The　fault　diagnosis　of　the　multilevel　reciprocating　compressor　is　very　complex.　Based　on　the　point　view　of　complementary　strategies,　a　fault　diagnosis　method　which　integrates　the　ART　and　the　artificial　immune　network　is　put　forward.　It　can　be　used　to　enrich　the　data.　Then　the　mapping　of　the　causation　between　reasons　and　results　can　be　created,　so　the　diagnosing　is　accomplished.　The　step　of　the　method　is　discussed.　Additionally,　a　criterion　about　the　data　enriching　is　put　forward.　The　correctness　and　the　effectiveness　of　the　method　are　proved　by　experiments.