The　complex　interactions　were　performed　among　non-pharmaceutical　interventions,　vaccinations,　and　hosts　for　all　epidemics　in　mainland　China　during　the　spread　of　COVID-19.　Specially,　the　small-scale　epidemic　in　the　city　described　by　SVEIR　model　was　less　found　in　the　current　studies.　The　SVEIR　model　with　control　was　established　to　analyze　the　dynamical　and　epidemiological　features　of　two　epidemics　in　Jinzhou　City　led　by　Omicron　variants　before　and　after　Twenty　Measures.　In　this　study,　the　total　population　(N)　of　Jinzhou　City　was　divided　into　five　compartments:　the　susceptible　(S),　the　vaccinated　(V),　the　exposed　(E),　the　infected　(I),　and　the　recovered　(R).　By　surveillance　data　and　the　SVEIR　model,　three　methods　(maximum　likelihood　method,　exponential　growth　rate　method,　next　generation　matrix　method)　were　governed　to　estimate　basic　reproduction　number,　and　the　results　showed　that　an　increasing　tendency　of　basic　reproduction　number　from　Omicron　BA.5.2　to　Omicron　BA.2.12.1.　Meanwhile,　the　effective　reproduction　number　for　two　epidemics　were　investigated　by　surveillance　data,　and　the　results　showed　that　Jinzhou　wave　1　reached　the　peak　on　November　1　and　was　controlled　7　days　later,　and　that　Jinzhou　wave　2　reached　the　peak　on　November　28　and　was　controlled　5　days　later.　Moreover,　the　impacts　of　non-pharmaceutical　interventions　(awareness　delay,　peak　delay,　control　intensity)　were　discussed　extensively,　the　variations　of　infection　scales　for　Omicron　variant　and　EG.5　variant　were　also　discussed.　Furthermore,　the　investigations　on　peaks　and　infection　scales　for　two　epidemics　in　dynamic　zero-COVID　policy　were　operated　by　the　SVEIR　model　with　control.　The　investigations　on　public　medical　requirements　of　Jinzhou　City　and　Liaoning　Province　were　analyzed　by　using　SVEIR　model　without　control,　which　provided　a　possible　perspective　on　variant　evolution　in　the　future.　©　2024　The　Authors