Eugene OBrien, Roughan & O Donovan and University College Dublin, Ireland. Peter Favai, Cestel, Slovenia (NATMEC Exhibitors)

Transcription

1 Eugene OBrien, Roughan & O Donovan and University College Dublin, Ireland Peter Favai, Cestel, Slovenia (NATMEC Exhibitors)

2 1. What is Bridge Weigh-in-Motion? 2. Accuracy of Bridge WIM 3. BridgeMon project 4. Improving Bridge WIM accuracy Improving weighing algorithms. Improving axle detection. Improving temperature/velocity calibration. Improving data quality assurance. 1

3 » WIM can be grouped into two categories: Pavement WIM Bridge WIM 2

4 » Bridge-WIM uses an existing bridge to weigh trucks.» The idea is to measure the structural response of the bridge while a truck crosses and use an algorithm to backcalculate the axle weights of the truck. 1) Install Sensors Under Bridge 2) Measure response of bridge as truck crosses 3) Use response to calculate weight 3

5 1) Install Sensors Under Bridge» Bridge-WIM uses an existing bridge to weigh trucks.» The idea is to measure the structural response of the bridge while a truck crosses and use an algorithm to backcalculate the axle weights of the truck. 1) Install Sensors Under Bridge 2) Measure response of bridge as truck crosses 3) Use response to calculate weight 4

6 2) Measure response of bridge as truck crosses» Bridge-WIM uses an existing bridge to weigh trucks.» The idea is to measure the structural response of the bridge while a truck crosses and use an algorithm to backcalculate the axle weights of the truck. 1) Install Sensors Under Bridge 2) Measure response of bridge as truck crosses 3) Use response to calculate weight 5

7 3) Use response to calculate axle weights» Bridge-WIM uses an existing bridge to weigh trucks. 3)» The idea is to measure the structural response of the bridge while a truck crosses and use an algorithm to backcalculate the axle weights of the truck. 1) Install Sensors Under Bridge 2) Measure response of bridge as truck crosses 3) Use response to calculate weight 6

8 » Bridge WIM systems are portable Allows temporary monitoring» Everything installed underneath the bridge. No road closure necessary. Can t be seen by truck drivers. You don t need to dig up the pavement! 7

9 8

10 9

11 Bridge WIM system can provide data Pavement WIM system can provide data for pavement design/assessment (MEPP) for economic studies for bridge design/ assessment/load control

12 Least Accurate Most Accurate» Accuracy of WIM systems classified using COST 323 specification. Accuracy is benchmarked against static weighing» There are seven accuracy classes: A(5) means that 95% of gross weights have accuracy of 5% Accuracy Class A(5) B+(7) B(10) C(15) D+(20) D(25) E 11

13 » Varies a lot between sites, etc.» We need independent tests to compare the technologies» But these are hard to find 12

14 » Cold Environment Test in Luleå, Sweden» Major European test of alternative WIM systems

15 » Cold Environment Test in Luleå, Sweden» Major European test of alternative WIM systems

16 E D(25) D(20) C(15) B(10) B+(7) A(5) DuWIM Bridge WIM 1 SiWIM Bridge WIM 2 Two piezoquartz Piezoceramic Bending Plate Bending Beam (prototype)

17 » BridgeMon is an EU funded project from the Research for the Benefit of SMEs scheme.» BridgeMon aims to improve the accuracy of current Bridge WIM technologies. Aims to increase accuracy by at least one accuracy class for a given installation. Ultimate goal is to consistently achieve class A(5) accuracy.» BridgeMon is exploring: Alternative weighing algorithms Alternative axle-detection strategies Data quality assurance Temperature/velocity autocalibration 16

18 » Computer simulations along with testing in the field (using 3 bridges in Slovenia) to assess the accuracy of various Bridge WIM approaches. 1. Box Culvert, Slovenia 17

19 » Computer simulations along with testing in the field (using 3 bridges in Slovenia) to assess the accuracy of various Bridge WIM approaches. 2. Girder Bridge, Slovenia 18

20 » Computer simulations along with testing in the field (using 3 bridges in Slovenia) to assess the accuracy of various Bridge WIM approaches. 3. Concrete Bridge, Slovenia 19

21 » Conventional Bridge WIM assumes axles are static» New approach takes bridge vibrations into account Improves accuracy when the bridge is very flexible 20

22 » Measured response of box culvert bridge was shown to change with temperature.» Weight predictions affected by temperature change.» Calibration curve used to correct the effects of temperature. 21

23 » The measured structural response changes for vehicles travelling at different speeds (dynamic effect)» This changing response can affect the Bridge WIM predicted weights.» Correction factor applied based on speed of vehicle Alternative to a full dynamic approach 22

24 » Bridge WIM uses strategically positioned sensors under the bridge to detect axles Needed for classification and for the Bridge WIM calculation 23

25 » BridgeMon will improve axle detection using: Better axle-detecting locations for some bridge types Advanced signal analysis (Wavelets) to accentuate peaks in the signal 24

26 » Do these axle weights/spacings look reasonable?» Have we seen anything like this vehicle before? Eg: Axle spacings in a 3-axle truck Axle Spacing 2 (m) Axle Spacing 1 (m)

27 Axle Spacing 2 (m)» We use Kernel Densities to fit a distribution» Hence the probability of any configuration Probability of Configuration» A confidence index is assigned to each vehicle based on its probability 2 Suspicious Axle Spacing 1 (m)

28 Gross weight error (%)» Testing the methods outlined in this presentation, and using those which provided the best results has led to a big improvement in Bridge WIM accuracy Mean 0.2% Std 4.4% Mean 2.6% Std 8.1% 27

29 » Bridge WIM requires no installation on the road surface no need to dig up the pavement» Until now, accuracy was similar to other pavement-based WIM technologies» Through this European research project, we have halved the standard deviation of inaccuracies» And have improved Quality Assurance» Hence developing the next generation of Bridge WIM 28

30 Bridgemon project (Project No ) is financed from the Research for the Benefit of SMEs scheme of the FP7 of the European Commission 29