To　solve　the　navigation　problem　for　mobile　robots,　we　present　a　model　based　on　the　operant　conditioning　mechanism　(OCM).　8　elements　consist　of　the　model,　including　state　set,　action　set,　learning　mechanism　and　system　entropy　etc.　As　the　core　of　the　model,　the　learning　mechanism　is　in　accordance　with　operant　conditioning　principles,　which　makes　agents　learn　the　actions　with　reward　and　avoid　the　actions　without　reward.　We　test　the　model＇s　function　in　several　ways　and　change　the　simulation　platform　and　environment　map.　The　results　in　both　experiments　show　that　the　proposed　model　is　effective.　©　2015　IEEE.