State Space Representation of a 1D Point Mass Floating in Space and Actuated by Two Lateral Thrusters

  • Thread starter Thread starter djulzz1982
  • Start date Start date
  • Tags Tags
    State space
AI Thread Summary
The discussion focuses on the state space representation of a 1D point mass in space, actuated by two lateral thrusters. The equations of motion are derived using Newton's second law, leading to a formulation that includes control inputs for each thruster. A state space representation is established, defining states related to position and velocity, which is necessary for designing a Linear Quadratic Regulator (LQR) controller. Participants emphasize the importance of clarity in mathematical notation, suggesting the use of LaTex for better readability. The conversation concludes with a request for feedback on a submitted PDF containing the solution to the posed homework questions.
djulzz1982
Messages
5
Reaction score
2
Homework Statement
. A point-mass not subjected to gravity is "floating" in space.
. The point-mass has a mass m, and has two lateral thrusters, opposite one another.
. The first thruster generates a max force T1, the second thruster generates a max force T2.
. For the sake of problem formulation, the point-mass can move along the x direction.
. Assume no friction (the point-mass is in space).
1) Write the equations of motion (EOM) for the 1D point-mass
2) Convert the derived EOS(s) to a state space representation
3) If feasible, design an Linear Quadratic Regulator (LQR), that drives the point-mass from x(0) = 0 to x(t) = 4
Relevant Equations
According to Newton's 2nd Law, Sum of forces = m . a, so
(T1-T2) = m a, where:
- T1 is the maximum force generated by the first thruster
- T2 is the maximum force generated by the second thruster
- a is the point-mass' acceleration, so a = d^2(x) / d^2(t).
(u1*T1) + (u2*T2) = m (x dot dot), [1.1]
where (x dot dot) is the 2nd derivative of the point-mass position with respect to time, u1 is the control input for the 1st thruster, u2 is the control input for the second thruster.
Rearranging equation 1.1 yields
(x dot dot) = (T1/m)*u1+ (T2/m)*u2 [1.2]
In order to design an LQR controller for such system, its state space formulation is required, as doing so allows to determine if the system is controllable.
The following states are thus defined
- x1 = x dot ↔ x dot dot = x1 dot
- x2 = x ↔ x2 dot = x1 [1.3]
Rearranging equation [1.2] with the state space variables defined in [1.3] yields:
- x1 dot = (T1/m)*u1 + (T2/m)*u2
- x2 dot = x1 [1.4]
The state space vector is thus x = [x1 x2]T
 
Physics news on Phys.org
djulzz1982 said:
The state space vector is thus x = [x1 x2]T
And your question is ##\dots##?
 
djulzz1982 said:
m (x dot dot)
Kind of hard to read. You'll get better response from us if you use LaTex. It's not hard to learn for basic stuff. There's a nice guide available at the link below your post window.

Also, it would be good to reserve "*" for inner products (or dot products) when vectors might be involved. There's an icon above the post window that looks like a little greek temple that you can use to insert some symbols, like "⋅"; or, better yet LaTex.

For example ##(u_1T_1) + (u_2T_2) =m \ddot {x}##
Or ##(\vec u_1 \cdot \vec T_1) + (\vec u_2 \cdot \vec T_2) =m \ddot {x}##
Or ##(u_1 \vec T_1) + ( u_2 \vec T_2) =m \ddot { \vec x}##

No worries, just some suggestions.
 
Last edited:
djulzz1982 said:
Homework Statement: . A point-mass not subjected to gravity is "floating" in space.
. The point-mass has a mass m, and has two lateral thrusters, opposite one another.
. The first thruster generates a max force T1, the second thruster generates a max force T2.
. For the sake of problem formulation, the point-mass can move along the x direction.
. Assume no friction (the point-mass is in space).
1) Write the equations of motion (EOM) for the 1D point-mass
2) Convert the derived EOS(s) to a state space representation
3) If feasible, design an Linear Quadratic Regulator (LQR), that drives the point-mass from x(0) = 0 to x(t) = 4
Relevant Equations: According to Newton's 2nd Law, Sum of forces = m . a, so
(T1-T2) = m a, where:
- T1 is the maximum force generated by the first thruster
- T2 is the maximum force generated by the second thruster
- a is the point-mass' acceleration, so a = d^2(x) / d^2(t).

(u1*T1) + (u2*T2) = m (x dot dot), [1.1]
where (x dot dot) is the 2nd derivative of the point-mass position with respect to time, u1 is the control input for the 1st thruster, u2 is the control input for the second thruster.
Rearranging equation 1.1 yields
(x dot dot) = (T1/m)*u1+ (T2/m)*u2 [1.2]
In order to design an LQR controller for such system, its state space formulation is required, as doing so allows to determine if the system is controllable.
The following states are thus defined
- x1 = x dot ↔ x dot dot = x1 dot
- x2 = x ↔ x2 dot = x1 [1.3]
Rearranging equation [1.2] with the state space variables defined in [1.3] yields:
- x1 dot = (T1/m)*u1 + (T2/m)*u2
- x2 dot = x1 [1.4]
The state space vector is thus x = [x1 x2]T
$$\underline{x} =\begin{bmatrix}
\dot{x_{1}} \\
\dot{x_{2}}
\end{bmatrix}
=
\begin{bmatrix}
0 & 0 \\
1 & 0
\end{bmatrix}
$$
djulzz1982 said:
Homework Statement: . A point-mass not subjected to gravity is "floating" in space.
. The point-mass has a mass m, and has two lateral thrusters, opposite one another.
. The first thruster generates a max force T1, the second thruster generates a max force T2.
. For the sake of problem formulation, the point-mass can move along the x direction.
. Assume no friction (the point-mass is in space).
1) Write the equations of motion (EOM) for the 1D point-mass
2) Convert the derived EOS(s) to a state space representation
3) If feasible, design an Linear Quadratic Regulator (LQR), that drives the point-mass from x(0) = 0 to x(t) = 4
Relevant Equations: According to Newton's 2nd Law, Sum of forces = m . a, so
(T1-T2) = m a, where:
- T1 is the maximum force generated by the first thruster
- T2 is the maximum force generated by the second thruster
- a is the point-mass' acceleration, so a = d^2(x) / d^2(t).

(u1*T1) + (u2*T2) = m (x dot dot), [1.1]
where (x dot dot) is the 2nd derivative of the point-mass position with respect to time, u1 is the control input for the 1st thruster, u2 is the control input for the second thruster.
Rearranging equation 1.1 yields
(x dot dot) = (T1/m)*u1+ (T2/m)*u2 [1.2]
In order to design an LQR controller for such system, its state space formulation is required, as doing so allows to determine if the system is controllable.
The following states are thus defined
- x1 = x dot ↔ x dot dot = x1 dot
- x2 = x ↔ x2 dot = x1 [1.3]
Rearranging equation [1.2] with the state space variables defined in [1.3] yields:
- x1 dot = (T1/m)*u1 + (T2/m)*u2
- x2 dot = x1 [1.4]
The state space vector is thus x = [x1 x2]T

djulzz1982 said:
Homework Statement: . A point-mass not subjected to gravity is "floating" in space.
. The point-mass has a mass m, and has two lateral thrusters, opposite one another.
. The first thruster generates a max force T1, the second thruster generates a max force T2.
. For the sake of problem formulation, the point-mass can move along the x direction.
. Assume no friction (the point-mass is in space).
1) Write the equations of motion (EOM) for the 1D point-mass
2) Convert the derived EOS(s) to a state space representation
3) If feasible, design an Linear Quadratic Regulator (LQR), that drives the point-mass from x(0) = 0 to x(t) = 4
Relevant Equations: According to Newton's 2nd Law, Sum of forces = m . a, so
(T1-T2) = m a, where:
- T1 is the maximum force generated by the first thruster
- T2 is the maximum force generated by the second thruster
- a is the point-mass' acceleration, so a = d^2(x) / d^2(t).

(u1*T1) + (u2*T2) = m (x dot dot), [1.1]
where (x dot dot) is the 2nd derivative of the point-mass position with respect to time, u1 is the control input for the 1st thruster, u2 is the control input for the second thruster.
Rearranging equation 1.1 yields
(x dot dot) = (T1/m)*u1+ (T2/m)*u2 [1.2]
In order to design an LQR controller for such system, its state space formulation is required, as doing so allows to determine if the system is controllable.
The following states are thus defined
- x1 = x dot ↔ x dot dot = x1 dot
- x2 = x ↔ x2 dot = x1 [1.3]
Rearranging equation [1.2] with the state space variables defined in [1.3] yields:
- x1 dot = (T1/m)*u1 + (T2/m)*u2
- x2 dot = x1 [1.4]
The state space vector is thus x = [x1 x2]T
 
I wrote the potential solution to the problem, showing that with the given problem statement, LQR, PID, or any other control approach will not allow the system to be controlled.
Please check out the attached PDF, and please provide feedback if you can.
 

Attachments

kuruman said:
And your question is ##\dots##?
I posted the answer to the questions:
1) what is the state space formulation
2) is the system controllable
all in the attached PDF.
Regards
 
djulzz1982 said:
I posted the answer to the questions:
1) what is the state space formulation
2) is the system controllable
all in the attached PDF.
So then why would we want to go get it and read it? Is it just another example of a solved HW problem (hint - it's your HW problem, not ours), or is there something particularly interesting about it?
 
DaveE said:
So then why would we want to go get it and read it? Is it just another example of a solved HW problem (hint - it's your HW problem, not ours), or is there something particularly interesting about it?
I am not sure of my answer, that's why I posted it. You are welcome to ignore it. And by solved, I solved it (hint).
 

Similar threads

State-Space model of spring-mass system
Replies
4
Views
5K
Modeling Fracture: Intuition, Formula for Elastic Potential & Nodal Forces
Replies
6
Views
6K
Chapter 10: A Tutorial on Causal Significance
Replies
20
Views
5K
I Progress In Calculating The HLbL Contribution To Muon g-2
Replies
0
Views
2K

Hot Threads

Recent Insights

Back
Top