【摘要】 根据径流量的影响因素往往具有滞后性的特点,在构建神经网络时,从模型复杂度、训练精度、预测精度等方面综合分析了该特性的影响大小,获得了分析滞后性影响的方法。实例表明该方法能够准确判断出滞后时段的大小,为提高径流预报的准确性提供了一条有效的途径。更多还原

【Abstract】 The hysteretic nature of runoff’s influencing factors’is considered,the influencing degree of the hysteretic nature is tested from aspects of complexity,training precision,prediction accuracy of model during structuring ANN model,then the method of analyzing hysteretic nature’s influence is advanced.The result of example study demonstrates that the lag effect time can been accurately predicted by the way,an effective way is given to improve the accuracy of runoff forecast.更多还原