The challenge here is seems to be to ensure that if the reviewers do stumble across your thesis, then the failure mode will be penetrating blinding rather than accusations of plagiarism.
It is my belief that with an "extract" paper like this, the thesis should be cited in any case. In most cases, there will be some connection to other portions of the thesis that could motivate such a citation (e.g., a motivation or an application). I also think that it is good to explicitly acknowledge the relationship to the thesis, e.g., "This manuscript is based on work also presented in [cite]", though the customs of your field may differ.
Then you can appropriateappropriately blind the citation to the thesis, e.g., "Ph.D. thesis, blinded for review." This makes the relationship clear without violating blinding. At that point, you are preserving blinding to the best of your ability, and while a reviewer can certainly try to penetrate blinding if they want, you certainly won't run into any problems with misunderstanding about plagiarism.