| 제목 |
확장 칼만필터를 활용한 토크 앵글 센서의 조향각 계산 알고리즘 고장 강건화 |
| 분야 |
지능형 샤시 및 제어 |
| 언어 |
Korean |
| 저자 |
성연빈(에이치엘만도), 허건행(에이치엘만도), 허진욱(에이치엘만도), 정상우(에이치엘만도) |
| Key Words |
Steering system(조향 시스템), Torque Angle Sensor(토크 앵글 센서), Extended Kalman Filter(확장 칼만 필터),Vernier Algorithm (버니어 알고리즘), Fault-tolerance(결함 허용) |
| 초록 |
In Electric Power Steering (EPS) systems, the vernier algorithm calculates the absolute steering angle by mapping torque and angle signals in a 1:1 ratio. However, when the travel distances of the two sensors become inconsistent due to gear wear or strong magnetic interference, a phenomenon known as "vernier reserve" occurs. This leads to distorted steering angle calculations and can provide unexpected assist outputs, posing safety risks. Existing fault-display methods prevent the presentation of distorted steering angles but have limitations in providing continuous steering angle inputs. This study proposes a method to enhance the robustness of the vernier algorithm against faults by utilizing an Extended Kalman Filter. The system model is constructed as a dynamic model of a rotating torsion bar applying Newton's laws, and the measurement model employs the Jacobian matrix of the vernier algorithm. This approach excludes incorrect measurements from the Torque Angle Sensor (TAS) during the estimation process and incorporates the system model's estimates to provide continuous torque and steering angle inputs even under fault conditions. The expected results include minimizing steering angle distortion caused by sensor mismeasurements and reducing control impact by leveraging the estimated steering angle. This is anticipated to achieve fault tolerance in angle calculations |
| 원문(PDF) |
다운로드
|
