Addressing　the　issues　of　complex　demolition　environments　in　nuclear　facility　decommissioning　projects,　the　difficulty　of　achieving　multi-axis　linkage　for　robotic　arms,　and　low　work　efficiency,　an　optimization　method　that　can　autonomously　reduce　the　moving　joints　of　robotic　arms　and　autonomously　find　the　best　cutting　path　is　proposed.　This　method　can　effectively　solve　for　the　cutting　path　that　requires　the　fewest　moving　joints　and　improve　work　efficiency.　Firstly,　forward　and　inverse　kinematics　analysis　is　conducted　on　the　robotic　arm,　dividing　it　into　two　parts:　the　pedestal　and　the　arm.　The　pedestal　is　responsible　for　moving　and　cutting,　while　the　arm　adjusts　the　posture.　Secondly,　for　the　arc　and　straight　trajectories　during　the　cutting　process,　corresponding　motion　synthesis　is　performed　on　the　pedestal.　The　inverse　kinematics　of　the　arm　is　solved　using　a　genetic　algorithm　based　on　the　principle　of　minimizing　the　degree　of　freedom　of　motion,　achieving　cutting　with　the　fewest　moving　joints.　Finally,　BP　neural　network　is　used　to　simulate　and　train　the　solved　data　samples,　establishing　a　path　optimization　model　that　realizes　optimization　of　the　cutting　path　for　the　robotic　arm.　This　algorithm　provides　reference　significance　for　nuclear　facility　decommissioning　robots　to　autonomously　reduce　the　degrees　of　freedom　required　for　cutting.　Through　simulation　verification　of　a　certain　area　inside　a　reactor,　the　simulation　results　show　that　the　algorithm　is　reasonable　and　reliable.　©　The　Author(s),　under　exclusive　license　to　Springer　Nature　Singapore　Pte　Ltd.　2025.