecast the scale and variation tendency of RLC accurately,by adopting a method of combined forecasting which combines GM of grey system theory with radial basis function (RBF) neural network together effectively and gives full scope to their double-edged advantages that the grey system can construct forecasting model with poor information and neural network is capable of high non-linear mapping uniquely,a new kind of non-linear combined forecasting model of RLC based on grey RBF neural network was constructed. Then,according to the quantization values of RLC of Sichuan Province from 1997 to 2005,the values of RLC from 2006 to 2008 were forecasted,and the reasonableness of short-term forecasting results was analyzed. The study shows that (1) the combined model is superior to any single grey model and it can reflect the change regularity of RLC perfectly;(2) it can still get accurate forecasting results even if under the condition of little sample and poor information,and has practicability.