With　the　acceleration　of　urbanization,　the　energy　consumption　of　residential　communities　is　increasing.　Therefore,　researching　feasible　control　strategies　to　address　the　frequent　power　fluctuations　in　these　communities　is　of　great　significance.　Energy　router　(ER),　which　offers　advantages　such　as　fast　transmission　efficiency,　high　energy　utilization,　and　stable　operation,　is　a　key　technology　in　future　residential　community　systems.　This　paper　first　proposes　a　small-capacity　energy　router　system　suitable　for　residential　communities.　The　system　adopts　a　dual-energy　storage　solution　consisting　of　a　supercapacitor　and　a　storage　battery.　In　terms　of　control,　the　improved　model　predictive　control　(IMPC)　method　is　adopted　to　combine　PI　control　with　model　predictive　control,　and　the　fast　stability　characteristics　of　MPC　and　the　low　ripple　characteristics　of　PI　are　used　to　achieve　bus　voltage　stability.　Then,　the　operating　states　of　the　energy　router　are　divided　into　two　modes　and　seven　working　states,　aimed　at　stabilizing　the　bus　voltage　of　the　energy　router　and　controlling　the　state　of　charge　(SOC)　of　the　storage　battery.　The　analysis　of　the　two　operational　modes　of　the　energy　router　was　finally　conducted　on　the　RT-LAB　platform.　The　experimental　results　show　that　the　proposed　modular　control　method　stabilizes　the　bus　voltage　effectively　and　ensures　the　longevity　of　the　energy　storage　system,　demonstrating　robust　overall　stability.