Method for estimating the angular position of a rotor of an electric drive system

Patent metadata

Field	Value
Publication number	US-2018226906-A1
Application number	US-201815890479-A
Country	US
Kind code	A1
Filing date	Feb 7, 2018
Priority date	Feb 8, 2017
Publication date	Aug 9, 2018
Grant date	—

Abstract

The present invention has for object a method for estimating the angular position of a rotor in relation to a stator in a rotary electric machine, such as an electric machine of an electric or hybrid motorisation system, comprising:
estimating the angular position and/or of the rotation speed of the rotor by a method of injecting high frequency signals as long as the absolute value of the rotation speed of the rotor, derived from said angular position, is less than a first predefined threshold; estimating the angular position and/or of the rotation speed of the rotor by a model coming from a learning method as long as the absolute value of the rotation speed of the rotor, derived from said angular position, is greater than a second predefined threshold.

First claim

1 . A method for estimating the angular position of a rotor in relation to a stator of a rotary electric machine, such as an electric machine of an electric or hybrid motorisation system, comprising:
estimating the angular position and/or of the rotation speed of the rotor via a method of injecting high frequency signals as long as the absolute value of the rotation speed of the rotor, derived from said angular position, is less than a first predefined threshold (V 1 ); estimating the angular position and/or of the rotation speed of the rotor via a model coming from a learning method as long as the absolute value of the rotation speed of the rotor, derived from said angular position, is greater than a second predefined threshold.

2 . The method according to claim 1 , wherein the model comprises a neural network of which the weights and/or internal connections come from the learning method.

3 . The method according to claim 1 , wherein the method of injecting high frequency signals is a pulsating method.

4 . The method according to claim 1 , wherein the method of injecting high frequency signals is a rotary method.

5 . A method for controlling the torque produced by a rotary electric machine, in particular with permanent magnets, said electric machine comprising a rotor and a stator, a physical sensor that measures the angular position and/or the rotation speed of the rotor, with said method comprising the controlling of the torque to be produced by the electric machine according to the angular position and/or of the rotation speed of the rotor measured by the physical sensor as long as said physical sensor is available and, if said physical sensor is in fault, the controlling of the torque to be produced by the electric machine according to an estimated angular position and/or of a rotation speed of the rotor, according to the method comprising:
estimating the angular position and/or of the rotation speed of the rotor via a method of injecting high frequency signals as long as the absolute value of the rotation speed of the rotor, derived from said angular position, is less than a first predefined threshold; estimating the angular position and/or of the rotation speed of the rotor via a model coming from a learning method as long as the absolute value of the rotation speed of the rotor, derived from said angular position, is greater than a second predefined threshold.

6 . The method as claimed in claim 5 , further comprising the torque generated by the electric machine having the maximum value corresponding to a maximum current that can be reached or to a maximum voltage available, wherein, in the event of a fault with the physical sensor, the maximum value of the torque generated by the electric machine is reduced in such a way as to correspond to a downgraded maximum torque.

7 . A device for controlling an inverter supplying the phases of a stator of a rotary electric machine comprising moreover a rotor, said control device comprising a module for estimating the angular position and/or the rotation speed of the rotor configured to implement the method comprising:
estimating the angular position and/or of the rotation speed of the rotor via a method of injecting high frequency signals as long as the absolute value of the rotation speed of the rotor, derived from said angular position, is less than a first predefined threshold, estimating the angular position and/or of the rotation speed of the rotor via a model coming from a learning method as long as the absolute value of the rotation speed of the rotor, derived from said angular position, is greater than a second predefined threshold, wherein said estimated angular position and/or said rotation speed derived from said angular position of the rotor is taken into account for the controlling of said inverter.

8 . The device as claimed in claim 7 , comprising a first estimator in order to estimate the angular position and/or the rotation speed of the rotor via a method of injecting high frequency signals, a second estimator for estimating the angular position and/or the rotation speed of the rotor by the model coming from a learning method and a decision module in order to carry out the selection, according to the speed of the rotor derived from the angular position coming from the first estimator or from the second estimator, of the first or of the second estimator as a source of the estimated angular position and/or rotation speed of the rotor taken into account for the controlling of said inverter.

9 . An electric or hybrid motorisation system for vehicle, comprising a drive shaft driven by a rotor of a rotary electric machine moreover comprising a stator supplied by an inverter, as well as a device for controlling the inverter according to claim 7 .

10 . A motor vehicle comprising an electric or hybrid motorisation system as claimed in claim 9 .

11 . The method according to claim 2 , wherein the method of injecting high frequency signals is a pulsating method.

12 . The method according to claim 2 , wherein the method of injecting high frequency signals is a rotary method.

13 . The method according to claim 5 , wherein the model comprises a neural network of which the weights and/or internal connections come from the learning method.

14 . The method according to claim 5 , wherein the method of injecting high frequency signals is a pulsating method.

15 . The method according to claim 5 , wherein the method of injecting high frequency signals is a rotary method.

16 . The device according to claim 7 , wherein the model comprises a neural network of which the weights and/or internal connections come from the learning method.

17 . The device according to claim 7 , wherein the method of injecting high frequency signals is a pulsating method.

18 . The device according to claim 7 , wherein the method of injecting high frequency signals is a rotary method.

19 . An electric or hybrid motorisation system for vehicle, comprising a drive shaft driven by a rotor of a rotary electric machine moreover comprising a stator supplied by an inverter, as well as a device for controlling the inverter according to claim 8 .

20 . A motor vehicle comprising an electric or hybrid motorisation system as claimed in claim 19 .

Assignees

• Valeo Siemens Eautomotive France Sas

Inventors

• Chen Keyu
• Zine Wided
• Yu Lengfei
• Makni Zaatar

Classifications

• G06N3/09

  Supervised learning · CPC title

• G06N3/0499

  Feedforward networks · CPC title

• H02P6/08

  Arrangements for controlling the speed or torque of a single motor (H02P6/10, H02P6/28 take precedence) · CPC title

• G06N3/08

  Learning methods · CPC title

• H02P2203/05

  Determination of the rotor position by using two different methods and/or motor models · CPC title