Mitigating sensor noise in legged robots
An example implementation involves receiving measurements from an inertial sensor coupled to the robot and detecting an occurrence of a foot of the legged robot making contact with a surface. The implementation also involves reducing a gain value of an amplifier from a nominal value to a reduced value upon detecting the occurrence. The amplifier receives the measurements from the inertial sensor and provides a modulated output based on the gain value. The implementation further involves increasing the gain value from the reduced value to the nominal value over a predetermined duration of time after detecting the occurrence. The gain value is increased according to a profile indicative of a manner in which to increase the gain value of the predetermined duration of time. The implementation also involves controlling at least one actuator of the legged robot based on the modulated output during the predetermined duration of time.
1. A method comprising:
receiving, at a control system of a robot, sensor data as a feedback input, the sensor data captured from an inertial sensor disposed on a portion of the robot;
determining, by the control system, that the sensor data comprises a signal fluctuation indicative of the portion of the robot contacting a surface external to the robot;
generating, by the control system, a time-varying gain signal based on the signal fluctuation of the sensor data, the time-varying gain signal suppressing the signal fluctuation of the sensor data; and
controlling, by the control system, at least one actuator on the portion of the robot using an output from an amplifier receiving the time-varying gain signal as an input.
2. The method of claim 1 , further comprising:
receiving, at the control system, updated sensor data from the inertial sensor disposed on the portion of the robot;
determining, by the control system, that the updated sensor data indicates an attenuation of the signal fluctuation; and
adjusting, by the control system, the time-varying gain signal based on the attenuation of the signal fluctuation.
3. The method of claim 1 , wherein the signal fluctuation corresponds to vibrations measured by the inertial sensor in response to the portion of the robot contacting the surface external to the robot.
4. The method of claim 1 , wherein the signal fluctuation corresponds to an oscillating acceleration measured by the inertial sensor in response to the portion of the robot contacting the surface external to the robot.
5. The method of claim 1 , wherein generating the time-varying gain signal is additionally based on a gait being executed by the robot when the control system receives the sensor data comprising the signal fluctuation.
6. The method of claim 1 , wherein the time-varying gain signal is designated by a gain signal profile corresponding to a gait cycle.
7. The method of claim 6 , further comprising selecting the gain signal profile to match a gait being executed by the robot when the control system receives the sensor data comprising the signal fluctuation.
8. The method of claim 1 , wherein the time-varying gain signal comprises a nominal gain value and a reduced gain value with respect to the nominal gain value, the time-varying gain signal changing from the reduced gain value to the nominal gain value over a predetermined duration of time, the predetermined duration of time based on a decay rate of vibrations from the contact between the portion of the robot and the surface external to the robot.
9. The method of claim 1 , wherein the time-varying gain signal comprises a nominal gain value and a reduced gain value with respect to the nominal gain value, the time-varying gain signal linearly increasing from the reduced gain value to the nominal gain value over a predetermined duration of time.
10. The method of claim 1 , wherein the time-varying gain signal comprises a nominal gain value and a reduced gain value with respect to the nominal gain value, the time-varying gain signal exponentially increasing from the reduced gain value to the nominal gain value over a predetermined duration of time.
11. The method of claim 1 , wherein the time-varying gain signal comprises a nominal gain value and a reduced gain value with respect to the nominal gain value, the time-varying gain signal changing from the reduced gain value to the nominal gain value over a predetermined duration of time, the predetermined duration of time based on a stance duration for a leg of the robot.
12. A robot comprising:
an inertial sensor disposed on a portion of the robot; and
a control system of the robot configured to perform operations comprising:
receiving, at a control system of robot, sensor data as a feedback input, the sensor data captured from the inertial sensor disposed on the portion of the robot;
determining, by the control system, that the sensor data comprises a signal fluctuation indicative of the portion of the robot contacting a surface external to the robot;
generating, by the control system, a time-varying gain signal based on the signal fluctuation of the sensor data, the time-varying gain signal suppressing the signal fluctuation of the sensor data; and
controlling, by the control system, at least one actuator on the portion of the robot using an output from an amplifier receiving the time-varying gain signal as an input.
13. The robot of claim 12 , wherein the operations further comprise:
receiving updated sensor data from the inertial sensor disposed on the portion of the robot;
determining that the updated sensor data indicates an attenuation of the signal fluctuation; and
adjusting the time-varying gain signal based on the attenuation of the signal fluctuation.
14. The robot of claim 12 , wherein the signal fluctuation corresponds to vibrations measured by the inertial sensor in response to the portion of the robot contacting the surface external to the robot.
15. The robot of claim 12 , wherein the signal fluctuation corresponds to an oscillating acceleration measured by the inertial sensor in response to the portion of the robot contacting the surface external to the robot.
16. The robot of claim 12 , wherein generating the time-varying gain signal is additionally based on a gait being executed by the robot when the control system receives the sensor data comprising the signal fluctuation.
17. The robot of claim 12 , wherein the time-varying gain signal is designated by a gain signal profile corresponding to a gait cycle.
18. The robot of claim 17 , further comprising selecting the gain signal profile to match a gait being executed by the robot when the control system receives the sensor data comprising the signal fluctuation.
19. The robot of claim 12 , wherein the time-varying gain signal comprises a nominal gain value and a reduced gain value with respect to the nominal gain value, the time-varying gain signal changing from the reduced gain value to the nominal gain value over a predetermined duration of time, the predetermined duration of time based on a decay rate of vibrations from the contact between the portion of the robot and the surface external to the robot.
20. The robot of claim 12 , wherein the time-varying gain signal comprises a nominal gain value and a reduced gain value with respect to the nominal gain value, the time-varying gain signal linearly increasing from the reduced gain value to the nominal gain value over a predetermined duration of time.
21. The robot of claim 12 , wherein the time-varying gain signal comprises a nominal gain value and a reduced gain value with respect to the nominal gain value, the time-varying gain signal exponentially increasing from the reduced gain value to the nominal gain value over a predetermined duration of time.
22. The robot of claim 12 , wherein the time-varying gain signal comprises a nominal gain value and a reduced gain value with respect to the nominal gain value, the time-varying gain signal changing from the reduced gain value to the nominal gain value over a predetermined duration of time, the predetermined duration of time based on a stance duration for a leg of the robot.