Efficient　capture　and　global　optimization　of　organic　Rankine　cycle　(ORC)　operating　features　in　road　environment　is　the　key　to　obtain　actual　waste　heat　recovery　potential.　The　complexity　of　time-varying　characteristics　intensifies　the　nonlinear　coupling　relationship　between　ORC　performance　and　parameters.　This　paper　first　designs　and　builds　an　ORC　experimental　system　for　recovering　waste　heat　from　internal　combustion　engine　(ICE).　The　experimental　system　includes　ICE　and　its　measurement　subsystem,　ORC　subsystem,　expander　lubrication　subsystem,　cooling　subsystem,　and　ORC　measurement　subsystem.　R245fa　is　used　as　working　fluid;　The　expander　adopts　a　specially　designed　single　screw　expander.　The　performance　of　ORC　waste　heat　recovery　is　tested　by　adjusting　the　operating　conditions　of　the　ICE,　expander,　pump,　and　valve　opening.　The　temperature　range　of　waste　heat　is　627-686　K.　Subsequently,　a　dynamic　feature　capture　and　global　multi-objective　optimization　approach　for　ORC　is　proposed.　The　results　show　that　the　precision　of　dynamic　feature　capture　can　be　improved　at　least　41.78%　with　the　introduction　of　information　theory　in　complex　environment.　Frequent　fluctuation　of　vehicle　speed　is　not　conducive　to　the　improvement　of　evaporator　inlet　temperature　at　working　fluid　side.　As　the　speed　of　the　vehicle　is　greatly　reduced,　the　evaporator　inlet　temperature　drops　sharply.　The　frequent　fluctuation　of　speed　will　intensify　the　nonlinear　correlation　between　operating　parameters　and　system　performance　and　make　the　system　performance　more　dependent　on　the　coupling　correlation　between　operating　parameters.　The　approach　proposed　in　this　paper　can　provide　a　new　idea　for　efficient　capture　of　dynamic　characteristics　and　global　multi-objective　optimization　of　ORC　in　road　environment.