Range of transmission line

I find that the concept of sending/receiving end voltage is somewhat meaningless in most cases. Actual transmission lines are connected in complex networks with distributed impedance (the line itself) and point impedance (loads tapped off) all over the network. There is no set beginning or end.

In the real world you have to keep all points on the line within the stated tolerance of the nominal voltage. Typically that's +/-10%. At lower, distribution level voltages where the source impedance is higher, they will often times use voltage regulators or on-load tap changers to maintain the bus voltage dynamically.

But for your text book question I think that they want the receiving end voltage to be the nominal 220kV and the sending end voltage would be chosen to allow for drop.

First technique : Model the load as a disturbance that tries to prevent you from reaching your goal, i.e. regulating the voltage. That way, you will have 2 transfer functions. The first one will be

G(s) = y(s)/u(s) where y is your output, u your input (setpoint)

H(s) = y(s)/d(s) where d is your disturbance (i.e load). d(s) might a be a step of a certain number of amps for example.

You simply have to balance the 2 transfer functions. For example, you could aim for a 60-degree phase margin for both.

Second technique,

Model 2 transfer functions and try to optimize both of them (phase margin, gain margin, sensitivity, etc.)

Gempty(s) = y(s)/u(s) at no load.

Gfull(s) = y(s)/u(s) at full load.

I don't know exactly what kind of power systems you're modeling, but for buck and boost converter for example, the transfer function in continuous-conduction mode is barely affected by load variations, while in discontinuous-conduction mode, a load variation can make you unstable.