scholarly journals LQG Online Learning

2017 ◽  
Vol 29 (8) ◽  
pp. 2203-2291 ◽  
Author(s):  
Giorgio Gnecco ◽  
Alberto Bemporad ◽  
Marco Gori ◽  
Marcello Sanguineti
Keyword(s):  
Optimal Control ◽  
Online Learning ◽  
Nonlinear Models ◽  
Machine Learning Techniques ◽  
Parameter Vector ◽  
Linear Quadratic ◽  
Linear Quadratic Gaussian ◽  
Learning Framework ◽  
Learning Techniques ◽  
Lqg Optimal Control

Optimal control theory and machine learning techniques are combined to formulate and solve in closed form an optimal control formulation of online learning from supervised examples with regularization of the updates. The connections with the classical linear quadratic gaussian (LQG) optimal control problem, of which the proposed learning paradigm is a nontrivial variation as it involves random matrices, are investigated. The obtained optimal solutions are compared with the Kalman filter estimate of the parameter vector to be learned. It is shown that the proposed algorithm is less sensitive to outliers with respect to the Kalman estimate (thanks to the presence of the regularization term), thus providing smoother estimates with respect to time. The basic formulation of the proposed online learning framework refers to a discrete-time setting with a finite learning horizon and a linear model. Various extensions are investigated, including the infinite learning horizon and, via the so-called kernel trick, the case of nonlinear models.

Linear Quadratic Gaussian Optimal Control for Human Eye Aberration Correction

2014 ◽  
Vol 51 (4) ◽  
pp. 041102
Author(s):  
王波 Wang Bo ◽  
钮赛赛 Niu Saisai ◽  
吴卫明 Wu Weiming
Keyword(s):  
Optimal Control ◽  
Aberration Correction ◽  
Linear Quadratic ◽  
Linear Quadratic Gaussian ◽  
Human Eye

Kalman Filter Control Embedded into the Reinforcement Learning Framework

2004 ◽  
Vol 16 (3) ◽  
pp. 491-499 ◽  
Author(s):  
István Szita ◽  
András Lőrincz
Keyword(s):  
Optimal Control ◽  
Kalman Filter ◽  
Reinforcement Learning ◽  
Learning Rule ◽  
Slight Modification ◽  
Linear Quadratic ◽  
Filter Model ◽  
Learning Framework ◽  
Value Estimation ◽  
On Line

There is a growing interest in using Kalman filter models in brain modeling. The question arises whether Kalman filter models can be used on-line not only for estimation but for control. The usual method of optimal control of Kalman filter makes use of off-line backward recursion, which is not satisfactory for this purpose. Here, it is shown that a slight modification of the linear-quadratic-gaussian Kalman filter model allows the on-line estimation of optimal control by using reinforcement learning and overcomes this difficulty. Moreover, the emerging learning rule for value estimation exhibits a Hebbian form, which is weighted by the error of the value estimation.

Risk-sensitive linear/quadratic/gaussian control

1981 ◽  
Vol 13 (4) ◽  
pp. 764-777 ◽  
Author(s):  
P. Whittle
Keyword(s):  
Optimal Control ◽  
Kalman Filter ◽  
Risk Aversion ◽  
Separation Theorem ◽  
Linear Quadratic ◽  
Criterion Function ◽  
Linear Quadratic Gaussian ◽  
State Estimate ◽  
Certainty Equivalence ◽  
Risk Sensitive

The conventional linear/quadratic/Gaussian assumptions are modified in that minimisation of the expectation of cost G defined by (2) is replaced by minimisation of the criterion function (5). The scalar –θ is a measure of risk-aversion. It is shown that modified versions of certainty equivalence and the separation theorem still hold, that optimal control is still linear Markov, and state estimate generated by a version of the Kalman filter. There are also various new features, remarked upon in Sections 5 and 7. The paper generalises earlier work of Jacobson.

scholarly journals Delivering Precision Medicine to Patients with Spinal Cord Disorders; Insights into Applications of Bioinformatics and Machine Learning from Studies of Degenerative Cervical Myelopathy

2021 ◽  
Author(s):  
Kalum J. Ost ◽  
David W. Anderson ◽  
David W. Cadotte
Keyword(s):  
Machine Learning ◽  
Precision Medicine ◽  
New Technologies ◽  
Machine Learning Techniques ◽  
Massive Datasets ◽  
Learning Framework ◽  
Learning Techniques ◽  
Machine Learning Approach ◽  
Spinal Cord Disorders ◽  
Degenerative Cervical Myelopathy

With the common adoption of electronic health records and new technologies capable of producing an unprecedented scale of data, a shift must occur in how we practice medicine in order to utilize these resources. We are entering an era in which the capacity of even the most clever human doctor simply is insufficient. As such, realizing “personalized” or “precision” medicine requires new methods that can leverage the massive amounts of data now available. Machine learning techniques provide one important toolkit in this venture, as they are fundamentally designed to deal with (and, in fact, benefit from) massive datasets. The clinical applications for such machine learning systems are still in their infancy, however, and the field of medicine presents a unique set of design considerations. In this chapter, we will walk through how we selected and adjusted the “Progressive Learning framework” to account for these considerations in the case of Degenerative Cervical Myeolopathy. We additionally compare a model designed with these techniques to similar static models run in “perfect world” scenarios (free of the clinical issues address), and we use simulated clinical data acquisition scenarios to demonstrate the advantages of our machine learning approach in providing personalized diagnoses.

Microclimate control of a greenhouse by adaptive Generalized Linear Quadratic strategy

2018 ◽  
Vol 11 (1) ◽  
pp. 377 ◽  
Author(s):  
Mohamed Essahafi ◽  
Mustapha Ait Lafkih
Keyword(s):  
Optimal Control ◽  
Least Squares ◽  
Recursive Least Squares ◽  
Linear Quadratic ◽  
Linear Quadratic Gaussian ◽  
Advanced Technique ◽  
Online Identification ◽  
State Models ◽  
The Cost ◽  
And Control

<p>To highlight the conceptual aspects related to the implementation of techniques optimal control in the form state, we present in this paper, the identification and control of the temperature and humidity of the air inside a greenhouse. Using respectively an online identification based on the recursive least squares with forgotten Factor method and the multivariable adaptive linear quadratic Gaussian approach which the advanced technique (LQG) is presented.  The design of this controller parameters is based on state models identified directly from measured greenhouse data. hence the performances of the controller developed are illustrated by different tests and simulations on identified models of a greenhouse. Discussions on the results obtained are then processed in the paper to show the effectiveness of the controller in terms of stability and optimization of the cost of control.</p>

LQG Control for Nonstandard Singularly Perturbed Discrete-Time Systems

2004 ◽  
Vol 126 (4) ◽  
pp. 860-864 ◽  
Author(s):  
Beom-Soo Kim ◽  
Young-Joong Kim ◽  
Myo-Taeg Lim
Keyword(s):  
Optimal Control ◽  
Optimal Control Problem ◽  
Control Problem ◽  
Discrete Time ◽  
Singularly Perturbed ◽  
Linear Quadratic ◽  
Linear Quadratic Gaussian ◽  
Reduced Order ◽  
Discrete Time Systems ◽  
Time Systems

In this paper we present a control method and a high accuracy solution technique in solving the linear quadratic Gaussian problems for nonstandard singularly perturbed discrete time systems. The methodology that exists in the literature for the solution of the standard singularly perturbed discrete time linear quadratic Gaussian optimal control problem cannot be extended to the corresponding nonstandard counterpart. The solution of the linear quadratic Gaussian optimal control problem is obtained by solving the pure-slow and pure-fast reduced-order continuous-time algebraic Riccati equations and by implementing the pure-slow and pure-fast reduced-order Kalman filters. In order to show the effectiveness of the proposed method, we present the numerical result for a one-link flexible robot arm.

scholarly journals Temperature Control of Stirred Tank Heater using Optimal Control Technique

2020 ◽  
Author(s):  
Mustefa Jibril ◽  
Messay Tadese ◽  
Eliyas Alemayehu
Keyword(s):  
Optimal Control ◽  
Temperature Control ◽  
Open Loop ◽  
Loop System ◽  
Stirred Tank ◽  
Control Technique ◽  
Closed Loop System ◽  
Linear Quadratic ◽  
Linear Quadratic Gaussian ◽  
Quadratic Integral

This paper presents the application of optimal control problem in modeling of stirred tank heater temperature control. The analysis of the open loop system shows that the system is not efficient without a controller. Linear Quadratic Gaussian (LQG) and Linear Quadratic Integral (LQI) controllers are used to increase the performance of the system. Comparison of the closed loop system with the proposed controllers have been done with Matlab/Simulink Toolbox and a promising results have been analyzed.

Sign in / Sign up

Export Citation Format

Share Document