To　solve　the　problem　of　the　complexity　and　danger　of　natural　disaster　rescue,　a　rescue　robotic　arm　with　high　load　to　weight　ratio　was　proposed,　and　a　three-dimensional　model　was　established.　The　forward　and　inverse　kinematics　models　of　the　multi-degree-of-freedom　mechanical　arm　were　established　through　the　Denavit-Hartenberg　method,　and　the　simulation　model　of　the　rescue　robotic　arm　was　established　by　using　the　MATLAB　robot　toolbox　to　verify　the　correctness　of　the　forward　and　inverse　kinematics　models.　A　10-input,　5-output,　3-layer　BP　neural　network　was　used　to　predict　the　inverse　kinematics,　and　the　4　group　kinematics　inverse　solution　was　screened　to　obtain　the　only　accurate　inverse　solution,　which　lays　a　foundation　for　the　realization　of　the　subsequent　trajectory　planning　of　the　robotic　arm.　©　2022,　Editorial　Department　of　Journal　of　Beijing　University　of　Technology.　All　right　reserved.