I don't recommend to use the notation $d_\mu$ since the derivative that appears in the E.L equations of motion is partial derivative $\partial_\mu$. Then if the Lagrangian density depends on the field $y$ and its derivatives $\partial_\mu y$ the E.L equations of motion read
$\begin{equation}
\partial_\mu\dfrac{\partial \mathcal L}{\partial (\partial_\mu y)}-\dfrac{\partial \mathcal{L}}{\partial y}
\end{equation}=0
$
Then if you want to find solutions to the E.L equations you will need to consider a specific expression for the Lagrangian density. In your case, you consider that the field can be expanded on a superposition of fields
$y=c_0y_0+c_1 y_1+c_2y_2+\ldots=\displaystyle\sum_{i=1}^{\infty}c_i y_i$
Then note that the Lagrangian density becomes a function of that many fields and its derivatives $\mathcal{L}(\partial_\mu y, y)\equiv \mathcal{L}(\partial_\mu y_0,\partial_\mu y_1,\ldots, y_0,y_1,\ldots)$. So that when you derive (through the variational principle) the Euler-Lagrange equations of motion you get a system of (possibly coupled) differential equations for the $y_0,y_1,\ldots$ fields.
$\begin{equation}
\displaystyle\sum_{i=1}^{\infty}\bigg[\partial_\mu\dfrac{\partial \mathcal L}{\partial (\partial_\mu y_i)}-\dfrac{\partial \mathcal{L}}{\partial y_i}\bigg]
\end{equation}=0.
$