基于BP人工神经网络中国光伏产业贸易摩擦预警研究

建立有效的贸易摩擦预警系统有利于中国新兴产业的健康发展,有利于中国实施制造强国战略。将警兆信号法与人工神经网络相结合,从宏观经济形势、产业供给能力、双边贸易状况、市场需求水平等四个维度切入,构建了中国光伏产品出口贸易摩擦预警系统模型。预警分析表明,模型的预警效果与现实拟合较好,具有可行性。从模型的预警结果来看,2019年~2020年中国输美光伏产品贸易摩擦警情均低于轻警级别,仍面临一定的贸易摩擦风险。

Research on trade friction early warning of China’s photovoltaic industry based on BP artificial neural network

The establishment of an effective early warning system for trade frictions is conducive to the healthy development of China’s emerging industries and China’s implementation of the strategy for making the manufacturing country powerful. Combining the warning signal method and artificial neural network, this paper builds the early warning system model for China’s photovoltaic product export trade friction from the four dimensions of the macroeconomic situation, industrial supply capacity, bilateral trade status and market demand level. The early warning analysis shows that the early warning effect of the model fits well with the reality, which is feasible. From the early warning results of the model, the trade friction warnings of China’s photovoltaic products exported to the United States from 2019 to 2020 are all below the level of light alarm, yet with certain risks of trade friction.