Prediction Method of Parking Space Based on Genetic Algorithm and RNN
| Autor: | Xuwang Lu, Jilun Qiu, Haipeng Chen, Jianrong Tian |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
050210 logistics & transportation
education.field_of_study Fitness function Mean squared error Parking guidance and information Artificial neural network Computer science 05 social sciences Population 02 engineering and technology computer.software_genre Set (abstract data type) Recurrent neural network 0502 economics and business Genetic algorithm 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Data mining education computer |
| Zdroj: | Advances in Multimedia Information Processing – PCM 2018 ISBN: 9783030007751 PCM (1) |
| DOI: | 10.1007/978-3-030-00776-8_79 |
| Popis: | With respect to the prediction of short-term unoccupied parking space of parking guidance and information system (PGIS),a prediction method using genetic algorithm combined with recurrent neural network (RNN) is proposed. First, set the parameters of the RNN population, including the search space of the neural network’s hidden layers, neuron number, and neuron type. Then by setting the parameters of the genetic algorithm to drive and control the RNN training process, and using the RMSE value of the prediction result as the fitness function of the genetic algorithm to perform the individual evaluation index of the RNN. Finally, the RMSE values of the predicted results of all RNN individuals on the experimental dataset are compared through two different scenarios of prediction examples to obtain the best prediction model. The results of experiments show that this method has excellent prediction accuracy and wide applicability for the prediction of short-term and parking spaces in parking guidance information systems. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|