Showing 1–4 of 4 results for author: Nayyer, S S

Unified Control Framework: A Novel Perspective on Constrained Optimization, Optimization-based Control, and Parameter Estimation

Authors: Revati Gunjal, Syed Shadab Nayyer, Sushama Wagh, Navdeep Singh

Abstract: A common theme in all the above areas is designing a dynamical system to accomplish desired objectives, possibly in some predefined optimal way. Since control theory advances the idea of suitably modifying the behavior of a dynamical system, this paper explores the role of control theory in designing efficient algorithms (or dynamical systems) related to problems surrounding the optimization frame… ▽ More A common theme in all the above areas is designing a dynamical system to accomplish desired objectives, possibly in some predefined optimal way. Since control theory advances the idea of suitably modifying the behavior of a dynamical system, this paper explores the role of control theory in designing efficient algorithms (or dynamical systems) related to problems surrounding the optimization framework, including constrained optimization, optimization-based control, and parameter estimation. This amalgamation of control theory with the above-mentioned areas has been made possible by the recently introduced paradigm of Passivity and Immersion (P\&I) based control. The generality and working of P\&I, as compared to the existing approaches in control theory, are best introduced through the example presented below. △ Less

Submitted 2 July, 2024; v1 submitted 30 June, 2024; originally announced July 2024.

A New Perspective of Accelerated Gradient Methods: The Controlled Invariant Manifold Approach

Authors: Revati Gunjal, Sushama Wagh, Syed Shadab Nayyer, Alex Stankovic, Navdeep M. Singh

Abstract: Gradient Descent (GD) is a ubiquitous algorithm for finding the optimal solution to an optimization problem. For reduced computational complexity, the optimal solution $\mathrm{x^*}$ of the optimization problem must be attained in a minimum number of iterations. For this objective, the paper proposes a genesis of an accelerated gradient algorithm through the controlled dynamical system perspective… ▽ More Gradient Descent (GD) is a ubiquitous algorithm for finding the optimal solution to an optimization problem. For reduced computational complexity, the optimal solution $\mathrm{x^*}$ of the optimization problem must be attained in a minimum number of iterations. For this objective, the paper proposes a genesis of an accelerated gradient algorithm through the controlled dynamical system perspective. The objective of optimally reaching the optimal solution $\mathrm{x^*}$ where $\mathrm{\nabla f(x^*)=0}$ with a given initial condition $\mathrm{x(0)}$ is achieved through control. △ Less

Submitted 31 May, 2023; v1 submitted 18 May, 2023; originally announced May 2023.

Passivity and Immersion based-modified gradient estimator: A control perspective in parameter estimation

Authors: Syed Shadab Nayyer, G. Revati, S. R. Wagh, N. M. Singh

Abstract: In this paper, a constructive and systematic strategy with more apparent degrees of freedom to achieve the accurate estimation of unknown parameters via a control perspective is proposed. By adding a virtual control in the final equation of the gradient dynamics, the Gradient Estimator (GE) and Memory Regressor and Extension (MRE) approaches are extended. The solution of the virtual control law is… ▽ More In this paper, a constructive and systematic strategy with more apparent degrees of freedom to achieve the accurate estimation of unknown parameters via a control perspective is proposed. By adding a virtual control in the final equation of the gradient dynamics, the Gradient Estimator (GE) and Memory Regressor and Extension (MRE) approaches are extended. The solution of the virtual control law is identified by the P&I approach. The P&I approach is based on the choice of an appropriate implicit manifold and the generation of a suitable passive output and a related storage function. This facilitates the virtual control law being obtained in a way that the parametric error converges asymptotically to zero. Because the above ideas connect with the P&I approach and GE, the developed methodology is labeled the passivity and immersion-based modified gradient estimator (MGE). The proposed P&I-based modified gradient estimator is extended via the MRE approach. This modification provides improved transient response and fast convergence. Based on certain PE and non-PE examples, a comparative analysis is carried out to show the efficacy of the proposed approaches. △ Less

Submitted 19 November, 2022; originally announced November 2022.

Towards a Constructive Framework for Stabilization and Control of Nonlinear Systems: Passivity and Immersion (P\&I) Approach

Authors: Syed Shadab Nayyer, Sushama R. Wagh, Navdeep M. Singh

Abstract: The varied and complex dynamics of real-world systems challenge the formulation of a systematic strategy for designing a stabilizing feedback law. Rather than taking a universal approach, the control strategies developed thus far to handle this problem are specific to the inherent structure of the system under consideration. Therefore, this paper attempts to develop a generalized theory for the de… ▽ More The varied and complex dynamics of real-world systems challenge the formulation of a systematic strategy for designing a stabilizing feedback law. Rather than taking a universal approach, the control strategies developed thus far to handle this problem are specific to the inherent structure of the system under consideration. Therefore, this paper attempts to develop a generalized theory for the design of the stabilizing feedback law wherever possible for a general class of systems, including the systems in standard structured and unstructured forms discussed in the existing literature. The theory behind this general controller design theory utilizes the idea of an invariant target manifold giving rise to a non-degenerate two form, through which the definition of certain passive outputs and storage functions leads to a generation of control law for stabilizing the system. Because the above concepts are linked with the Immersion and Invariance (I&I) design policy and the passivity theory of controller design, the proposed methodology is labeled as the "Passivity and Immersion (P&I) approach". Furthermore, being a constructive methodology, the various worked examples in this paper exemplify and demonstrate the various design paradigms such as Backstep**, Incremental Backstep**, Forwarding, I&I, and the techniques based on the generation of Control Contraction Metrics (CCM) can be unified in the P&I methodology. △ Less

Submitted 11 December, 2022; v1 submitted 22 August, 2022; originally announced August 2022.