Dynamic multi-interval bus travel time prediction using bus transit data

Chang, Hyun‐Kyung; Park, Dongjoo; Lee, Seungjae; Lee, Ho-Sang; Baek, Seungjae

doi:10.1080/18128600902929591

Cited by 105 publications

(52 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…-nf_ns .] N (3) , p, m , l ' l-l n=l Third, historical travel times are incorporated into the hybrid method. The historical data is retrieved based on certain attributes of the current state such as seasonal, day of the week, time of the day or service characteristics, such as extend of delay or demand levels.…”

Section: Hybrid Schemementioning

confidence: 99%

“…There is an extensive literature on applying various statistical and meta-heuristic methods for bus arrival predictions including regression models [3], artificial neural networks [4], Kalman filter [5], support vector machines [6] and Markov chain [7]. While being very efficient, machine learning techniques do not provide a tractable formulation of the prognosis logic which relates predictions to the underlying service mechanisms.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Real-time bus departure time predictions: Vehicle trajectory and countdown display analysis

Oshyani

Cats

2014

17th International IEEE Conference on Intelligent Transportation Systems (ITSC)

View full text Add to dashboard Cite

Uncertainty is an important challenge in operating bus systems. Accurate real-time predictions can therefore facilitate adaptive decision making process of both operations and passengers. This scheme should be tractable, fast and reliable to be used in real time applications. This paper presents a hybrid prediction scheme to generate real-time information concerning downstream vehicle trajectories and next bus arrival. The prediction generated by the proposed hybrid scheme integrates three travel time components: schedule, instantaneous and historical data. Genetic algorithm is applied in order to specify the contribution of each data source component to the prediction scheme. The benefits, transferability and estimation form of the proposed scheme were tested by applying it on three trunk bus lines in Stockholm, Sweden. Its performance was compared to a commonly deployed scheme. The results indicate that the proposed scheme reduces significantly the overall mean absolute error for all routes from both operators' and passengers' perspectives.

show abstract

Section: Hybrid Schemementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Real-time bus departure time predictions: Vehicle trajectory and countdown display analysis

Oshyani

Cats

2014

17th International IEEE Conference on Intelligent Transportation Systems (ITSC)

View full text Add to dashboard Cite

show abstract

“…Chang et al. () developed a KNN model to estimate bus travel time; the results proved that the model is effective according to the accuracy and computing time of prediction.…”

Section: Introductionmentioning

confidence: 99%

Prediction of Bus Travel Time Using Random Forests Based on Near Neighbors

Wang

Shan

et al. 2017

Computer aided Civil Eng

133

View full text Add to dashboard Cite

The prediction of bus arrival time is important for passengers who want to determine their departure time and reduce anxiety at bus stops that lack timetables. The random forests based on the near neighbor (RFNN) method is proposed in this article to predict bus travel time, which has been calibrated and validated with real‐world data. A case study with two bus routes is conducted, and the proposed RFNN is compared with four methods: linear regression (LR), k‐nearest neighbors (KNN), support vector machine (SVM), and classic random forest (RF). The results indicate that the proposed model achieves high accuracy. That is, one bus route has the results of 13.65 mean absolute error (MAE), 6.90% mean absolute percentage error (MAPE), 26.37 root mean squared error (RMSE) and 13.77 (MAE), 7.58% (MAPE), 29.01 (RMSE), respectively. RFNN has a longer computation time of 44,301 seconds for a data set with 14,182 data. The proposed method can be optimized by the technology of parallel computing and can be applied to real‐time prediction.

show abstract

“…Other methods including NPR and LR are not so popular, but they are quite simple in calibration and calculation (Park et al 2007;Chang et al 2010;Maiti et at. 2014).…”

Section: Literaturementioning

confidence: 99%

Bus arrival time prediction using mixed multi-route arrival time data at previous stop

Hua

Wang

et al. 2017

Transport

View full text Add to dashboard Cite

Abstract. The primary objective of this paper is to develop models to predict bus arrival time at a target stop using actual multi-route bus arrival time data from previous stop as inputs. In order to mix and fully utilize the multiple routes bus arrival time data, the weighted average travel time and three Forgetting Factor Functions (FFFs) -F1, F2 and F3 -are introduced. Based on different combinations of input variables, five prediction models are proposed. Three widely used algorithms, i.e. Support Vector Machine (SVM), Artificial Neutral Network (ANN) and Linear Regression (LR), are tested to find the best for arrival time prediction. Bus location data of 11 road segments from Yichun (China), covering 12 bus stops and 16 routes, are collected to evaluate the performance of the proposed approaches. The results show that the newly introduced parameters, the weighted average travel time, can significantly improve the prediction accuracy: the prediction errors reduce by around 20%. The algorithm comparison demonstrates that the SVM and ANN outperform the LR. The FFFs can also affect the performance errors: F1 is more suitable for ANN algorithm, while F3 is better for SVM and LR algorithms. Besides, the virtual road concept in this paper can slightly improve the prediction accuracy and halve the time cost of predicted arrival time calculation.

show abstract

Dynamic multi-interval bus travel time prediction using bus transit data

Cited by 105 publications

References 17 publications

Real-time bus departure time predictions: Vehicle trajectory and countdown display analysis

Real-time bus departure time predictions: Vehicle trajectory and countdown display analysis

Prediction of Bus Travel Time Using Random Forests Based on Near Neighbors

Bus arrival time prediction using mixed multi-route arrival time data at previous stop

Contact Info

Product

Resources

About