This could cause optimization issues because shifting the turbine slightly would not cause a change in power (gradient = 0). Is there a good way we could add a spline or a sin/cos function to the overlap or effective U function so the optimizer will have a non-zero gradient to work with?
This could cause optimization issues because shifting the turbine slightly would not cause a change in power (gradient = 0). Is there a good way we could add a spline or a sin/cos function to the overlap or effective U function so the optimizer will have a non-zero gradient to work with?