In order to alleviate the parking problem in the city, based on the analysis of parking demand scale of the influencing factors, we selected the appropriate demand forecasting models according to the different service objects and mechanism of parking behavior of accessory parking facilities, off-street parking facilities and on-street parking facilities. At the same time, the influence coefficient of road network capacity and the influence coefficient of location conditions were introduced, and the berth turnover rate was used for conversion and correction. A reasonable scale prediction model of district parking based on location conditions was established. Through the division of the macro-level parking strategy in Xining City, we forecasted the scale of parking facilities in the first-class area of Xining City, and compared the prediction results with the results of the traditional parking generation rate method and the results of the parking system planning in Xining City. The results showed that the error rate between the predicted result of the traditional parking generation rate and the planned parking berth was 12.76%, and the error rate of the calculation result of the district parking demand forecasting model and the planned parking berth was 7.6%, which was lower than the traditional parking generation rate method 5%. This method had certain rationality and applicability.

CICTP