An　adaptive　line　enhancer　(ALE)　was　introduced　into　the　secondary　path　identification　part　of　the　adaptive　active　control　system　for　engine　exhaust　noise.　The　modeling　accuracy　of　the　secondary　path　is　greatly　improved,　and　an　additional　transfer　function　of　the　secondary　path　is　introduced.　An　improved　algorithm　was　put　forward,　and　theoretical　analysis,　computer　simulation　and　experimental　set-up　of　this　method　were　implemented.　Comparative　experiments　were　carried　out　using　this　method　against　the　offline　adaptive　modeling　method　and　the　dither　noise　excited　online　modeling　method.　Experimental　results　showed　that　the　online　modeling　technique　with　ALE　and　the　updated　adaptive　algorithm　had　greatly　decreased　the　deformation　of　the　secondary　path　estimation,　improved　the　estimation　accuracy,　accelerated　the　convergence　speed　of　the　control　system,　and　lowered　the　need　of　the　pseudo　signal　power　for　the　secondary　path　modeling.　A　test　on　SUZUKI　AX-100　motorcycle　engine　exhaust　noise　by　active　noise　control　(ANC)　was　carried　out.　The　results　showed　that　the　proposed　ANC　system　could　greatly　eliminate　the　narrowband　components　of　the　noise　source,　and　improve　the　whole　reduction　in　noise.