Demand-responsive pricing
Open Data and Source Code
PayByPhone
The City of Melbourne has installed in-ground parking bay sensors in most CBD parking bays. These sensors record when a vehicle arrives and when it departs(4600 sensors)
The largest and most accurate parking database in the world.
Over 84,000 locations in over 64 countries.
Location d’un parking privé
Solution de Parking privé
Smart Parking Tags: Bluetooth patches for parking permits
Dynamic Parking Prediction research project
L’intérêt du Machine Learning est qu’une fois que la machine apprend à modéliser à partir des données de stationnement, il est possible de prédire le fonctionnement d’une rue sans horodateur ou même d’une ville entière.
Model Selection:
ARIMA (Autoregressive Integrated moving average)
Linear Regression
Support Vector Regression
Feed Forward Neural Network (90mn training )
Result :Neural Network gives us the best prediction among these models.
Error Reduction by Aggregation/cross-correlation coefficients
Error Table
The key element associate the geographical lots with the aggregated level of prediction error (cross-correlation coefficients)
Clustering
An aggregation of N parking lots (many spots)
In order to find out if prediction errors are independent, we try to explore the daily parking profiles for different regions.
More interesting aggregation phenomena is observed from the daily shape pattern analysis, including Dual Peak, Noon Peak, and Drop before Ramp up.
Profiling such groups will give huge benefits for predictions. It will help us to decompose the prediction error, and targeting different type of travelers.
Prediction of parking space availability in real time
Real-time availability forecast (RAF) algorithm
The three main subroutines of the RAF algorithm are:
The RAF algorithm is validated using a simulation exercise and real case data
The RAF is using combined current (on-line) and historical information.
This algorithm requires a discrete choice model (DCM) !!!
Operates with mixed information (real and simulated)
This algorithm is fed over time with updated records of arrivals and departures at each parking alternative, in order to update parking availability forecast.
http://parknews.biz
Principal Component Analysis