A study of modeling signal dependent noise in robotic systems
Robotic systems are stochastic in nature. That is, there is an inherent error associated with motions commanded to the robot and those executed by the robot. Further, this error itself is a random variable with a particular probability distribution. In this project, we aim to characterize the uncert...
Saved in:
| 主要作者: | |
|---|---|
| 其他作者: | |
| 格式: | Final Year Project |
| 語言: | English |
| 出版: |
2017
|
| 主題: | |
| 在線閱讀: | http://hdl.handle.net/10356/72233 |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 總結: | Robotic systems are stochastic in nature. That is, there is an inherent error associated with motions commanded to the robot and those executed by the robot. Further, this error itself is a random variable with a particular probability distribution. In this project, we aim to characterize the uncertainty model associated with a given mobile robot. Specifically, we investigate whether signal-dependent noise model where the uncertainty depends on the robot’s motion and control inputs are a more suitable as compared to constant additive noise model. The formulation used in the parameterizes the uncertainty and recovers the parameters through maximization of log-likelihood of the state or measurement trajectory. We used Clearpath Ridgeback mobile robot as the experimental platform in the project. |
|---|