Actuator Dynamics - Revision history

Hjkarssies at 22:22, 24 October 2020

2020-10-24T22:22:33Z

← Older revision		Revision as of 16:22, 24 October 2020
Line 28:		Line 28:
	Used for the setup of this example:		Used for the setup of this example:
	* [https://hobbyking.com/en_us/orange-rpm-sensor.html Orange RPM Sensor] for RPM measurements		* [https://hobbyking.com/en_us/orange-rpm-sensor.html Orange RPM Sensor] for RPM measurements
	* [https://www.saleae.com/ Saleae Logic 8] for recording measurements		* [https://www.saleae.com/ Saleae Logic 8 Digital Analyzer] for recording measurements
	* [https://www.mathworks.com/products/matlab.html MATLAB] for analyzing measurements<br><br>		* [https://www.mathworks.com/products/matlab.html MATLAB] for analyzing measurements<br><br>

	In order to get a representative model for the whole range of the motor, control the motor with varying speeds and varying step sizes in the PWM signal. You can use Paparazzi's module ''servo_tester'', which runs a predefined program. Try to experiment a bit with the module's settings to create a test program suitable for your purpose, and make sure to firmly attach your motor before running the program.		In order to get a representative model for the whole range of the motor, control the motor with varying speeds and varying step sizes in the PWM signal. You can use Paparazzi's module ''servo_tester'', which runs a predefined program. Try to experiment a bit with the module's settings to create a test program suitable for your purpose, and make sure to firmly attach your motor before running the program.<br><br>

	==~~WORK IN PROGRESS~~==		You can record the PWM and RPM signals with your digital analyzer. Make sure to select appropriate bitrates. Saleae provides a function in their software with which you can export your measurements to a MATLAB file, which you can read with the MATLAB script below.<br><br>

			==Data Processing==
			With the following MATLAB script, you can make the estimation of you first order actuator model's time constant. In the script, indicate the file name of your measurements, the channel numbers for the different types of data, and the timeframe in which you recorded your relevant data. You can also specify different values of tau to try. The script will try these values, and interpolate results to find the tau with which the simulated response best fits the measured response. Don't forget to set the correct frequency at which your module runs if you want to get a correct value for alpha as well. The minimal period settings are used to filter out any signal changes that are too quick, and are therefore probably misreadings.<br><br>

	<source lang="Matlab">		<source lang="Matlab">

Hjkarssies: /* Measuring Actuator Response */

2020-10-24T21:41:47Z

Measuring Actuator Response

← Older revision		Revision as of 15:41, 24 October 2020
Line 30:		Line 30:
	* [https://www.saleae.com/ Saleae Logic 8] for recording measurements		* [https://www.saleae.com/ Saleae Logic 8] for recording measurements
	* [https://www.mathworks.com/products/matlab.html MATLAB] for analyzing measurements<br><br>		* [https://www.mathworks.com/products/matlab.html MATLAB] for analyzing measurements<br><br>

			In order to get a representative model for the whole range of the motor, control the motor with varying speeds and varying step sizes in the PWM signal. You can use Paparazzi's module ''servo_tester'', which runs a predefined program. Try to experiment a bit with the module's settings to create a test program suitable for your purpose, and make sure to firmly attach your motor before running the program.

	==WORK IN PROGRESS==		==WORK IN PROGRESS==

Hjkarssies at 20:47, 18 October 2020

2020-10-18T20:47:18Z

← Older revision		Revision as of 14:47, 18 October 2020
Line 22:		Line 22:

	==Measuring Actuator Response==		==Measuring Actuator Response==
	The first step in approximating actuator dynamics is to measure actuator response.		The first step in approximating actuator dynamics is to measure actuator response. This can be done with by connecting an RPM sensor to the motor's wires, and recording both the ESC's input PWM and output RPM signals. An schematic representation of the setup used for this example is shown in the figure below:<br><br>

			[[File:Actuator_Dynamics_Measurements_Schematic.svg\|A schematic overview of a test setup to measure actuator dynamics\|400px]]<br><br>

			Used for the setup of this example:
			* [https://hobbyking.com/en_us/orange-rpm-sensor.html Orange RPM Sensor] for RPM measurements
			* [https://www.saleae.com/ Saleae Logic 8] for recording measurements
			* [https://www.mathworks.com/products/matlab.html MATLAB] for analyzing measurements<br><br>

			==WORK IN PROGRESS==

	<source lang="Matlab">		<source lang="Matlab">

Hjkarssies at 17:13, 18 October 2020

2020-10-18T17:13:01Z

@@ Line 4: / Line 4: @@
 [[File:Actuator_dynamics_matlab.svg|An example of estimated actuator dynamics in Matlab|thumb]]
-==Work in progess==
+Some stabilization modules require an estimation of the current actuator states. Examples are the [[Working with INDI|INDI]] and INCA modules. When actuator feedback is not available, this estimation can be mathematical model of the actuators. For brushless motors, a first order transfer function often suffices as actuator model. This guide describes how to accurately determine a first order approximation for a motor with a propeller. Note that the actuator dynamics vary for each configuration of different motors, propellers and ESCs (Electronic Speed Controllers). Used for this example are:
-Some stabilization modules require an estimation of the current actuator states. Examples are the INDI and INCA modules. When actuator feedback is not available, this estimation can be mathematical model of the actuators. This guide describes how to accurately determine a first order approximation for a motor with a propeller.
+==Actuator model==
 <source lang="Matlab">

Hjkarssies at 19:33, 23 September 2020

2020-09-23T19:33:06Z

← Older revision		Revision as of 13:33, 23 September 2020
Line 1:		Line 1:
	~~[[File:Actuator_dynamics_matlab.svg\|An example of estimated actuator dynamics in Matlab\|500px\|thumb]]~~
	[[File:Rpm_sensor.jpg\|An RPM sensor\|thumb]]		[[File:Rpm_sensor.jpg\|An RPM sensor\|thumb]]
			[[File:Saleae-logic.jpg\|A Saleae logic analyzer\|thumb]]
			[[File:Saleae-analog-example.png\|A screenshot from the Saleae logic analyser software\|thumb]]
			[[File:Actuator_dynamics_matlab.svg\|An example of estimated actuator dynamics in Matlab\|thumb]]

	==Work in progess==		==Work in progess==

Hjkarssies at 19:23, 23 September 2020

2020-09-23T19:23:46Z

← Older revision		Revision as of 13:23, 23 September 2020
Line 1:		Line 1:
	Work in progess		[[File:Actuator_dynamics_matlab.svg\|An example of estimated actuator dynamics in Matlab\|500px\|thumb]]
			[[File:Rpm_sensor.jpg\|An RPM sensor\|thumb]]

			==Work in progess==

	Some stabilization modules require an estimation of the current actuator states. Examples are the INDI and INCA modules. When actuator feedback is not available, this estimation can be mathematical model of the actuators. This guide describes how to accurately determine a first order approximation for a motor with a propeller.		Some stabilization modules require an estimation of the current actuator states. Examples are the INDI and INCA modules. When actuator feedback is not available, this estimation can be mathematical model of the actuators. This guide describes how to accurately determine a first order approximation for a motor with a propeller.

	~~[[File:Actuator_dynamics_matlab.svg\|An example of estimated actuator dynamics in Matlab\|frame\|right\|300px]]~~

	<source lang="Matlab">		<source lang="Matlab">

Hjkarssies at 19:19, 23 September 2020

2020-09-23T19:19:08Z

← Older revision		Revision as of 13:19, 23 September 2020
Line 3:		Line 3:
	Some stabilization modules require an estimation of the current actuator states. Examples are the INDI and INCA modules. When actuator feedback is not available, this estimation can be mathematical model of the actuators. This guide describes how to accurately determine a first order approximation for a motor with a propeller.		Some stabilization modules require an estimation of the current actuator states. Examples are the INDI and INCA modules. When actuator feedback is not available, this estimation can be mathematical model of the actuators. This guide describes how to accurately determine a first order approximation for a motor with a propeller.

	[[File:Actuator_dynamics_matlab.svg\|An example of estimated actuator dynamics in Matlab]]		[[File:Actuator_dynamics_matlab.svg\|An example of estimated actuator dynamics in Matlab\|frame\|right\|300px]]

	<source lang="Matlab">		<source lang="Matlab">

Hjkarssies at 19:18, 23 September 2020

2020-09-23T19:18:21Z

← Older revision		Revision as of 13:18, 23 September 2020
Line 2:		Line 2:

	Some stabilization modules require an estimation of the current actuator states. Examples are the INDI and INCA modules. When actuator feedback is not available, this estimation can be mathematical model of the actuators. This guide describes how to accurately determine a first order approximation for a motor with a propeller.		Some stabilization modules require an estimation of the current actuator states. Examples are the INDI and INCA modules. When actuator feedback is not available, this estimation can be mathematical model of the actuators. This guide describes how to accurately determine a first order approximation for a motor with a propeller.

			[[File:Actuator_dynamics_matlab.svg\|An example of estimated actuator dynamics in Matlab]]

	<source lang="Matlab">		<source lang="Matlab">

Hjkarssies: Added code

2020-09-23T19:05:28Z

Added code

Show changes

Hjkarssies at 19:05, 23 September 2020

2020-09-23T19:05:04Z

← Older revision		Revision as of 13:05, 23 September 2020
Line 3:		Line 3:
	Some stabilization modules require an estimation of the current actuator states. Examples are the INDI and INCA modules. When actuator feedback is not available, this estimation can be mathematical model of the actuators. This guide describes how to accurately determine a first order approximation for a motor with a propeller.		Some stabilization modules require an estimation of the current actuator states. Examples are the INDI and INCA modules. When actuator feedback is not available, this estimation can be mathematical model of the actuators. This guide describes how to accurately determine a first order approximation for a motor with a propeller.

	~~{{inline-code\|test}}~~		<source lang="Matlab">
			Test
			</source>