GPS/GIS Based Electronic Road Pricing System Design


2.2 In-vehicle unit
The proposed in-vehicle unit in next ERP system is shown as in figure 2. Integrated GPS/Dead Reckoning (DR) positioning provide continuous vehicular position output even in urban area where dense high buildings or the tunnel will block the GPS signal. To further improve the positioning accuracy, Map Matching (MM) could be applied in the in-vehicle unit if there is map database and enough processing power or implemented in the management center. The accurate positioning information could be used to assist the traffic flow analysis on the exact road. The Data fusion is mainly for fusing all the sensors’ data to get more accurate, robust positioning information. Status provides the vehicular or road status such as alarm, breaking down or congestion. Display shows some related information to the driver including guidance, charging information or electronic map if there has. Transceiver provides two-way data link via wireless network and IP based packet data transmission. The data would be broadcast into the Internet so that any authorized user can make full use of them to provide value-added services. Control card is the crucial part for organizing all these information. Smart card interface is for charging via smart card inserted into the in-vehicle unit.


Fig. 2 Design of the in-vehicle unit for ERP


2.3 Communication link
Wireless communication is used to transmit all the vehicles position and status to management center, and broadcast the updates of the traffic information when required or periodically.

For Singapore next ERP system, current cellular mobile systems or third generation wireless communication could be employed and the mobile set could be embedded in the IU. As 2.5 Generation wireless communication, General Packet Radio Service (GPRS) has quicker session setup, permanent connection, lower cost, higher data rate performance[5]. Furthermore, it is IP based data transmission so the transceiver is unnecessary in the management center, which is connected with Internet. It is better choice for the next ERP system and can be easily upgraded into the next generation wireless communication.

3 System Modules
Based on the proposed architecture of in-vehicle unit and management center, the main modules include electronic map database, integrated positioning, and map matching.

3.1 Electronic map database
In next ERP system in Singapore, electronic road map could be used for assist positioning, charging scheme and route guidance. The contents and data structure in the electronic map database are described as follows.

1) Spatial database: This includes node, which may be the junction and cross point of road, link representing road segment. The points between two adjacent nodes are also included in the spatial database.

2) Non-spatial database: This includes address name, road name and attribute etc.

3) Topological relationship: It includes relationship between node and node, node and link, among links.

3.2 Positioning of vehicle
In next ERP system, widespread installation of expensive gantries is not necessary because the position of vehicle will be used for charging. There are many positioning systems are available for the civil application; among them GPS could provide 10 meters accuracy without SA and can work in all weather, at any time of the day, and under specified conditions of radio-frequency interference, signal availability. However, in an urban environment, satellite signals are blocked by high buildings or heavy foliage, and the urban canyon limits satellite visibility. To solve these problems, many aided positioning methods such as dead reckoning (DR), wireless communication were adopted [3]. In comparison to DR, wireless aided method is still dependant on the radio environment. DR uses the distance and direction changes to provide the vehicular position relative to the original point. As DR has error accumulated over time and reliable GPS solution are often not available in urban area, multi-sensor integration systems, therefore, are required for next ERP systems, which is shown in the diagram of figure 1. The GPS receiver is for the absolute positioning and set up the original point for DR, while the distance and direction sensor provide the distance and direction information for DR algorithm. When the GPS data is available, the fusion algorithm predicts the errors of distance and direction sensor; the DR algorithm will provide the positioning while GPS data is not available.

The traffic flow statistics will be dependant on the road where vehicle travels while its position is transmitted to management center. However, sometimes the GPS and/or DR cannot give an accurate position, which cannot determine which road the vehicle travels on. Therefore map matching is needed to integrate with GPS/DR in next ERP system. Map matching is simply a method of using stored electronic map data information about a region to improve the ability of a position determination system to handle errors[4]. Its algorithm has basically two categories: as a search problem or as a statistic estimation problem. As a search problem, map matching is to match the measured vehicle position to the closest node or point within a predefined range in the road network. This approach is fast and fairly easy to implement and it is suitable for the vehicles that have fixed track such as buses or travel in accordance with planning route. As a statistics estimation problem, map matching use probabilistic approaches to make the match more accurate through modeling errors in an ellipse or rectangle of uncertainty that represents the area in which the vehicle is likely to be located. It is suitable for any vehicle that does not have fixed route in traveling.

3.3 Map matching
Map matching algorithm in CCS processes the received vehicle position and searches the map database to create the list of road links the vehicle traveled. The required payment is calculated according to the road link price in database and deduction is made from prepaid card or stored to the vehicle ID account for monthly-based payment.

Different from vehicle navigation system[6], map matching for ERP does not require real time calculation. This means that the searching of road link vehicle traveling on current can use not only the past and present vehicle position but also the following vehicle position. The off time matching makes it much more accurate and reliable. The matching processes (as shown in figure 3) are: initialization (O), position between intersections (M-N), position near intersection and no turn detected (M), position near intersection and turn detected (P), and re-initialization.


Fig. 3. Map matching process


Page 2 of 3
| Previous | Next |