LSTM neural network short-term traffic flow prediction method based on ant colony algorithm optimization

Abstract

The invention discloses an LSTM neural network short-time traffic flow prediction method based on ant colony algorithm optimization. The method comprises the steps: importing historical data from a certain intersection passage database into a Pands module of Python for preprocessing, and carrying out grouping aggregation according to a time period; carrying out normalization processing on the processed data; constructing a training model based on a long-term and short-term neural network, and dividing the training set and the verification set according to the proportion of 80% of the trainingset and 20% of the verification set; and optimizing the neural network of the LSTM based on a heuristic thought through an At-Cycle ant colony algorithm, comparing the neural network with real-time data, evaluating the accuracy of the model by using an MASE and decision coefficient index, and finally, completing the prediction of the short-time traffic flow of the intersection. According to the invention, the ant colony algorithm is used to optimize the setting of hyper-parameters, so that the problem of difficulty in setting hyper-parameters is avoided, the determination coefficient is increased by 8%, and the accuracy of intersection short-term traffic flow prediction can be effectively improved.

Description

technical field [0001] The present invention relates to the traffic flow forecasting method of city crossing, particularly a kind of LSTM neural network short-term traffic flow forecasting method based on an ant colony algorithm optimization. Background technique [0002] Intelligent transportation is a real-time, efficient and accurate comprehensive transportation system that integrates effective information technology, data transmission technology, electronic sensing and control technology, and computer technology. It is a data system worth building for the entire transportation system. [0003] With the further development of information technology, traffic data has gradually evolved into static data, small-scale data and large-scale dynamic data. Static data is data that is relatively stable over a fixed period of time, such as infrastructure, transportation, etc. Dynamic data is data that changes continuously in time and space, such as traffic. The most important thin...

AI Technical Summary

Problems solved by technology

[0005] Purpose of the invention: In view of the fact that traffic flow at intersections is affected by many environments such as weather, holidays, and rush hours, it is difficult to make accurate predictions on traffic flow at intersections. The main difficulty is the combination of short-term traffic and various influencing factors. , resulting in a large deviation between the final forecast data and the actual, the method of the present invention uses the ant colony algorithm to optimize the setting of hyperparameters to avoid the difficulty of setting hyperparameters in the original method, can effectively improve the prediction accuracy, and increase the coefficient of determination by 8 percentage points

Images