To　estimate　the　contribution　of　main　vibration　and　noise　sources　of　vehicles,　a　source　contribution　estimation　method　based　on　an　improved　independent　component　analysis　(ICA)　algorithm　is　proposed.　A　measure　of　the　similarity　between　the　independent　components　and　the　reference　signals　with　given　characteristics　is　introduced.　A　widely　used　contrast　function,　namely　kurtosis,　is　enhanced　by　the　measure　to　obtain　an　improved　contrast　function.　By　means　of　fixed-point　iteration　and　deflation　approach,　the　improved　contrast　function　is　optimized　and　the　improved　ICA　algorithm　is　attained.　The　contribution　is　computed　by　the　reduced　energy　in　each　extraction,　the　reduction　of　the　energy　corresponds　to　the　contribution　of　the　extracted　independent　components.　The　correspondence　relationship　between　the　independent　components　(ICs)　and　source　signals　can　be　obtained　by　the　signal　characteristics.　The　effectiveness　of　the　proposed　algorithm　is　verified　by　numerical　simulation　and　experiment.