18/04/2011

Backstepping - Wikipedia, the free enc

Backstepping
From Wikipedia, the free encyclopedia

In control theory, backstepping is a technique developed circa 1990 by Petar V. Kokotovic and others[1][2] for designing stabilizing controls for a special class of nonlinear dynamical systems. These systems are built from subsystems that radiate out from an irreducible subsystem that can be stabilized using some other method. Because of this recursive structure, the designer can start the design process at the known-stable system and "back out" new controllers that progressively stabilize each outer subsystem. The process terminates when the final external control is reached. Hence, this process is known as backstepping.[3]

Contents
1 Recursive Control Design Overview 2 Integrator Backstepping 2.1 Single-integrator Equilibrium 2.2 Single-integrator Backstepping 2.3 Motivating Example: Two-integrator Backstepping 2.4 Many-integrator backstepping 3 Generic Backstepping 3.1 Single-step Procedure 3.2 Many-step Procedure 4 Flexible-joint manipulators (the Lozano-Brogliato's scheme) 5 See also 6 References The backstepping approach provides a recursive method for stabilizing the origin of a system in strict-feedback form. That is, consider a system of the form[3]

where with , are scalars,

u is a scalar input to the system,

vanish at the origin (i.e., are nonzero over the domain of interest (i.e., Also assume that the subsystem ), for ).

is stabilized to the origin (i.e., ) by some known control such that . It is also assumed that a Lyapunov function Vx for this stable subsystem is known. That is, this subsystem is stabilized by some other method and backstepping extends its stability to the shell around it. In systems of this strict-feedback form around a stable subsystem,

The backstepping-designed control input u has its most immediate stabilizing impact on state zn. The state zn then acts like a stabilizing control on the state zn 1 before it. This process continues so that each state zi is stabilized by the fictitious "control" zi + 1. The backstepping approach determines how to stabilize the subsystem using z1, and then proceeds with determining how to make the next state z2 drive z1 to the control required to stabilize . Hence, the process "steps backward" from out of the strict-feedback form system until the ultimate control u is designed.

Recursive Control Design Overview

1. It is given that the smaller (i.e., lower-order) subsystem

is already stabilized to the origin by some control where . That is, choice of ux to stabilize this system must occur using some other method. It is also assumed that a Lyapunov function Vx for this stable subsystem is known. Backstepping provides a way to extend the controlled stability of this subsystem to the larger system. 2. A control is designed so that the system

en.wikipedia.org/wiki/Backstepping

1/9

18/04/2011

Backstepping - Wikipedia, the free enc

is stabilized so that z1 follows the desired u x control. The control design is based on the augmented Lyapunov function candidate

The control u1 can be picked to bound 3. A control

away from zero.

is designed so that the system

is stabilized so that z2 follows the desired u 1 control. The control design is based on the augmented Lyapunov function candidate

The control u2 can be picked to bound

away from zero.

4. This process continues until the actual u is known, and The real control u stabilizes zk to fictitious control uk 1. The fictitious control uk 1 stabilizes zk 1 to fictitious control uk 2. The fictitious control uk 2 stabilizes zk 2 to fictitious control uk 3. The fictitious control u2 stabilizes z2 to fictitious control u 1. The fictitious control u1 stabilizes z1 to fictitious control u x . The fictitious control ux stabilizes to the origin. This process is known as backstepping because it starts with the requirements on some internal subsystem for stability and progressively steps back out of the system, maintaining stability at each step. Because

fi vanish at the origin for gi are nonzero for the given control u x has

, , , , z1 = 0, z2 = 0, , zk 1 = 0, and zk = 0) that is globally asymptotically stable.

then the resulting system has an equilibrium at the origin (i.e., where

Integrator Backstepping
Before describing the backstepping procedure for general strict-feedback form dynamical systems, it is convenient to discuss the approach for a smaller class of strict-feedback form systems. These systems connect a series of integrators to the input of a system with a known feedback-stabilizing control law, and so the stabilizing approach is known as integrator backstepping. With a small modification, the integrator backstepping approach can be extended to handle all strict-feedback form systems.

Single-integrator Equilibrium
Consider the dynamical system

where subsystem). We assume that

and z1 is a scalar. This system is a cascade connection of an integrator with the , and so if u 1 = 0, and z1 = 0, then

subsystem (i.e., the input u enters an integrator, and the integral z1 enters the

So the origin

is an equilibrium (i.e., a stationary point) of the system. If the system ever reaches the origin, it will remain there forever after.

Single-integrator Backstepping
In this example, backstepping is used to stabilize the single-integrator system in Equation (1) around its equilibrium at the origin. To be less precise, we wish to design a control law that ensures that the states return to after the system is started from some arbitrary initial condition. First, by assumption, the subsystem

with

has a Lyapunov function

such that

where is a positive-definite function. That is, we assume that we have already shown that this existing simpler Lyapunov). Roughly speaking, this notion of stability means that:

subsystem is stable (in the sense of

en.wikipedia.org/wiki/Backstepping

2/9

18/04/2011

Backstepping - Wikipedia, the free enc

The function Vx is like a "generalized energy" of the subsystem. As the states of the system move away from the origin, the energy also grows. By showing that over time, the energy decays to zero, then the states must decay toward . That is, the origin will be a stable equilibrium of the system the states will continuously approach the origin as time increases. Saying that is positive definite means that everywhere except for , and . The statement that means that is bounded away from zero for all points except where . That is, so long as the system is not at its equilibrium at the origin, its "energy" will be decreasing. Because the energy is always decaying, then the system must be stable; its trajectories must approach the origin. Our task is to find a control u that makes our cascaded system also stable. So we must find a new Lyapunov function candidate for this new system. That candidate will depend upon the control u , and by choosing the control properly, we can ensure that it is decaying everywhere as well. Next, by adding and subtracting it becomes (i.e., we don't change the system in any way because we make no net effect) to the part of the larger system,

which we can re-group to get

So our cascaded supersystem encapsulates the known-stable We now can change variables from to by letting

subsystem plus some error perturbation generated by the integrator. . So

Additionally, we let

so that

and

We seek to stabilize this error system by feedback through the new control v1. By stabilizing the system at e1 = 0, the state z1 will track the desired control ux which will result in stabilizing the inner subsystem. From our existing Lyapunov function Vx , we define the augmented Lyapunov function candidate

So

By distributing

, we see that

To ensure that

(i.e., to ensure stability of the supersystem), we pick the control law

with k1 > 0, and so

After distributing the e1 through,

en.wikipedia.org/wiki/Backstepping

3/9

18/04/2011

Backstepping - Wikipedia, the free enc

So our candidate Lyapunov function V1 is a true Lyapunov function, and our system is stable under this control law v1 (which corresponds the control law u 1 because ). Using the variables from the original coordinate system, the equivalent Lyapunov function

As discussed below, this Lyapunov function will be used again when this procedure is applied iteratively to multiple-integrator problem. Our choice of control v1 ultimately depends on all of our original state variables. In particular, the actual feedback-stabilizing control law

The states and z1 and functions fx and gx come from the system. The function ux comes from our known-stable affects the convergence rate or our system. Under this control law, our system is stable at the origin

subsystem. The gain parameter k1 > 0 .

Recall that u 1 in Equation (3) drives the input of an integrator that is connected to a subsystem that is feedback-stabilized by the control law u x . Not surprisingly, the control term that will be integrated to follow the stabilizing control law plus some offset. The other terms provide damping to remove that offset and any other perturbation effects that would be magnified by the integrator.

u1 has a

So because this system is feedback stabilized by in another single-integrator cascade system.

and has Lyapunov function

with

, it can be used as the upper subsystem

Motivating Example: Two-integrator Backstepping

Before discussing the recursive procedure for the general multiple-integrator case, it is instructive to study the recursion present in the two-integrator case. That is, consider the dynamical system

where and z1 and z2 are scalars. This system is a cascade connection of the single-integrator system in Equation (1) with another integrator (i.e., the input u2 enters through an integrator, and the output of that integrator enters the system in Equation (1) by its u1 input). By letting , ,

then the two-integrator system in Equation (4) becomes the single-integrator system

stabilizes the upper z2-to- subsystem using the Lyapunov function , and so Equation (5) is a new single-integrator system that is structurally equivalent to the single-integrator system in Equation (1). So a stabilizing control u 2 can be found using the same single-integrator procedure that was used to find u 1. By the single-integrator procedure, the control law

Many-integrator backstepping
In the two-integrator case, the upper single-integrator subsystem was stabilized yielding a new single-integrator system that can be similarly stabilized. This recursive procedure can be extended to handle any finite number of integrators. This claim can be formally proved with mathematical induction. Here, a stabilized multiple-integrator system is built up from subsystems of already-stabilized multiple-integrator subsystems. First, consider the dynamical system

that has scalar input u x and output states

. Assume that . In this case, the origin is called an equilibrium of the system.

so that the zero-input (i.e., ux = 0) system is stationary at the origin The feedback control law stabilizes the system at the equilibrium at the origin. A Lyapunov function corresponding to this system is described by .

That is, if output states are fed back to the input ux by the control law , then the output states (and the Lyapunov function) return to the origin after a single perturbation (e.g., after a nonzero initial condition or a sharp disturbance). This subsystem is stabilized by feedback control law u x . Next, connect an integrator to input u x so that the augmented system has input u 1 (to the integrator) and output states . The resulting augmented dynamical system is

en.wikipedia.org/wiki/Backstepping

4/9

18/04/2011

Backstepping - Wikipedia, the free enc

This "cascade" system matches the form in Equation (1), and so the single-integrator backstepping procedure leads to the stabilizing control law in Equation (3). That is, if we feed back states z1 and to input u 1 according to the control law

with gain k1 > 0, then the states z1 and will return to z1 = 0 and corresponding Lyapunov function from Equation (2) is

after a single perturbation. This subsystem is stabilized by feedback control law u 1, and the

That is, under feedback control law u 1, the Lyapunov function V1 decays to zero as the states return to the origin. Connect a new integrator to input u 1 so that the augmented system has input u 2 and output states . The resulting augmented dynamical system is

which is equivalent to the single-integrator system

Using these definitions of

, f1, and g 1, this system can also be expressed as

This system matches the single-integrator structure of Equation (1), and so the single-integrator backstepping procedure can be applied again. That is, if we feed back states z1, z2, and to input u2 according to the control law

with gain k2 > 0, then the states z1, z2, and will return to z1 = 0, z2 = 0, and u2, and the corresponding Lyapunov function is

after a single perturbation. This subsystem is stabilized by feedback control law

That is, under feedback control law u 2, the Lyapunov function V2 decays to zero as the states return to the origin. Connect an integrator to input u 2 so that the augmented system has input u 3 and output states . The resulting augmented dynamical system is

which can be re-grouped as the single-integrator system

By the definitions of

, f1, and g 1 from the previous step, this system is also represented by

en.wikipedia.org/wiki/Backstepping

5/9

18/04/2011
Further, using these definitions of

Backstepping - Wikipedia, the free enc

, f2, and g2, this system can also be expressed as

So the re-grouped system has the single-integrator structure of Equation (1), and so the single-integrator backstepping procedure can be applied again. That is, if we feed back states z1, z2, z3, and to input u 3 according to the control law

with gain k3 > 0, then the states z1, z2, z3, and will return to z1 = 0, z2 = 0, z3 = 0, and feedback control law u3, and the corresponding Lyapunov function is

after a single perturbation. This subsystem is stabilized by

That is, under feedback control law u 3, the Lyapunov function V3 decays to zero as the states return to the origin. This process can continue for each integrator added to the system, and hence any system of the form

has the recursive structure

and can be feedback stabilized by finding the feedback-stabilizing control and Lyapunov function for the single-integrator subsystem (i.e., with input z2 and output ) and iterating out from that inner subsystem until the ultimate feedback-stabilizing control u is known. At iteration i, the equivalent system is

The corresponding feedback-stabilizing control law is

with gain ki > 0. The corresponding Lyapunov function is

By this construction, the ultimate control

(i.e., ultimate control is found at final iteration i = k).

en.wikipedia.org/wiki/Backstepping

6/9

18/04/2011

Backstepping - Wikipedia, the free enc

Hence, any system in this special many-integrator strict-feedback form can be feedback stabilized using a straight-forward procedure that can even be automated (e.g., as part of an adaptive control algorithm).

Generic Backstepping
Systems in the special strict-feedback form have a recursive structure similar to the many-integrator system structure. Likewise, they are stabilized by stabilizing the smallest cascaded system and then backstepping to the next cascaded system and repeating the procedure. So it is critical to develop a single-step procedure; that procedure can be recursively applied to cover the many-step case. Fortunately, due to the requirements on the functions in the strict-feedback form, each single-step system can be rendered by feedback to a single-integrator system, and that single-integrator system can be stabilized using methods discussed above.

Single-step Procedure
Consider the simple strict-feedback system

where ,

z1 and u 1 are scalars, For all and z1,

Rather than designing feedback-stabilizing control u 1 directly, introduce a new control u a1 (to be designed later) and use control law

which is possible because

. So the system in Equation (6) is

which simplifies to

This new ua1-to- system matches the single-integrator cascade system in Equation (1). Assuming that a feedback-stabilizing control law for the upper subsystem is known, the feedback-stabilizing control law from Equation (3) is

and Lyapunov function

with gain k1 > 0. So the final feedback-stabilizing control law is

with gain k1 > 0. The corresponding Lyapunov function from Equation (2) is

Because this strict-feedback system has a feedback-stabilizing control and a corresponding Lyapunov function, it can be cascaded as part of a larger strict-feedback system, and this procedure can be repeated to find the surrounding feedback-stabilizing control.

Many-step Procedure
As in many-integrator backstepping, the single-step procedure can be completed iteratively to stabilize an entire strict-feedback system. In each step, 1. 2. 3. 4. The smallest "unstabilized" single-step strict-feedback system is isolated. Feedback is used to convert the system into a single-integrator system. The resulting single-integrator system is stabilized. The stabilized system is used as the upper system in the next step.

That is, any strict-feedback system

en.wikipedia.org/wiki/Backstepping

7/9

18/04/2011

Backstepping - Wikipedia, the free enc

has the recursive structure

and can be feedback stabilized by finding the feedback-stabilizing control and Lyapunov function for the single-integrator subsystem (i.e., with input z2 and output ) and iterating out from that inner subsystem until the ultimate feedback-stabilizing control u is known. At iteration i, the equivalent system is

By Equation (7), the corresponding feedback-stabilizing control law is

with gain ki > 0. By Equation (8), the corresponding Lyapunov function is

By this construction, the ultimate control (i.e., ultimate control is found at final iteration i = k). Hence, any strict-feedback system can be feedback stabilized using a straight-forward procedure that can even be automated (e.g., as part of an adaptive control algorithm).

Flexible-joint manipulators (the Lozano-Brogliato's scheme)

Almost simultaneously with Kanellakopoulos-Kokotovic-Morse (whose paper appeared in the 1991 November issue of IEEE Transactions on Automatic Control), Lozano and Brogliato published a paper (see reference [2] below) in the 1992 February issue of the same journal, which proposed a solution for the adaptive tracking control of a class of underactuated mechanical systems that possess a triangular form: flexible-joint manipulators. This controller, that belongs to the class of what has been named afterwards passivitybased control, is in essence designed from a backstepping method, though this was not called this way by the authors (and has remained a largely ignored fact in the Control literature). This is obvious looking at equation (6) and the derivations of Appendix B in [2]: the signal q2d in (6) plays the role of the "fictitious" input of the backstepping procedure. This control algorithm may be seen also as the extension towards flexible-joint systems, of the celebrated Slotine-Li's scheme for rigid manipulators. It is noteworthy that this adaptive scheme remains the only available solution for the global tracking control of such mechanical systems, without the a priori knowledge of the inertial parameters and of the stiffness matrix K. This last point is quite important because the stiffness matrix is precisely the term which pre-multiplies the "fictitious" input in the backstepping process. Relaxing its a priori knowledge is therefore a hard task. Even in the fixed-parameter case the proposed controller has a strong interest due to its intrinsic robustness properties. Experiments and comparisons with other control methods have been presented in.[4][5] In [6] it is shown that the fixed-parameter Lozano-Brogliato's scheme has the advantage over other backstepping-based schemes, that the Lyapunov function of the closed-loop system is fixed in advance (it looks like the total mechanical energy of the system), and allows one to avoid undesirable behaviors that a "blind" backstepping design may have (like high-gain in the loop that may create undesirable oscillations which may hamper the practical applications to work well). Further informations may be found in the monograph.[7]

en.wikipedia.org/wiki/Backstepping

8/9

18/04/2011

Backstepping - Wikipedia, the free enc

See also
Nonlinear control Strict-feedback form Robust control Adaptive control

References
1. ^ Kokotovic, P.V. (1992). "The joy of feedback: nonlinear and adaptive" (http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=165507) . Control Systems Magazine, IEEE 12 (3): 7 17. doi:10.1109/37.165507 (http://dx.doi.org/10.1109%2F37.165507) . http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=165507. Retrieved 2008-04-13. 2. ^ R. Lozano, B. Brogliato (1992). "Adaptive control of robot manipulators with flexible joints". IEEE Transactions on Automatic Control, 37 (2): 174181. 3. ^ a b Khalil, H.K. (2002). onlinear Systems (http://www.egr.msu.edu/~khalil/NonlinearSystems/) (3rd ed.). Upper Saddle River, NJ: Prentice Hall. ISBN 0-13-067389-7. http://www.egr.msu.edu/~khalil/NonlinearSystems/. 4. ^ B., Brogliato et al (1998). "Experimental comparison of nonlinear controllers for flexible joint manipulators". Int. Journal of Robotics Research 17 (3): 260281. 5. ^ B. Brogliato, D. Rey (1998). "Further experimental results on nonlinear control of flexible joint manipulators". Proc. of the American Control Conference 4: 22092211. 6. ^ B., Brogliato et al (1995). "Global tracking controllers for flexible-joint manipulators: a comparative study". Automatica 31 (7): 941956. 7. ^ B. Brogliato, et al (2007). "Dissipative Systems Analysis and Control (2nd Edition)". Springer Verlag, Communications and Control Engineering.

Retrieved from "http://en.wikipedia.org/wiki/Backstepping" Categories: Control theory | Nonlinear control This page was last modified on 3 February 2011 at 15:21. Text is available under the Creative Commons Attribution-ShareAlike License; additional terms may apply. See Terms of Use for details. Wikipedia is a registered trademark of the Wikimedia Foundation, Inc., a non-profit organization.

en.wikipedia.org/wiki/Backstepping

9/9