Data-driven control system

Last updated

Data-driven control systems are a broad family of control systems, in which the identification of the process model and/or the design of the controller are based entirely on experimental data collected from the plant. [1]

Contents

In many control applications, trying to write a mathematical model of the plant is considered a hard task, requiring efforts and time to the process and control engineers. This problem is overcome by data-driven methods, which fit a system model to the experimental data collected, choosing it in a specific models class. The control engineer can then exploit this model to design a proper controller for the system. However, it is still difficult to find a simple yet reliable model for a physical system, that includes only those dynamics of the system that are of interest for the control specifications. The direct data-driven methods allow to tune a controller, belonging to a given class, without the need of an identified model of the system. In this way, one can also simply weight process dynamics of interest inside the control cost function, and exclude those dynamics that are out of interest.

Overview

The standard approach to control systems design is organized in two-steps:

  1. Model identification aims at estimating a nominal model of the system , where is the unit-delay operator (for discrete-time transfer functions representation) and is the vector of parameters of identified on a set of data. Then, validation consists in constructing the uncertainty set that contains the true system at a certain probability level.
  2. Controller design aims at finding a controller achieving closed-loop stability and meeting the required performance with .

Typical objectives of system identification are to have as close as possible to , and to have as small as possible. However, from an identification for control perspective, what really matters is the performance achieved by the controller, not the intrinsic quality of the model.

One way to deal with uncertainty is to design a controller that has an acceptable performance with all models in , including . This is the main idea behind robust control design procedure, that aims at building frequency domain uncertainty descriptions of the process. However, being based on worst-case assumptions rather than on the idea of averaging out the noise, this approach typically leads to conservative uncertainty sets. Rather, data-driven techniques deal with uncertainty by working on experimental data, and avoiding excessive conservativism.

In the following, the main classifications of data-driven control systems are presented.

Indirect and direct methods

There are many methods available to control the systems. The fundamental distinction is between indirect and direct controller design methods. The former group of techniques is still retaining the standard two-step approach, i.e. first a model is identified, then a controller is tuned based on such model. The main issue in doing so is that the controller is computed from the estimated model (according to the certainty equivalence principle), but in practice . To overcome this problem, the idea behind the latter group of techniques is to map the experimental data directly onto the controller, without any model to be identified in between.

Iterative and noniterative methods

Another important distinction is between iterative and noniterative (or one-shot) methods. In the former group, repeated iterations are needed to estimate the controller parameters, during which the optimization problem is performed based on the results of the previous iteration, and the estimation is expected to become more and more accurate at each iteration. This approach is also prone to on-line implementations (see below). In the latter group, the (optimal) controller parametrization is provided with a single optimization problem. This is particularly important for those systems in which iterations or repetitions of data collection experiments are limited or even not allowed (for example, due to economic aspects). In such cases, one should select a design technique capable of delivering a controller on a single data set. This approach is often implemented off-line (see below).

On-line and off-line methods

Since, on practical industrial applications, open-loop or closed-loop data are often available continuously, on-line data-driven techniques use those data to improve the quality of the identified model and/or the performance of the controller each time new information is collected on the plant. Instead, off-line approaches work on batch of data, which may be collected only once, or multiple times at a regular (but rather long) interval of time.

Iterative feedback tuning

The iterative feedback tuning (IFT) method was introduced in 1994, [2] starting from the observation that, in identification for control, each iteration is based on the (wrong) certainty equivalence principle.

IFT is a model-free technique for the direct iterative optimization of the parameters of a fixed-order controller; such parameters can be successively updated using information coming from standard (closed-loop) system operation.

Let be a desired output to the reference signal ; the error between the achieved and desired response is . The control design objective can be formulated as the minimization of the objective function:

Given the objective function to minimize, the quasi-Newton method can be applied, i.e. a gradient-based minimization using a gradient search of the type:

The value is the step size, is an appropriate positive definite matrix and is an approximation of the gradient; the true value of the gradient is given by the following:

The value of is obtained through the following three-step methodology:

  1. Normal Experiment: Perform an experiment on the closed loop system with as controller and as reference; collect N measurements of the output , denoted as .
  2. Gradient Experiment: Perform an experiment on the closed loop system with as controller and 0 as reference ; inject the signal such that it is summed to the control variable output by , going as input into the plant. Collect the output, denoted as .
  3. Take the following as gradient approximation: .

A crucial factor for the convergence speed of the algorithm is the choice of ; when is small, a good choice is the approximation given by the Gauss–Newton direction:

Noniterative correlation-based tuning

Noniterative correlation-based tuning (nCbT) is a noniterative method for data-driven tuning of a fixed-structure controller. [3] It provides a one-shot method to directly synthesize a controller based on a single dataset.

Suppose that denotes an unknown LTI stable SISO plant, a user-defined reference model and a user-defined weighting function. An LTI fixed-order controller is indicated as , where , and is a vector of LTI basis functions. Finally, is an ideal LTI controller of any structure, guaranteeing a closed-loop function when applied to .

The goal is to minimize the following objective function:

is a convex approximation of the objective function obtained from a model reference problem, supposing that .

When is stable and minimum-phase, the approximated model reference problem is equivalent to the minimization of the norm of in the scheme in figure.

The idea is that, when G is stable and minimum phase, the approximated model reference problem is equivalent to the minimization of the norm of
e
{\displaystyle \varepsilon }
. Noniterative Correlation-based Tuning Scheme.svg
The idea is that, when G is stable and minimum phase, the approximated model reference problem is equivalent to the minimization of the norm of .

The input signal is supposed to be a persistently exciting input signal and to be generated by a stable data-generation mechanism. The two signals are thus uncorrelated in an open-loop experiment; hence, the ideal error is uncorrelated with . The control objective thus consists in finding such that and are uncorrelated.

The vector of instrumental variables is defined as:

where is large enough and , where is an appropriate filter.

The correlation function is:

and the optimization problem becomes:

Denoting with the spectrum of , it can be demonstrated that, under some assumptions, if is selected as:

then, the following holds:

Stability constraint

There is no guarantee that the controller that minimizes is stable. Instability may occur in the following cases:

Consider a stabilizing controller and the closed loop transfer function . Define:

Theorem
The controller stabilizes the plant if
  1. is stable
  2. s.t.

Condition 1. is enforced when:

The model reference design with stability constraint becomes:

A convex data-driven estimation of can be obtained through the discrete Fourier transform.

Define the following:

For stable minimum phase plants, the following convex data-driven optimization problem is given:

Virtual reference feedback tuning

Virtual Reference Feedback Tuning (VRFT) is a noniterative method for data-driven tuning of a fixed-structure controller. It provides a one-shot method to directly synthesize a controller based on a single dataset.

VRFT was first proposed in [4] and then extended to LPV systems. [5] VRFT also builds on ideas given in [6] as .

The main idea is to define a desired closed loop model and to use its inverse dynamics to obtain a virtual reference from the measured output signal .

The main idea is to define a desired closed loop model M and to use its inverse dynamics to obtain a virtual reference from the measured output signal y. Virtual Reference Feedback Tuning Scheme.svg
The main idea is to define a desired closed loop model M and to use its inverse dynamics to obtain a virtual reference from the measured output signal y.

The virtual signals are and

The optimal controller is obtained from noiseless data by solving the following optimization problem:

where the optimization function is given as follows:

See also

Related Research Articles

Linear elasticity is a mathematical model of how solid objects deform and become internally stressed by prescribed loading conditions. It is a simplification of the more general nonlinear theory of elasticity and a branch of continuum mechanics.

<span class="mw-page-title-main">Step response</span> Time behavior of a system controlled by Heaviside step functions

The step response of a system in a given initial state consists of the time evolution of its outputs when its control inputs are Heaviside step functions. In electronic engineering and control theory, step response is the time behaviour of the outputs of a general system when its inputs change from zero to one in a very short time. The concept can be extended to the abstract mathematical notion of a dynamical system using an evolution parameter.

In physics, the Hamilton–Jacobi equation, named after William Rowan Hamilton and Carl Gustav Jacob Jacobi, is an alternative formulation of classical mechanics, equivalent to other formulations such as Newton's laws of motion, Lagrangian mechanics and Hamiltonian mechanics.

<span class="mw-page-title-main">Large eddy simulation</span> Mathematical model for turbulence

Large eddy simulation (LES) is a mathematical model for turbulence used in computational fluid dynamics. It was initially proposed in 1963 by Joseph Smagorinsky to simulate atmospheric air currents, and first explored by Deardorff (1970). LES is currently applied in a wide variety of engineering applications, including combustion, acoustics, and simulations of the atmospheric boundary layer.

<span class="mw-page-title-main">Wigner distribution function</span>

The Wigner distribution function (WDF) is used in signal processing as a transform in time-frequency analysis.

<span class="mw-page-title-main">Lattice Boltzmann methods</span> Class of computational fluid dynamics methods

The lattice Boltzmann methods (LBM), originated from the lattice gas automata (LGA) method (Hardy-Pomeau-Pazzis and Frisch-Hasslacher-Pomeau models), is a class of computational fluid dynamics (CFD) methods for fluid simulation. Instead of solving the Navier–Stokes equations directly, a fluid density on a lattice is simulated with streaming and collision (relaxation) processes. The method is versatile as the model fluid can straightforwardly be made to mimic common fluid behaviour like vapour/liquid coexistence, and so fluid systems such as liquid droplets can be simulated. Also, fluids in complex environments such as porous media can be straightforwardly simulated, whereas with complex boundaries other CFD methods can be hard to work with.

In conformal geometry, the tractor bundle is a particular vector bundle constructed on a conformal manifold whose fibres form an effective representation of the conformal group.

The Poisson–Boltzmann equation describes the distribution of the electric potential in solution in the direction normal to a charged surface. This distribution is important to determine how the electrostatic interactions will affect the molecules in solution. The Poisson–Boltzmann equation is derived via mean-field assumptions. From the Poisson–Boltzmann equation many other equations have been derived with a number of different assumptions.

The Newman–Penrose (NP) formalism is a set of notation developed by Ezra T. Newman and Roger Penrose for general relativity (GR). Their notation is an effort to treat general relativity in terms of spinor notation, which introduces complex forms of the usual variables used in GR. The NP formalism is itself a special case of the tetrad formalism, where the tensors of the theory are projected onto a complete vector basis at each point in spacetime. Usually this vector basis is chosen to reflect some symmetry of the spacetime, leading to simplified expressions for physical observables. In the case of the NP formalism, the vector basis chosen is a null tetrad: a set of four null vectors—two real, and a complex-conjugate pair. The two real members often asymptotically point radially inward and radially outward, and the formalism is well adapted to treatment of the propagation of radiation in curved spacetime. The Weyl scalars, derived from the Weyl tensor, are often used. In particular, it can be shown that one of these scalars— in the appropriate frame—encodes the outgoing gravitational radiation of an asymptotically flat system.

In many-body theory, the term Green's function is sometimes used interchangeably with correlation function, but refers specifically to correlators of field operators or creation and annihilation operators.

<span class="mw-page-title-main">Critical state soil mechanics</span>

Critical state soil mechanics is the area of soil mechanics that encompasses the conceptual models representing the mechanical behavior of saturated remoulded soils based on the critical state concept. At the critical state, the relationship between forces applied in the soil (stress), and the resulting deformation resulting from this stress (strain) becomes constant. The soil will continue to deform, but the stress will no longer increase.

The Cauchy momentum equation is a vector partial differential equation put forth by Cauchy that describes the non-relativistic momentum transport in any continuum.

Laser linewidth is the spectral linewidth of a laser beam.

In the Newman–Penrose (NP) formalism of general relativity, independent components of the Ricci tensors of a four-dimensional spacetime are encoded into seven Ricci scalars which consist of three real scalars , three complex scalars and the NP curvature scalar . Physically, Ricci-NP scalars are related with the energy–momentum distribution of the spacetime due to Einstein's field equation.

<span class="mw-page-title-main">Moving load</span> Load that changes in time

In structural dynamics, a moving load changes the point at which the load is applied over time. Examples include a vehicle that travels across a bridge and a train moving along a track.

In orbital mechanics, Gauss's method is used for preliminary orbit determination from at least three observations of the orbiting body of interest at three different times. The required information are the times of observations, the position vectors of the observation points, the direction cosine vector of the orbiting body from the observation points and general physical data.

In astrophysics, the Chandrasekhar virial equations are a hierarchy of moment equations of the Euler equations, developed by the Indian American astrophysicist Subrahmanyan Chandrasekhar, and the physicist Enrico Fermi and Norman R. Lebovitz.

Tau functions are an important ingredient in the modern mathematical theory of integrable systems, and have numerous applications in a variety of other domains. They were originally introduced by Ryogo Hirota in his direct method approach to soliton equations, based on expressing them in an equivalent bilinear form.

In statistics, the Innovation Method provides an estimator for the parameters of stochastic differential equations given a time series of observations of the state variables. In the framework of continuous-discrete state space models, the innovation estimator is obtained by maximizing the log-likelihood of the corresponding discrete-time innovation process with respect to the parameters. The innovation estimator can be classified as a M-estimator, a quasi-maximum likelihood estimator or a prediction error estimator depending on the inferential considerations that want to be emphasized. The innovation method is a system identification technique for developing mathematical models of dynamical systems from measured data and for the optimal design of experiments.

Ball divergence is a non-parametric two-sample statistical test method in metric spaces. It measures the difference between two population probability distributions by integrating the difference over all balls in the space. Therefore, its value is zero if and only if the two probability measures are the same. Similar to common non-parametric test methods, ball divergence calculates the p-value through permutation tests.

References

  1. Bazanella, A.S., Campestrini, L., Eckhard, D. (2012). Data-driven controller design: the approach. Springer, ISBN   978-94-007-2300-9, 208 pages.
  2. Hjalmarsson, H., Gevers, M., Gunnarsson, S., & Lequin, O. (1998). Iterative feedback tuning: theory and applications. IEEE control systems, 18(4), 26–41.
  3. van Heusden, K., Karimi, A. and Bonvin, D. (2011), Data-driven model reference control with asymptotically guaranteed stability. Int. J. Adapt. Control Signal Process., 25: 331–351. doi : 10.1002/acs.1212
  4. Campi, Marco C., Andrea Lecchini, and Sergio M. Savaresi. "Virtual reference feedback tuning: a direct method for the design of feedback controllers." Automatica 38.8 (2002): 1337–1346.
  5. Formentin, S., Piga, D., Tóth, R., & Savaresi, S. M. (2016). Direct learning of LPV controllers from data. Automatica, 65, 98–110.
  6. Guardabassi, Guido O., and Sergio M. Savaresi. "Approximate feedback linearization of discrete-time non-linear systems using virtual input direct design." Systems & Control Letters 32.2 (1997): 63–74.

An Introduction to Data-Driven Control Systems Ali Khaki-Sedigh

ISBN: 978-1-394-19642-5 November 2023 Wiley-IEEE Press 384 Pages