Feedforward multi layer neural networks have very strong mapping capability that is based on the non linearity of the activation function, however, the non linearity of the activation function can cause the multiple ...Feedforward multi layer neural networks have very strong mapping capability that is based on the non linearity of the activation function, however, the non linearity of the activation function can cause the multiple local minima on the learning error surfaces, which affect the learning rate and solving optimal weights. This paper proposes a learning method linearizing non linearity of the activation function and discusses its merits and demerits theoretically.展开更多
冲压空气涡轮(Ram Air Turbine,RAT)最大释放冲击载荷是飞机结构设计重要参数。当前RAT释放冲击载荷的试验仅试飞或高速风洞试验可以得到,寻找一种有效的RAT最大冲击载荷预测方法很有必要。通过分析得到RAT最大释放冲击载荷的影响因素...冲压空气涡轮(Ram Air Turbine,RAT)最大释放冲击载荷是飞机结构设计重要参数。当前RAT释放冲击载荷的试验仅试飞或高速风洞试验可以得到,寻找一种有效的RAT最大冲击载荷预测方法很有必要。通过分析得到RAT最大释放冲击载荷的影响因素与飞行高度和飞行空速有直接关系,采用线性回归及BP神经网络研究飞行高度和空速对RAT最大释放载荷的影响,并从平均绝对误差及均方根百分误差进行评价。研究将试验及仿真结果作为训练样本,训练完成后将已知输入层参数输入后预测RAT最大释放载荷。对比预测最大释放载荷与试验结果,线性回归预测值平均绝对误差及均方根百分误差小于10%,BP神经网络预测值平均绝对误差及均方根百分误差的平均值小于5%。展开更多
文摘Feedforward multi layer neural networks have very strong mapping capability that is based on the non linearity of the activation function, however, the non linearity of the activation function can cause the multiple local minima on the learning error surfaces, which affect the learning rate and solving optimal weights. This paper proposes a learning method linearizing non linearity of the activation function and discusses its merits and demerits theoretically.
文摘冲压空气涡轮(Ram Air Turbine,RAT)最大释放冲击载荷是飞机结构设计重要参数。当前RAT释放冲击载荷的试验仅试飞或高速风洞试验可以得到,寻找一种有效的RAT最大冲击载荷预测方法很有必要。通过分析得到RAT最大释放冲击载荷的影响因素与飞行高度和飞行空速有直接关系,采用线性回归及BP神经网络研究飞行高度和空速对RAT最大释放载荷的影响,并从平均绝对误差及均方根百分误差进行评价。研究将试验及仿真结果作为训练样本,训练完成后将已知输入层参数输入后预测RAT最大释放载荷。对比预测最大释放载荷与试验结果,线性回归预测值平均绝对误差及均方根百分误差小于10%,BP神经网络预测值平均绝对误差及均方根百分误差的平均值小于5%。
