Energy　storage　power　stations　are　endowed　with　excellent　regulatory　capabilities,　capable　of　achieving　frequency　regulation　on　a　minute　scale.　However,　the　current　proportion　of　power　and　capacity　of　energy　storage　power　stations　in　the　electrical　grid　is　limited,　which　only　permits　them　to　assist　thermal　power　units　in　accomplishing　grid　frequency　regulation.　Consequently,　considering　the　distinct　response　characteristics,　a　multi-scale　frequency　regulation　coordination　strategy　for　thermal　power　and　energy　storage　is　proposed.　Furthermore,　a　rolling　optimization　control　method　is　employed　to　enhance　frequency　regulation　performance.　Initially,　a　high-order　frequency　domain　model　of　the　dynamic　characteristics　of　thermal　power　units　is　constructed.　The　model　is　reduced　through　partial　fraction　expansion,　and　the　frequency-time　domain　conversion　of　the　dynamic　output　characteristics　of　thermal　power　units　is　realized　based　on　the　inverse　Laplace　transform.　Subsequently,　based　on　the　mechanism　of　capacity　degradation　due　to　power　exchange,　the　capacity　degradation　calculation　deviation　is　corrected　using　the　rain　flow-counting　algorithm,　and　an　energy　storage　capacity　degradation　model　is　established.　Thereafter,　a　multi-scale　frequency　regulation　collaborative　strategy　for　thermal　power　and　energy　storage　was　proposed,　wherein　the　scheduling　period　length　of　units　is　expressed　as　a　constraint　on　the　number　of　unit　output　changes.　By　employing　a　relaxation-tightening　constraint　method,　the　scheduling　period　length　for　the　frequency　regulation　of　units　is　adjusted　to　match　the　response　characteristics　of　the　units　with　the　frequency　regulation　scheduling　period　length.　Ultimately,　the　method　proposed　is　validated　through　numerical　examples,　demonstrating　its　effectiveness　in　reducing　the　system’s　frequency　regulation　deviation　and　enhancing　the　stability　of　the　grid’s　frequency.　©　2025　Power　System　Technology　Press.　All　rights　reserved.