A　new　wind　power　modeling　method　is　proposed　for　multiple　wind　farms　considering　fluctuation　tendency　and　correlation　characteristics　for　planning　and　operation　of　power　system.　These　wind　power　time　series　are　divided　into　several　wind　processes,　which　are　defined　as　local　fluctuation　processes　and　are　quantitatively　expressed.　Classifying　the　months　with　similar　characteristics,　the　correlations　of　these　fitting　parameters　are　analyzed,　and　the　Copula　function　is　established.　Combining　with　Copula　function,　the　fluctuation　parameters　are　sampled　sequentially,　hence　the　constructed　similar　time　series　have　remarkable　correlation　characteristics.　This　method　focuses　on　the　local　correlation　characteristics　with　fluctuation　feature　and　avoids　concerning　only　the　entire　correlation.　A　simulation　for　some　wind　farms　in　China　is　performed　and　comparisons　of　correlation　between　two　wind　farms　and　between　historical　and　simulated　time　series　show　that　the　proposed　method　is　effective　and　the　local　correlation　is　well　considered.　©　2018,　Editorial　Office　of　Journal　of　Xi＇an　Jiaotong　University.　All　right　reserved.