Vibration　signals　from　buildings　are　crucial　for　analysis,　safety　prediction,　and　early　warnings.　However,　acquiring　and　analyzing　these　signals　requires　complex　systems　including　sensor　systems,　storage　devices,　and　computing　equipment.　All　the　part　of　the　system　rely　on　external　power.　This　poses　a　challenge　for　buildings　where　the　installation　of　complex　equipment　and　power　systems　is　inconvenient.　This　study　proposes　a　self-　powered,　high-speed,　and　highly　sensitive　vibration　detection　system.　It　integrates　a　triboelectric　nano-　generator　(TENG)　and　an　organic　field-effect　synaptic　transistor.　A　synaptic　transistor　with　analog　biomimetic　synapse　characteristics　is　proposed.　The　TENG　and　synaptic　transistor＇s　working　principles　and　carrier　transport　characteristics　are　studied.　Using　TENG＇s　output　and　the　synaptic　device＇s　memory,　the　system　detects　and　evaluates　building　vibration　signals.　The　system＇s　adaptability　to　one-dimensional　signals　allows　for　vibration　classification　and　recognition　using　1D-CNN,　achieving　88.9%　accuracy.　This　innovative　strategy　has　broad　prospects　for　solving　vibration　detection　problems　in　special　buildings　and　achieving　lightweight,　real-time,　and　intelligent　monitoring.