1 M-P神经元模型的工作原理和几何意义
1943年,MoCulloch和Pitts[1]根据神经元传递规律,第一次提出了神经元的数学模型.M-P神经元模型一直沿用至今,它对神经网络的发展起到了奠基性的作用.每个神经元的状态由M-P方程决定:S=f(∑W X -θ)...1 M-P神经元模型的工作原理和几何意义
1943年,MoCulloch和Pitts[1]根据神经元传递规律,第一次提出了神经元的数学模型.M-P神经元模型一直沿用至今,它对神经网络的发展起到了奠基性的作用.每个神经元的状态由M-P方程决定:S=f(∑W X -θ),θ为阈值,f为激励函数,一般取符号函数.令:它代表了n维空间中,以X为坐标变量,以W为坐标系数,θ为常数项的一个超平面.当样本点X落入超平面的正半区,即I(X)>0时,有f(I)＝1;当样本点X落入超平面的负半区,即I(X)<0时,有f(I)＝0.从分类的角度看,一个神经元按输入将样本划分成为两类(0和1).现在广泛使用的BP模型采用Sigmoid函数作为激励函数,但是它没有改变神经元分类的本质.神经网络实际上就是多个神经元组织起来的一种网状结构.展开更多
A new insight into the constant current-constant voltage （CC-CV） charge protocol based on the spherical diffusion model was presented. From the model, the CV-charge process compensates, to a large extent, the capaci...A new insight into the constant current-constant voltage （CC-CV） charge protocol based on the spherical diffusion model was presented. From the model, the CV-charge process compensates, to a large extent, the capacity loss in the CC process, and the capacity loss increases with increasing the charging rate and decreases with increasing the lithium-ion diffusion coefficient and using a smaller r value （smaller particle-size and larger diffusion coefficient） and a lower charge rate will be helpful to decreasing the capacity loss. The results show that the CC and the CV charging processes, in some way, are complementary and the capacity loss during the CC charging process due to the large electrochemical polarization can be effectively compensated from the CV charging process.展开更多
文摘1 M-P神经元模型的工作原理和几何意义
1943年,MoCulloch和Pitts[1]根据神经元传递规律,第一次提出了神经元的数学模型.M-P神经元模型一直沿用至今,它对神经网络的发展起到了奠基性的作用.每个神经元的状态由M-P方程决定:S=f(∑W X -θ),θ为阈值,f为激励函数,一般取符号函数.令:它代表了n维空间中,以X为坐标变量,以W为坐标系数,θ为常数项的一个超平面.当样本点X落入超平面的正半区,即I(X)>0时,有f(I)＝1;当样本点X落入超平面的负半区,即I(X)<0时,有f(I)＝0.从分类的角度看,一个神经元按输入将样本划分成为两类(0和1).现在广泛使用的BP模型采用Sigmoid函数作为激励函数,但是它没有改变神经元分类的本质.神经网络实际上就是多个神经元组织起来的一种网状结构.
基金Projects(20676152, 20876178) supported by the National Natural Science Foundation of China
文摘A new insight into the constant current-constant voltage （CC-CV） charge protocol based on the spherical diffusion model was presented. From the model, the CV-charge process compensates, to a large extent, the capacity loss in the CC process, and the capacity loss increases with increasing the charging rate and decreases with increasing the lithium-ion diffusion coefficient and using a smaller r value （smaller particle-size and larger diffusion coefficient） and a lower charge rate will be helpful to decreasing the capacity loss. The results show that the CC and the CV charging processes, in some way, are complementary and the capacity loss during the CC charging process due to the large electrochemical polarization can be effectively compensated from the CV charging process.
