Traditional　method　is　difficult　to　be　efficiently　applied　to　series　arc　fault　detection　of　complex　electricity　environment　because　of　the　complex　home　load　working　mechanism　and　its　variety.　For　such　a　problem,　an　effective　method　based　on　multidimensional　feature　fusion　and　principal　component　analysis　in　this　paper.　Firstly,　current　waveforms　of　various　loads　are　analyzed　from　the　perspectives　of　time　domain,　frequency　domain　and　time-　frequency　domain　to　extract　arc　fault　features.　Then,　the　root　mean　square　of　current　time　domain,　harmonic　amplitude　of　frequency　domain,　frequency　domain　energy　of　2k–5 kHz　and　decomposition　energy　entropy　of　wavelet　packet　3　layer　are　selected　as　the　multidimensional　characteristics　for　arc　fault.　And　principal　component　analysis　is　used　for　dimensionality　reduction　for　multidimensional　features.　Finally,　the　multidimensional　features　are　input　to　the　neural　network　to　establish　arc　fault　diagnosis　model　to　realize　the　fault　detection.　The　results　of　the　experiments　demonstrate　that　the　proposed　method　realizes　the　multi-dimensional　fusion　and　arc　fault　diagnosis　model　can　effectively　identify　faults　and　realize　load　identification,　with　an　accuracy　of　99%.　©　Beijing　Paike　Culture　Commu.　Co.,　Ltd.　2025.