please tell me the reasons for this situation,thank you very much.
Try search this forum for exponential cone and you will learn a lot about what you can do.
Btw no one can tell you what the issue is based on the limited information you provide.
You shouldn’t worry about the unbounded listed in the Successive Approximation log. The final assessment by CVX is that the problem is infeasible. All but the first section of Debugging infeasible models - YALMIP also applies to CVX.