中文说明：
自己编写RBF神经网络程序，RBF神经网络隐层采用标准Gaussian径向基函数，输出层采用线性激活函数，其中数据中心、扩展常数和输出权值均用梯度法求解，它们的学习率均为0.001。其中隐节点数选为10，初始输出权值取[-0.1，0.1]内的随机值，初始数据中心取[-1，1]内的随机值，初始扩展常数取[0.1，0.3]内的随机值，输入采用[0 1]的随机阶跃输入
English Description：
The hidden layer of RBF neural network adopts standard Gaussian radial basis function, and the output layer adopts linear activation function. The data center, expansion constant and output weight are solved by gradient method, and their learning rate is 0.001. Among them, the number of hidden nodes is 10, the initial output weight is taken as the random value within [- 0.1, 0.1], the initial data center is taken as the random value within [- 1, 1], the initial expansion constant is taken as the random value within [0.1, 0.3], and the input adopts the random step input of [0.1]