| Autor: |
Dai Cong, Yong-zhi LlU, Ll Jie, Jin-long Song, Hao-shui Sun |
| Rok vydání: |
2018 |
| Předmět: |
|
| Zdroj: |
2018 IEEE CSAA Guidance, Navigation and Control Conference (CGNCC). |
| DOI: |
10.1109/gncc42960.2018.9018648 |
| Popis: |
In this paper, for the problem that a position sensor will add weight and complexity to switched reluctance motor (SRM) system and therefore negatively affect the system reliability, RBF neural network and terminal sliding mode control (TSMC) are applied, and adaptive neural terminal sliding mode observer is designed as a position sensorless method. The RBF neural network is used to approximate the control input of the observer, regardless of the upper bound of the disturbance, and the observation of the combined variable iωσLσθ can be achieved by controlling the current deviation to zero through terminal sliding mode. The working range of the motor is divided into the inductance approximate linear zone and the nonlinear zone according to the inductance characteristic. Thus, the complete period inductance mathematical model is established. The purpose of accurate tracking of the rotor position of switched reluctance motor is achieved by substituting the inductance and the current into the observation variable. The simulation results verify the accuracy of the estimating results under the circumstance of steady state and the sudden change of loading torque. The proposed method has good application prospect. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
