[Purposes]The paper explores the use of "bigdata"to innovate the prediction method of urban residents'travel volume(OD), improve the time-consuming,laborious and inaccurate problems of traditional urban residents'travel survey methods, and also provide reliable data support for urban public transportation planning and management. [Methods] Combining the characteristics and advantages of mobile phone signaling data,bus IC card, bus GPS and subway gate data, the OD matrix of urban residents'public transportation travel was obtained by cluster analysis and other methods,and the wavelet neural network combined with optimized whale algorithm (IWOA-WNN)was used to predict the travel OD matrix of future time period. Taking Changsha City as an example, the original data during the 60 d evening peak period were selected, the IWOA-WNN was used for prediction, and the time series method was combined for analysis. [Findings]Compared with the wavelet neural network before optimization, the prediction results of IWOA-WNN are closer to the actual situation, and the accuracy reaches 93.36%. [Conclusions]The methods of data processing and prediction proposed in this study have higher accuracy.