In　recent　years,　the　market　for　real-world　robots　has　become　more　and　more　extensive;　robots　in　warehousing　and　logistics,　public　services,　home　management,　and　other　fields　are　becoming　more　and　more　popular.　The　market　demand　for　better　navigation　algorithms　is　becoming　stronger,　and　algorithms　based　on　reinforcement　learning　are　the　key　focus　of　future　research.　In　this　overview,　we　analyze　the　existing　three　main　types　of　problems　in　this　field　and　summarize　the　existing　cutting-edge　algorithms.　Then,　we　reveal　the　remaining　problems　and　optimizations　in　this　area　and　give　possible　solutions　and　future　research　directions.　©　The　Author(s),　under　exclusive　license　to　Springer　Nature　Singapore　Pte　Ltd.　2024.