pinn二维转换成三维报错

最新推荐文章于 2024-06-26 18:25:09 发布

weixin_56937768

最新推荐文章于 2024-06-26 18:25:09 发布

阅读量215

点赞数 3

文章标签： python

本文链接：https://blog.csdn.net/weixin_56937768/article/details/138341063

版权

在使用TensorFlow进行板改案例的PINN（物理informedneuralnetworks）训练时，遇到MatMul操作的维度不匹配错误，问题出在输入矩阵的维度上。作者在尝试解决矩阵不配对问题，但尚未找到具体输入矩阵的问题所在。

摘要由CSDN通过智能技术生成

pinn二维转换成三维报错ValueError: Dimensions must be equal, but are 3 and 4 for 'MatMul_503' (op: 'MatMul') with input shapes: [?,3], [4,70].

这是我的traceback

Traceback (most recent call last):

File "C:\Users\yyx\Desktop\复材\案例\板改\PlateHoleQuarter\train\train.py", line 1032, in <module>
model = PINN(XYZT_c, IC, LF, RT, UP, LW, DIST, uvw_layers, dist_layers, part_layers, lb, ub)

File "C:\Users\yyx\Desktop\复材\案例\板改\PlateHoleQuarter\train\train.py", line 177, in __init__
self.u_pred, self.v_pred, self.w_pred, self.s11_pred, self.s22_pred, self.s12_pred, self.s33_pred, self.s13_pred, self.s23_pred = self.net_uvw(self.x_tf, self.y_tf, self.z_tf, self.t_tf)

File "C:\Users\yyx\Desktop\复材\案例\板改\PlateHoleQuarter\train\train.py", line 418, in net_uvw
uvw_sig = self.neural_net(tf.concat([x, y, z, t], 1), self.uvw_weights, self.uvw_biases)

File "C:\Users\yyx\Desktop\复材\案例\板改\PlateHoleQuarter\train\train.py", line 357, in neural_net
H = tf.tanh(tf.add(tf.matmul(H, W), b))

File "C:\Users\yyx\.conda\envs\tensorflow-pinn\lib\site-packages\tensorflow\python\ops\math_ops.py", line 2018, in matmul
a, b, transpose_a=transpose_a, transpose_b=transpose_b, name=name)

File "C:\Users\yyx\.conda\envs\tensorflow-pinn\lib\site-packages\tensorflow\python\ops\gen_math_ops.py", line 4455, in mat_mul
name=name)

File "C:\Users\yyx\.conda\envs\tensorflow-pinn\lib\site-packages\tensorflow\python\framework\op_def_library.py", line 787, in _apply_op_helper
op_def=op_def)

File "C:\Users\yyx\.conda\envs\tensorflow-pinn\lib\site-packages\tensorflow\python\util\deprecation.py", line 454, in new_func
return func(*args, **kwargs)

File "C:\Users\yyx\.conda\envs\tensorflow-pinn\lib\site-packages\tensorflow\python\framework\ops.py", line 3155, in create_op
op_def=op_def)

File "C:\Users\yyx\.conda\envs\tensorflow-pinn\lib\site-packages\tensorflow\python\framework\ops.py", line 1731, in __init__
control_input_ops)

File "C:\Users\yyx\.conda\envs\tensorflow-pinn\lib\site-packages\tensorflow\python\framework\ops.py", line 1579, in _create_c_op
raise ValueError(str(e))

ValueError: Dimensions must be equal, but are 3 and 4 for 'MatMul_503' (op: 'MatMul') with input shapes: [?,3], [4,70].

在网上找了许多，说是矩阵不配对，但是不太清楚自己的输入矩阵在哪里