Travel Time Estimation for Optimal Planning in Internal Transportation
Optimal planning depends on precise and exact estimation of the operation costs of mobile robots. Unfortunately, determining the current and future state of a vehicle implies identifying all the parameters in its model. Rather than broadening the number of factors, in this work we adopt the approach...
| Autores: | , |
|---|---|
| Tipo de recurso: | artículo |
| Fecha de publicación: | 2024 |
| País: | España |
| Institución: | Universitat Autònoma de Barcelona |
| Repositorio: | Dipòsit Digital de Documents de la UAB |
| Idioma: | inglés |
| OAI Identifier: | oai:ddd.uab.cat:307912 |
| Acceso en línea: | https://ddd.uab.cat/record/307912 https://dx.doi.org/urn:doi:10.3390/wevj15120565 |
| Access Level: | acceso abierto |
| Palabra clave: | Autonomous mobile robot systems Cost parameter estimation Cost efficiency Kalman filtering Optimal planning |
| Sumario: | Optimal planning depends on precise and exact estimation of the operation costs of mobile robots. Unfortunately, determining the current and future state of a vehicle implies identifying all the parameters in its model. Rather than broadening the number of factors, in this work we adopt the approach of using a higher-level abstraction model to identify only a few cost parameters. Based on the observation that arc travel times accurately reflect the effect of physical states, this work proposes using them as the key parameters to compute accurate path traversal costs in the context of indoor transportation. This approach eliminates the need to model all factors in order to derive the cost for every robot. The resulting model organizes those parameters in a bilinear state-space form and includes the evolution of actual travel times with changing states. We show that the proposed model accurately estimates arc travel times with respect to actual observations gathered from real robots traversing a few arcs of a traffic network until battery exhaustion. We experimentally obtained minimum-cost paths from random origin and destination nodes when using heuristics and the "closer-to-reality" (bilinear-state version of our model) path costs, finding that it can save an average of 15% in transportation time compared to conventional methods. |
|---|