Class-8 Heavy Truck Duty Cycle Project Final Report - Center for ...

More documents

Recommendations

Info

Given the size and complexity of the data collected, a significant effort was devoted to the task of preparing the raw data for the analysis. As explained elsewhere in this report, the information was collected using a data acquisition system that saved the information in its proprietary format (i.e., SIF files or SoMat Information Files; SoMat was the manufacturer of the equipment used in the data collection task). These binary files are saved using a data structure format that optimizes the size and accessibility of the information. Unfortunately, however, it is a format that cannot be read unless the associated data structure is known. Because the data structure is SoMat proprietary information, the company provided ORNL with a dynamic-link library, or dll (i.e., software in the form of a shared library of functions) that could be incorporated into the special software developed by ORNL to cleanse and organize the information. This dll was used to translate the data from the proprietary format (SIF) to a flat text (i.e., ASCII, or American Standard Code for Information Interchange) format that is readable by any user. The software was tested and debugged during the pre-FOT timeframe with data collected by one of the test trucks (Truck 4) during preliminary testing as an instrumented truck. The preliminary testing data was also used to: 1) check that all the equipment was working as expected, 2) that the data gathered was in the right format, and 3) map Truck 4 during this preliminary effort to better understand the geographical area that would be covered by the project (see Fig. 54). LEGEND Week 1 Week 2 Week 3 Week 4 Fig. 54. Truck 4 Trips, by Week for a Month of Preliminary Testing Once the FOT was initiated, the protocol that was used to save the information collected by the instrumented truck was as follows. Every one-to-two weeks (usually during a weekend when the instrumented trucks returned to the Schrader garage/parking lot), the information collected and stored on-board was retrieved and saved to a laptop computer. This created a SIF file with a size of about 0.5 GB per truck per week of data gathering3. The week-long SIF file was then translated into an easily readable format as explained above, and the information was parsed into the 60 channels that were collected and saved to files that stored one-day’s worth of data collection data. These files were named with the calendar date in which the data was collected (e.g., 20061113, for information 3 The size of the file was directly correlated to the time that the truck was powered (i.e., the ignition key was in the “on” position), whether it was moving or stationary. 60
collected on November 13 th , 2006), plus a one-digit identifier that identified the truck (1 to 6) that gathered the information. The apparently easy-sounding task of parsing the information into calendar date files required substantial manipulation of the raw data. Each record (i.e., each reading of the 60 channels, which occurred once every 0.2 seconds) was time stamped by the DAS using Universal Time Coordinates (UTC or Greenwich Time). This time information had to be adjusted based on the geographical location of the truck (e.g., a truck traveling within the Eastern Time Zone would have a time stamp of UTC - 4 hrs or UTC - 5 hrs, depending on the time of the year and the year itself since the daylight saving time start and end dates are year dependent4). This, in turn, required determining when a given truck crossed the border between one time zone and the next one. When that line was crossed, the local timestamp has to be immediately reduced by one hour for westbound trucks and increased by one hour for eastbound trucks. As a result, instead of being 24 hours long, some of the calendar date files could be up to 25 hrs long (i.e., they could have up to two “eleven pm-to-midnight” hours) for westbound trips and some could be 23 hrs long (even if the trips started at midnight) for eastbound trucks that crossed time zones. Obtaining the correct local time stamps was important for later analyses involving peak and off-peak hour periods. Simultaneous with the time stamp corrections, the software developed by ORNL also checked the validity of the data provided by the 60 channels. There were instances in which the on-board sensors did not provide information due to, for example, a malfunction of the sensor (sensor invalid data), and other cases where the DAS had problems storing the data (eDAQ invalid data). Each channel had a numeric code that identified these types of problems, and those codes were stored in the file for that particular channel and record. The data cleansing software read this information and generated statistics that allowed the ORNL researchers to determine when a certain sensor was systematically failing or when the failures were just “glitches” that occurred sporadically. In the first case, troubleshooting, repair/replacement of the particular sensor was accomplished the next time the team was in contact with the particular truck with the problem. Other errors such as out-of-range errors and errors attributed to the GPS system were also identified and indications of their occurrence were added to the summary reports. Any record that had any kind of error for any channel was marked with a code that allowed for the identification of those errors, and this information was added to the corresponding calendar date file as a new field. Table 11 below shows a sample of the general statistics generated for the first nine channels for Truck 1 during the week of October 21 to 28 (only October 21st and 23rd are shown). It should be noted that statistics were generated for all channels but because of space constraints only the first nine channels are shown in Table 11. For each data collection date, the table shows the statistics (count, minimum observed value, maximum observed value, etc.) for each channel (column header), followed by a detailed accounting of the different errors observed, going from errors reported by the data acquisition system, sensor errors, out of range errors, and other errors (e.g., errors attributed to the GPS system). Consider, for example, the statistics for October 23, 2006. The first channel, AcPdlPos (Acceleration Pedal Position), generated 73,132 readings without errors which ranged from 0-to- 100, with a mean of 19.28 and a standard deviation of 25.78 (no integration5 was computed for this channel). There were also 14,136 instances in which the sensor provided invalid data, but there were no eDAQ invalid 4 See http://nationalatlas.gov/articles/boundaries/a_savingtime.html for daylight saving time start and end dates. 5 For some channels such as speed or instantaneous fuel rate, the software integrated the information across time to generate, for example, the distance traveled and total fuel consumed. 61
Page 1 and 2:
Class-8 Heavy Truck Duty Cycle Proj
Page 3:
Vehicle Systems Program ORNL/TM-200
Page 6 and 7:
2.7 LAUNCH OF THE FOT (OCTOBER 23,
Page 8 and 9:
40 Connection and Support Structure
Page 11:
LIST OF TABLES Table Page 1 Pilot T
Page 15:
ACKNOWLEDGEMENTS The Oak Ridge Nati
Page 18 and 19:
Sixty channels of data were collect
Page 20 and 21:
xviii
Page 22 and 23:
The data analyses performed in Phas
Page 24 and 25:
Fig. 2. HHDDT Transient Mode In add
Page 26 and 27:
vehicle; 3) a transit bus; 4) a nei
Page 28 and 29:
payload and accessory load would al
Page 30 and 31: Data # Performance Measure Benefit
Page 32 and 33: Data # Performance Measure Benefit
Page 34 and 35: Fig. 10. NGK NOx/O2 Sensors Fig. 11
Page 36 and 37: 1.3.4 De-instrumentation: Following
Page 38 and 39: Fig. 16. Pilot Test North-South Rou
Page 40 and 41: GPS information collected with the
Page 42 and 43: 2.1 REVIEW OF PERFORMANCE MEASURE R
Page 44 and 45: 2.3.2 Performance Measures Fig. 18.
Page 46 and 47: Signal Sensor 53 Road Grade Calcula
Page 48 and 49: 2.3.6 Tractor and Trailer Details F
Page 50 and 51: determined that the particular CANo
Page 52 and 53: Cable runs to and from sensors were
Page 54 and 55: 2.8 CONDUCTION OF FOT The FOT laste
Page 57 and 58: 3. INTERNET DATA ACCESS TOOL The HT
Page 59: Fig. 30. Search Results The file se
Page 62 and 63: When the button “Extract Data”
Page 64 and 65: Data Fig. 33. Segment Merging In ad
Page 66 and 67: Fig. 36 shows the synthetic cycle g
Page 68 and 69: Fig. 39. A Synthetic Duty Cycle wit
Page 70 and 71: tractors) as well as the maximum to
Page 72 and 73: vehicle was carrying a heavy load o
Page 74 and 75: Vehicle Speed [mph] Vehicle Speed [
Page 76 and 77: 5.2 REAL-WORLD DUTY CYCLES Informat
Page 78 and 79: Vehicle Speed [mph] 70 60 50 40 30
Page 82 and 83: data or out-of-range data. Therefor
Page 84 and 85: 6.1.1 Spatial Information Using the
Page 86 and 87: For those cases in which an instrum
Page 88 and 89: Fig. 56. DCGenT Prototype Search Cr
Page 90 and 91: at constant speed. Parameters m, b,
Page 92 and 93: weight. Notice that, as expected, m
Page 94 and 95: the loss of information, and conseq
Page 96 and 97: consumption per unit of distance tr
Page 98 and 99: q q ( − ) ( feb'− feb) 100 '
Page 100 and 101: Table 15. General Statistics: Total
Page 102 and 103: 70% 60% 50% 40% 30% 20% 10% 0% Idli
Page 104 and 105: Fig. 68. All Trips with Dual Tires
Page 106 and 107: The almost 700,000 miles logged dur
Page 108 and 109: Frequency 300 250 200 150 100 50 0
Page 110 and 111: Speed [mph] 80 70 60 50 40 30 20 10
Page 112 and 113: vehicle is hauling light or heavy l
Page 114 and 115: Frequency 350 300 250 200 150 100 5
Page 116 and 117: A summary of the statistics describ
Page 118 and 119: analysis controlling for the type o
Page 120 and 121: 6.3.2.4 Summary of Fuel Efficiency
Page 122 and 123: Lesson Learned: A technology progre
Page 124 and 125: Lesson Learned: Adding external sen
Page 126 and 127: know if the vehicle was “filled u
Page 129 and 130: 8. SUMMARY OF RESULTS AND CONCLUSIO
Page 131 and 132:
9. FUTURE DIRECTIONS The Heavy Truc
Page 133:
Appendix A PILOT TEST LETTER REPORT
Page 136 and 137:
1.0 BACKGROUND 2.0 PROJECT OVERVIEW
Page 138 and 139:
ACKNOWLEDGEMENTS The authors wish t
Page 140 and 141:
evaluate the data collection system
Page 142 and 143:
performance of their next generatio
Page 144 and 145:
March and provided to the ORNL Prog
Page 146 and 147:
Between January, 2005 and May, 2005
Page 148 and 149:
Figure 4: NGK Sensor Figure 5: Omeg
Page 150 and 151:
o Date and Time o Wind Speed o Wind
Page 152 and 153:
Figures 28 through 31 show the ball
Page 154 and 155:
Figure 22: Airweigh System Figure 2
Page 156 and 157:
DVD. Figures 35 and 36 show some of
Page 158 and 159:
3.3.7.1 Pre-Analysis Efforts: As th
Page 160 and 161:
Task 1: Identification of Fleet(s)
Page 162 and 163:
Data # Performance Measure Performa
Page 164 and 165:
Sensor or eDaq Performance Performa
Page 166 and 167:
45 46 47 48 49 50 51 52 53 Measure
Page 168 and 169:
Data # 62 63 64 65-67 68-69 70-71 7
Page 170 and 171:
Data # Performance Measure 87 Headi
Page 172 and 173:
Appendix A-B Final Pilot Test Plan
Page 174 and 175:
• DAS/sensor suite/cabling operab
Page 176 and 177:
9 10 11 12 13 14 15 Lateral Velocit
Page 178 and 179:
34 Rain Intensity 35 36 37 38 39 40
Page 180 and 181:
59 60 61 62 63 64 65-67 68-69 70-71
Page 182 and 183:
85 Longitude 86 Velocity 87 Heading
Page 184 and 185:
3.2 Sensors The PMs listed in Table
Page 186 and 187:
vehicles will be dedicated to the p
Page 188 and 189:
DATE TEST ENGINEER: DRIVER: DEPARTI
Page 190 and 191:
Jim Ridge, Dana Product Engineer, w
Page 192 and 193:
Table 5.1 October Route and Schedul
Page 194 and 195:
Table 5.2 December Route and Schedu
Page 196 and 197:
Appendix A-C Field Test Signals 64
Page 198 and 199:
9 Performance Measure Performance M
Page 200 and 201:
Performance Measure 25 Rain Duratio
Page 202 and 203:
Letter Report SUMMARY OF RELEVANT H
Page 204 and 205:
Sensor or Performance Data Sensor o
Page 206 and 207:
GAWR 44 lb AirWeigh 5800 Road Surfa
Page 211:
Appendix B SCHRADER MEMORANDUM OF U
Page 214 and 215:
MOU-UTB-2007001 Schrader Trucking C
Page 216 and 217:
MOU-UTB-2007001 Schrader Trucking C
Page 219:
Appendix C FIELD OPERATIONAL TEST P
Page 222 and 223:
Collection, Analysis, and Archiving
Page 224 and 225:
4.0 APPROACH Phases 2 and 3 of the
Page 226 and 227:
Signal Sensor 60 Tractor-Trailer Ma
Page 228 and 229:
Schrader Trucking will provide six
Page 230 and 231:
Sensitive Information] [Partner Sen
Page 232 and 233:
4.6 Sensor and DAS Integration to T
Page 234 and 235:
5.5 Vehicle Deployment Methodology
Page 236 and 237:
Start address: 2360 Cherahala Blvd
Page 238 and 239:
Figure 6.2: Schrader Shop, Far Left
Page 240 and 241:
Figure 6.7: eDAQ Lite with two Vehi
Page 242 and 243:
6.5 Safety Information For Student
Page 245:
Appendix D SYSTEM DESIGN AND OPERAT
Page 248 and 249:
TABLE OF CONTENTS Collection, Analy
Page 250 and 251:
9.1.2. Shutdown Procedure 9.1.3. SI
Page 252 and 253:
Signal Sensor 26 Cruise Control Acc
Page 254 and 255:
Figure 4.1: Air-Weigh System Compon
Page 256 and 257:
4.5. EDAQ The eDAQ Lite is connecte
Page 258 and 259:
the equipment inside from overheati
Page 260 and 261:
5.4. Weather Station Connections Fi
Page 262 and 263:
Figure 5.5.3: VBOX Cable Table 5.5.
Page 264 and 265:
Figure 6.3: Weather Station Locatio
Page 266 and 267:
6.6 Power Supply Use a second cable
Page 268 and 269:
can also be changed from this windo
Page 270 and 271:
on and the yellow light should begi
Page 272 and 273:
Figure 7.1.3-1: Weather Station and
Page 274 and 275:
Speed unit: m/s Sampling frequency:
Page 276 and 277:
iii) Click on Browse to find and op
Page 278 and 279:
c) Power Controller (Power) - no co
Page 280 and 281:
1) Select the Transducer and Messag
Page 282 and 283:
Figure 7.2.4-2: Adding Transducer C
Page 284 and 285:
Figure 7.2.4-3: Editing DIO Vehicle
Page 286 and 287:
A. Changing Units - Converting tran
Page 288 and 289:
C. Coordinates - Assign correct sig
Page 290 and 291:
switch) Table 7.2.5: Contents of Ti
Page 292 and 293:
Trigger Delay Setup: Check Enable d
Page 294 and 295:
Figure 7.2.5-6: Time Base Shifter S
Page 296 and 297:
Units: Full Scale Estimate - Min: 0
Page 298 and 299:
Figure 7.2.6-2: Burst History Data
Page 300 and 301:
Press Ctrl+3 (or click Run in the T
Page 302 and 303:
8.2. Procedures for Heavy Truck Dut
Page 304 and 305:
d) Turn on the ignition. e) Turn on
Page 306 and 307:
4) Change the IP address field to t
Page 308 and 309:
5) Check to see that the computer c
Page 310 and 311:
e) When all SIC files have been sav
Page 312 and 313:
4) To verify the time setting, go t
Page 314 and 315:
e) Alter the id, mask, value, etc.
Page 316 and 317:
Longitudinal CAN 0x87FF 0x0304 Long
Page 318 and 319:
ULSB 0x98EAFFFBC1FE00 MaxVehSpLm CA
Page 320 and 321:
Figure 9.1.6-2: Performing a ColdSt
Page 322 and 323:
Table 9.2.2-1: Descriptions and Uni
Page 324 and 325:
Channel ID Sensor Format Integer? P
Page 326 and 327:
Channel ID eDAQ Invalid Data Value
Page 328 and 329:
Channel ID Resolution Accuracy Note
Page 330 and 331:
9.3. Appendix C: Additional DAS Not
Page 332 and 333:
e replicated and saved as five iden
Page 334 and 335:
2) Type in “cmd” 3) Type “ipc
Page 336 and 337:
Producing multiple setup files 1) G
Page 338 and 339:
9.4. Appendix D: Checklists and For
Page 340 and 341:
D-96
Page 342 and 343:
9.5. Appendix E: References and Web
Page 345 and 346:
E-3
Page 347 and 348:
E-5
Page 349 and 350:
E-7
Page 351 and 352:
E-9
Page 353 and 354:
E-11
Page 355 and 356:
E-13
Page 357 and 358:
E-15
Page 359 and 360:
E-17
Page 361 and 362:
E-19
Page 363 and 364:
E-21
Page 365 and 366:
E-23
Page 367 and 368:
E-25
Page 369 and 370:
E-27
Page 371 and 372:
E-29
Page 373 and 374:
E-31
Page 375 and 376:
E-33
Page 377 and 378:
E-35
Page 379 and 380:
E-37
Page 381 and 382:
E-39
Page 383:
E-41
show all

Class-8 Heavy Truck Duty Cycle Project Final Report - Center for ...

Create successful ePaper yourself

Delete template?

Save as template?