In order to improve the accuracy of the measurement of the export product quality, the existing quality estimation method was modified, a measurement model of China's export product quality was established, and its nonlinearity through the functional link neural network was self-corrected. The results show that the export product quality indices of industries are mostly larger than 0 and distribute evenly, namely there are three kinds of industries whose qualities are high, medium and low, respectively. In general, the average level of export products quality has a downward trend, while there are four industries in which the quality level of export products presents upward trend. In terms of the nonlinear self-correction function of the measurement model, it can better reflect the evolution path of product quality whose actual trend is time nonlinear, and can enhance the accuracy of the measurement of the export product quality significantly.%为提高机电类出口产品质量测度的精度,对已有的质量估算方法进行修正,构建衡量机电类出口产品质量测度估计模型,并采用函数链神经网络对其非线性自校正.研究结果表明:各行业机电类出口产品质量指数大多在0以上,且分布较均衡,质量指数在高位、中位和低位上运行的行业各3种;各行业出口产品的平均质量指数总体上呈持续下降趋势,但其中4个行业出口产品的质量指数呈上升趋势,其他行业呈下降趋势;具有非线性自校正功能的机电类出口产品质量测度模型更能反映产品质量演进路径呈时间非线性变化实际趋势,可在较大程度上提高机电类出口产品质量测度精度.
展开▼
