Abstract: Traditional point prediction is difficult to analyze the randomness and uncertainty of wind power

inadequately. Aiming at the shortcomings of point prediction，an interval prediction model based on improved

whale optimization algorithm and fast learning network（IWOA-FLN）was proposed. Firstly，the convergence

speed and accuracy of IWOA was enhanced by adjusting the convergence factor，adding adaptive inertia weight and chaos search strategy. Secondly，a new evaluation index was proposed according to the lower and upper bound estimation method. Finally，the new evaluation index was taken as fitness function，the FLN parameters were optimized by improved whale optimization algorithm to output final prediction interval. Actual examples show that the method could be employed to improve the interval coverage，and reduce the interval bandwidth，which has strong practical significance.

      Key words: whale optimization algorithm（WOA）；fast learning network（FLN）；wind power；interval

prediction