Based　on　the　wavelet　decomposition　and　conditions　of　the　support　vector　kernel　function,　a　nonlinear　wavelet　basis　is　introduced　to　construct　the　kernel　function　of　support　vector　machine　(SVM),　and　a　compact　wavelet　support　vector　machine　(WSVM)　with　a　strong　generalization　ability　is　obtained.　Wavelet　packet　decomposition　is　applied　to　the　structural　response　signals　under　ambient　vibrations,　feature　vectors　are　obtained　by　a　feature　extraction　method　with　＇energy-damage　state　＇.　The　feature　vectors　are　used　for　training　and　classification　as　the　inputs　of　the　compact　support　vector　machine.　Hence,　a　new　structural　damage　detection　method　called　＇　complete　wavelet　support　vector　machine　＇　is　established.　This　method　is　used　to　a　single-layer　spherical　lattice　shell　for　damage　diagnosis.　The　structural　damage　position　and　degree　can　be　identified　and　classified,　and　the　result　is　highly　accurate.　This　approach　has　some　advantages,　such　as　engineering　oriented,　low　cost　and　convenient.