Enabling Smart Urban Services with GPS Trajectory Data

Preface......Page 5
Acknowledgments......Page 10
Contents......Page 11
Part I: Foundations......Page 18
1.1 Introduction......Page 19
1.2 Related Work......Page 20
1.3 Main Concepts......Page 21
1.4.2 Identifying Top-k Candidate Edges for a Given GPS Point......Page 23
1.4.2.1 Spatial Probability......Page 24
1.4.3 Finding Paths Between Two Consecutive GPS Points......Page 25
1.4.4 Refining Paths for a Given Raw Trajectory Segment......Page 28
1.5.1 Data Preparation......Page 30
1.5.3 Baseline Methods......Page 31
1.5.4.2 Varying k......Page 32
1.5.5 Efficiency Study......Page 33
1.5.5.2 Varying k......Page 34
1.5.5.3 Time Cost at Different Stages......Page 37
1.6 Conclusions and Future Work......Page 38
References......Page 39
2.1 Introduction......Page 41
2.2.1 Line-Simplification-Based Compression......Page 42
2.2.2 Map-Matching-Based Compression......Page 43
2.3.1 Basic Concepts......Page 44
2.3.2 System Overview......Page 45
2.4.1 Motivation......Page 46
2.4.2 Algorithm Details......Page 47
2.4.3 Trajectory Decompression......Page 48
2.5.1 Baselines......Page 49
2.5.3 Varying l......Page 50
2.5.5 Quality of Response for When-and-Where Query......Page 52
2.6.1 Experimental Setup......Page 54
2.6.2 Effectiveness Study......Page 56
2.6.3.2 Energy Consumption......Page 58
2.6.3.3 Memory Consumption......Page 59
2.7 Conclusions and Future Work......Page 60
References......Page 61
3.1 Introduction......Page 63
3.2 Related Work......Page 64
3.3 Preliminary......Page 65
3.4 Trajectory Protection Mechanism......Page 69
3.5 Evaluations......Page 75
3.6 Conclusions and Future Work......Page 82
References......Page 83
Part II: Enabling Smart Urban Services: Drivers......Page 85
4.1 Introduction......Page 86
4.2 Related Work......Page 88
4.3 Empirical Study......Page 89
4.3.1.1 Shift Handover Event Detection......Page 91
4.3.1.2 Statistical Study of Taxi Shift Handover Location and Time......Page 92
4.3.2.1 Taxi Performance Quantification......Page 94
4.3.2.2 Number of Passenger Delivery Trips......Page 96
4.3.2.3 Popular Passenger Pickup and Drop-Off Areas......Page 97
4.4.1 Taxi Service Strategy Extraction......Page 98
4.4.1.1 Passenger-Searching Strategies......Page 99
4.4.1.3 Service-Area Preference......Page 100
4.5.1 Discovering Good Taxi Service Strategies......Page 101
4.5.1.1 Passenger-Searching Strategies......Page 102
4.5.1.2 Passenger-Delivery Strategies......Page 104
4.5.1.3 Service-Area Preference......Page 105
4.5.2 Performance Prediction Based on Historical Strategies......Page 107
4.6 Conclusions and Future Work......Page 108
References......Page 109
5.1 Introduction......Page 110
5.2.1 Fuel Consumption Modelling......Page 112
5.2.2 Solutions to Saving Fuel Consumption......Page 113
5.3.1 Basic Concepts......Page 114
5.3.2 Problem Statement......Page 115
5.4 Personal Fuel Consumption Model Building......Page 116
5.4.1 Necessary Fuel Consumption Rate......Page 117
5.4.2 Extra Fuel Consumption Rate......Page 118
5.4.3 Personalized Fuel Consumption Model (PFCM)......Page 119
5.5.1 Real-Time Traffic Information Collection......Page 121
5.5.3 Fuel-Efficient Path-Planning......Page 122
5.6.1.2 Baseline Models......Page 124
5.6.2.1 Comparison Results to Baselines......Page 125
5.6.2.2 Model Performance on Paths Under Varied Lengths......Page 126
5.6.3 A Deep Look at PFCM Coefficients......Page 127
5.6.4 Deriving a General PFCM......Page 129
5.6.5.1 Effectiveness of Real-Time Traffic Information Collection Method......Page 130
5.6.5.2 Determination of Length of Sub-route......Page 131
5.6.5.3 Potential Savings on Fuel Consumption......Page 133
References......Page 135
Part III: Enabling Smart Urban Services: Passengers......Page 137
6.1 Introduction......Page 138
6.2 Related Work......Page 140
6.3 Preliminaries and Problem Definition......Page 141
6.4 Isolation-Based Online Anomalous Trajectory Detection......Page 144
6.4.1 Offline Preprocessing......Page 145
6.4.2.1 Basic Idea......Page 146
6.4.2.2 Anomaly Score......Page 148
6.5.2 Evaluation Criteria......Page 150
6.5.3 Results......Page 151
6.5.4 Adaptive Versus Fixed Window Approach......Page 153
6.5.5 iBOAT Versus iBAT......Page 155
6.6 Conclusions and Future Work......Page 157
References......Page 158
7.1 Introduction......Page 160
7.3.1 Basic Concepts......Page 162
7.3.2 Problem Statement......Page 163
7.4.1 Urban Activity Region Identification......Page 164
7.4.2 Candidate Activity Area Identification......Page 165
7.4.3 Trip Purpose Imputation......Page 168
7.5 Phase II: Enabling Real-Time Response......Page 172
7.6.1 Evaluation on Candidate Activity Area Identification......Page 174
7.6.2 Evaluation on Trip Purpose Imputation Algorithm......Page 175
7.6.3 Evaluation on Response Time......Page 178
7.7 Conclusions and Future Work......Page 180
References......Page 181
Part IV: Enabling Smart Urban Services: Urban Planners......Page 183
8.1 Introduction......Page 184
8.2.2 Measuring GPS Positioning Performance......Page 186
8.3 Basic Concepts......Page 187
8.4.1 Overview of the Framework......Page 190
8.4.2 Map Matching-Based GPS Error Matrix Construction......Page 191
8.4.3 GPS Error Estimation with Additional Environment Information Integration......Page 192
8.4.3.2 Measure the Relative Advantage of GPS Receivers´ Qualities......Page 193
8.4.3.3 Measure the POI Information of Road Segments......Page 195
8.4.3.5 Measure the Layout Information around Road Segments......Page 196
8.4.3.6 Optimization of the Objective Function......Page 197
8.4.4 Weighted Estimation of GEF......Page 198
8.5.1 Dataset Description......Page 199
8.5.3 Evaluation of the Matrix Completion Result......Page 200
References......Page 201
9.1 Introduction......Page 203
9.2.1 Taxi GPS Traces Mining......Page 206
9.2.2 Bus Network Design......Page 207
9.3.2 Cluster Merging and Splitting......Page 208
9.3.3 Candidate Bus Stop Location Selection......Page 211
9.4.2 Bus Route Graph Building and Pruning......Page 212
9.4.3 Automatic Candidate Bus Route Generation......Page 215
9.5 Experimental Evaluations......Page 218
9.5.2 Evaluation on Bus Route Selection Algorithm......Page 219
9.5.3 Bidirectional vs. Single Directional Bus Route......Page 226
9.5.4 Comparison with Real Routes and Impacts on Taxi Services......Page 227
9.5.5 Bus Capacity Analysis......Page 229
9.6 Conclusions and Future Work......Page 230
References......Page 231
Part V: Enabling Smart Urban Services: Travellers......Page 233
10.1 Introduction......Page 234
10.2.1 Construction of POI Network......Page 236
10.3 Trip Planner System......Page 237
10.3.1 Key Terminologies......Page 238
10.3.3 Framework......Page 239
10.4.1 Node Modeling......Page 240
10.4.2 Edge Modeling......Page 241
10.5 The Two-Phase Approach......Page 242
10.5.1 Phase I: Route Search......Page 243
10.5.2 Phase II: Route Augmentation......Page 244
10.6.2 Parameter Sensitivity Study......Page 249
10.6.3 Efficiency Evaluation......Page 250
10.7 Conclusions and Future Work......Page 253
References......Page 255
11.1 Introduction......Page 257
11.2.2 Travel Route Planning......Page 259
11.3.1 Basic Concepts......Page 260
11.3.2 Problem Formulation......Page 261
11.4.1 System Overview......Page 262
11.4.2 Phase I: Scenic Road Network Modeling......Page 263
11.4.3 Phase II: The MA for SSR Planning......Page 267
11.5.1 Experimental Setup......Page 272
11.5.2 Evaluation on Scenic Road Network Modeling......Page 273
11.5.3 Evaluation on Skyline Routes Planning Algorithm......Page 275
11.6 Conclusions and Future Work......Page 280
References......Page 282
Part VI: Enabling Smart Urban Services: Beyond People Transportation......Page 284
12.1 Introduction......Page 285
12.2 Related Work......Page 288
12.3.1 Basic Concepts......Page 289
12.3.2 Assumptions......Page 291
12.4 Overview of CrowdDeliver......Page 292
12.5.1 Phase I: Offline Trajectory Data Mining......Page 294
12.5.2 Phase II: Online Package Routing......Page 300
12.6.1 Experimental Setup......Page 302
12.6.3 Experimental Objectives......Page 304
12.6.4 Experimental Results......Page 305
12.7 Conclusions and Future Work......Page 313
References......Page 314
13.1 Introduction......Page 316
13.2.1.1 Basic Concepts......Page 318
13.2.2 Problem Statement......Page 322
13.2.3 System Overview......Page 323
13.3.2.1 From Trajectory Data to Passenger Flow......Page 324
13.3.2.2 From Passenger Flow to Time Distributions......Page 325
13.4 Online Taxi Scheduling and Package Routing......Page 326
13.4.2.1 Probability Computation if Hitchhiking the Current......Page 327
13.4.2.2 Probability Computation if Waiting for the Future......Page 329
13.5 Evaluations......Page 330
13.5.1.2 Package Delivery Request......Page 331
13.5.2 Baseline Algorithms......Page 332
13.5.3 Experimental Objectives......Page 333
13.5.4.1 Results of Response Time......Page 334
13.5.4.2 Results of Success Rate......Page 335
13.5.4.3 Results of Number of Relays......Page 338
13.5.4.4 Results of Package Throughput......Page 340
References......Page 342
Part VII: Open Issues, Future Directions and Conclusions......Page 344
14.1.2 Model Issue: Effectiveness or Lightweight?......Page 345
14.2.2 Lane-Level Trajectory-Based Services......Page 346
14.2.4 Promoting Smart City with Smart Mobility......Page 347
14.3 Conclusions......Page 348
References......Page 351