原文传递
Automatic Detection of RWIS Sensor Malfunctions (Phase II). Final rept.
| 题名： | Automatic Detection of RWIS Sensor Malfunctions (Phase II). Final rept. |
| 作者： | Crouch-C.; Crouch-D.; Maclin-R.; Plumetla-A. |
| 关键词： | *Roads-; *Weather-conditions; *Information-systems; *Sensors-.;Automation-; Algorithms-; Weather-forecasting; Hidden-Markov-model; Regression-analysis; Occurrence-; Evaluation-. |
| 摘要： | The overall goal of this project (Phases I & II) was to develop computerized procedures that detect Road Weather Information System (RWIS) sensor malfunctions. In the first phase of the research we applied three classification algorithms and six regression algorithms to data generated by RWIS sensors in order to predict malfunctions. In this phase we investigate the use of Hidden Markov models as predictors of sensor values. The Hidden Markov model (HMM) is a technique used to model a sequence of temporal events. For example, suppose we have the sequence of values produced by a given sensor over a fixed time period. An HMM can be used to produce this sequence and then to determine the probability of occurrence of another sequence of values, such as that produced by the given sensor for a subsequent time period. If the actual values produced by the sensor deviate from the predicted sequence then a malfunction may have occurred. This report provides an overview of the Hidden Markov model and three algorithms, namely, the Forward-Backward algorithms, the Baum-Welch algorithm, and the Viterbi algorithm, that were used in our development of Hidden Markov models to predict sensor values. We performed a series of experiments to evaluate the use of HMMs as predictors of temperature and precipitation sensor values. |
| 报告类型： | 科技报告 |
相关文献
- Automatic Detection of RWIS Sensor Malfunctions (Phase I). Final rept.
- Models for Predicting RWIS Sensor Misalignments and Their Causes.
- Development of a Road Weather Information System (RWIS) for Florida's Intelligent Transportation System, Phase 2. Final rept.
- Indications of Propulsion System Malfunctions - Sustained Thrust Anomaly Study.
- Automation in General Aviation: Two Studies of Pilot Responses to Autopilot Malfunctions
- FUZZY VARIABLE SPEED LIMIT DEVICE MODIFICATION AND TESTING PROJECT, PHASE II.
检索历史
应用推荐
