The sine function is non-concave and non-convex, which is difficult to handle.
I would like to ask whether the sine function can adopt the first-order Taylor expansion, sin(x)=sin(xl)+cos(xl)(x-xl) with the xl is the feasible point in the l-th iteration.
Since the sine function rarely appears in optimization, I don’t know whether it’s reasonable to do so
I am a student who is new to optimization. I hope someone can put forward their valuable suggestions and ideas.
The same answer I gave in the other thread you posted in:
Trigonometric functions are neither convex nor concave,. They can only be “convexified” by means of an approximation, such as a suitable Taylor series. The one term Taylor series approximation for sin(x) is x, and the two term Taylor series approximation is x - x^3/6, is convex for x < 0 and concave for x > 0, hence neither convex nor concave…Any higher order Taylor series approximation for sin(x) will be neither convex nor concave. cos(x) does allow the Taylor series approximation 1-x^2/2, which is concave, and may or may not be adequate for purposes of optimization.
I recommend you use a non-convex nonlinear solver, which you can not do via CVX. You can consider using YALMIP…