Scribd
Upload
252 views

Exercises 16.06 16.07 Matlab Simulink

This document provides instructions for completing MATLAB and Simulink tutorials. It includes two exercises: 1) Using MATLAB to model and simulate the velocity of a Mars lander during descent. Students are instructed to modify scripts to compute and plot the lander's velocity and acceleration over time. 2) Using Simulink to design a controller for the longitudinal dynamics of an aircraft. Students are instructed to create a model with blocks representing the aircraft dynamics, sensor, controller, and step input. The goal is to iteratively improve the controller transfer function until the output response is acceptable.

Uploaded by

flowerboy
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
252 views

Exercises 16.06 16.07 Matlab Simulink

This document provides instructions for completing MATLAB and Simulink tutorials. It includes two exercises: 1) Using MATLAB to model and simulate the velocity of a Mars lander during descent. Students are instructed to modify scripts to compute and plot the lander's velocity and acceleration over time. 2) Using Simulink to design a controller for the longitudinal dynamics of an aircraft. Students are instructed to create a model with blocks representing the aircraft dynamics, sensor, controller, and step input. The goal is to iteratively improve the controller transfer function until the output response is acceptable.

Uploaded by

flowerboy
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

16.06 & 16.07 Fall 2007 | MATLAB & Simulink Tutorials | violeta@mit.

edu

EXERCISES 16.06/16.07: MATLAB & Simulink Tutorials I. Class Materials


1. Download matlab_simulink_tutorial.tar From a web browser:
Download matlab_simulink_tutorial.tar from http://web.mit.edu/acmath/matlab/course16 to a local directory. On Windows, if you do not have WinZip, download the ZIP file instead.

Alternatively, on Athena:
athena% add acmath athena% cp /mit/acmath/matlab/course16/matlab_simulink_tutorial.tar .

2. Extract this sessions sub-directories and files On Athena (or the UNIX shell on Mac OS X):
tar xvf matlab_simulink_tutorial.tar

On laptops:
Use your computers utilities, such as double click or WinZip on Windows or StuffIt on Mac. Your local work directory matlab_simulink_tutorial should now contain the following directories and files: matlab_simulink_tutorial Exercise1 odeLanderVelocity.m MarsLander.m longTimeResponse.m f8long.m Exercise2 f8.mdl

II. Start MATLAB


On Athena:
athena% cd matlab_simulink_tutorial athena% add matlab athena% matlab & >> desktop

On laptops:
Launch MATLAB and navigate to the work directory matlab_simulink_tutorial.

16.06 & 16.07 Fall 2007 | MATLAB & Simulink Tutorials | [email protected]

III. Exercise 1: Matrices & Differential Equations


Purpose
To practice the following in MATLAB: Creating and referencing the elements of matrices and vectors. Using MATLABs differential equation solvers and other built-in functions. Understanding MATLAB programs with script and function M-files. To prepare for the Simulink tutorial and exercise.

1-A Mars Lander Velocity Background


The velocity v of a Mars Lander during EDL (Entry, Descent, Landing) can be described with the differential equation for velocity of a body in free fall: dv ( t ) k = g v2 dt m

Problem
Compute the velocity, v, and acceleration, dv/dt, for the last stage of landing (after the retrorockets are launched) for a Mars Lander with mass m=150 kg and drag coefficient k=1.2. The gravity constant on Mars is 3.688 m/s2. Plot v and dv/dt vs. time on the same plot and show that the acceleration approaches zero as the velocity approaches a terminal velocity value.

1. Open M-files odeLanderVelocity.m and MarsLander.m


>> cd Exercise1 >> edit odeLanderVelocity.m >> edit MarsLander.m Lines that start with % are comments. The rest are MATLAB commands.

2. Follow the instructions in MarsLander.m


M-file odeLanderVelocity.m defines a differential equation for velocity in free fall. Follow instructions in the file MarsLander.m and add the missing code. In A1, define global variables M and K for the mass and drag coefficient. In A3, define a time vector tspan for the numerical solution. In A4, use the built-in MATLAB ODE solver ode45 to compute the numerical solution, which uses function odeLanderVelocity, tspan, and initial velocity V0 as arguments. The solver should return two output arguments: a velocity vector V and a time vector t.

3. Execute the M-file in the Command Window


>> MarsLander Look at the plot that is created and interpret the results1.
1

For another Dynamics example for 16.07, see files getF.m and Central_Force_Motion.m.

16.06 & 16.07 Fall 2007 | MATLAB & Simulink Tutorials | [email protected]

1-B F-8 Longitudinal Time Response Background


The following system of first-order differential equations describes the longitudinal time response of an airplane due to elevator deflection:

i q a11 a12 i u = a21 a22 i a31 a32 i a41 a42


i

a13 a23 a33 a43

a14 q b11 a24 u b21 + [ ] a34 b31 e a44 b31

In matrix form the system can be represented as:

x = Ax + Bv
where the variables in the x vector are called state variables and define the state of the aircraft at any time, while the v vector contains the control variables. The following state variables represent perturbations from an equilibrium condition: q(t) is the pitch rate (deg/s) u(t) is the perturbation of horizontal velocity (ft/s) (t) is the perturbation of the angle of attack from trim (deg) (t) is the perturbation of the pitch angle from trim (deg) The control variable is e: the elevator deflection (deg) from equilibrium (trimmed) condition. When the equations are written in the following form, they are said to be in State Space form:
i

x = Ax + Bv y = Cx + Dv
where y is the output vector (which could be the same as or different from x). A, B, C, and D are called state, input, output, and feedthrough matrices, respectively. The free longitudinal response of an aircraft can be studied by setting the control input to zero:
i

x = Ax
Two natural modes of the system, called Short Period mode and Phugoid mode, are related to the eigenvalues of the matrix A.

16.06 & 16.07 Fall 2007 | MATLAB & Simulink Tutorials | [email protected]

Problem
The following linear differential equations model the longitudinal motions of the F-8 aircraft:
q ( t ) = 0.7q ( t ) 0.0344u ( t ) 12 ( t ) 19 e ( t ) u ( t ) = 0.014u ( t ) 0.290 ( t ) 0.562 ( t ) 0.0115 e ( t )
i i

( t ) = q ( t ) 0.0057u ( t ) 1.3 ( t ) 0.16 e ( t ) (t ) = q (t )


about the following equilibrium flight conditions: Altitude: Speed: Dynamic pressure: Trim pitch angle: Trim angle of attack: Trim elevator angle: 20,000 ft = 6094 m Mach 0.9 = 916 ft/s = 281.6 m/s 550 lbs/ft2 = 26,439 N/m2 2.25 deg 2.25 deg -2.65 deg
i

Define the state and input matrices A and B for the State Space form of the system. Compute the transfer functions relating the state variables to the control input. Compute and plot the change over time of the elements of the state vector x due to a perturbation in the elevator angle, e.

1. Open M-files longTimeResponse.m and f8long.m


>> edit longTiimeResponse.m >> edit f8long.m Lines that start with % are comments. The rest are MATLAB commands.

2. Follow the instructions in f8long.m



M-file longTimeReponse.m defines the differential equations that describe the longitudinal time response of an aircraft due to input of elevator deflection angle. Follow instructions in the file f8long.m and add the missing code. In B3, define the state and input matrices A and B. In B4, compute the eigenvalues and eigenvectors of A. In B5, compute the numerator and denominator of the transfer functions relating the state variables to the control input. Hint: Use the built-in function ss2tf. In B8, add the code for plotting the perturbation of the pitch angle (t) over time.

3. Execute the M-file in the Command Window


>> f8long What is the transfer function relating the horizontal velocity to the elevator deflection? u (s) G (s) = = ... Write it here: e ( s )

16.06 & 16.07 Fall 2007 | MATLAB & Simulink Tutorials | [email protected]

V. Exercise 2: F-8 Controller Design


Purpose
To practice the following in Simulink: Creating and saving models. Using blocks to create block diagrams. Configuring parameters and running simulations.

Background
In this exercise, we will be designing a controller for the longitudinal dynamics of the F-8. The transfer function of the horizontal speed of the F-8 to the elevator angle is approximately:

G (s) =

u (s) 964 = 2 e ( s ) 976s + 11.25s + 1

The controller takes the error in the speed as input and outputs the elevator angle. The F-8 dynamics then converts this elevator angle to airspeed. We will also model a sensor, which will have a small amount of lag. We are modeling the behavior of the system when the desired speed suddenly changes, which will be represented with a step function as input.

Problem
Create the Simulink model shown in the figure below (also saved in file f8.mdl). Run the simulation. Change the transfer function of the controller and re-run the simulation. Repeat until the output has improved.

1. Create the models block diagram

Create a new model in Simulink.

16.06 & 16.07 Fall 2007 | MATLAB & Simulink Tutorials | [email protected]

Add a step function input to the model: use Step Function from the Sources library. Add a scope to the model: use Scope from the Sinks library. Add three transfer functions to the model: Transfer Fcn from the Continuous library. Change the names of the transfer functions to Controller, Aircraft, and Sensor. Add a sum operator to the model: Sum from the Math Operations library. Change the sum operator to perform subtraction rather than addition. (Double click on the operator and change the list of signs to |+ - in the Block Parameters: Sum window). Change the transfer function of the aircraft. (Double click on the aircraft icon and change the nominator and denominator using MATLAB matrix notation in the Function Block Parameters: Aircraft window. In this exercise both the nominator and denominator are row vectors that represent the coefficients of a polynomial in decreasing exponent of s.) 1 Change the transfer function of the sensor to (i. e. a sensor with 0.01 sec lag). 0.01s + 1 Add an output to the MATLAB plot interface: To Workspace from the Sinks library. Connect the blocks of the model as shown in the figure.

2. Run the simulation


Set the simulation to run for 30 seconds: Simulation->Configuration Parameters. Run the simulation: Simulation->Start. Examine the output of the simulation in the Scope window. (Click on the binocular icon to show the most appropriate scale.) Is this an appropriate way for an aircraft to respond?

3. Change the controller


2.5s . 0.15s + 1 Run the simulation again and observe the output of the scope. Has the new controller improved the output? Your model should now look like this:
Change the transfer function of the controller to

In the MATLAB command line window, type the command who to see the variables in the workplace. Then plot simout vs. tout. What does this plot look like? 6

You might also like