Proposed framework for applying adaptive critics in real-time realm

Lendaris, George G.; Santiago, R.A.; Carroll, Michael S.

doi:10.1109/ijcnn.2002.1007791

Cited by 9 publications

(12 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The control input can be calculated using Equation (34) once the critic network and system state density models become available. The implementation of this two stage optimisation method can be performed efficiently by utilising the modular approach constituting of functional modules and algorithmic modules (Ferrari and Stengel, 2004;Lendaris et al, 2002;Herzallah, 2007). The main functional modules are the action and critic networks.…”

Section: Generalised Nonlinear Probabilistic Control Algorithmmentioning

confidence: 99%

Fully probabilistic control for stochastic nonlinear control systems with input dependent noise

Herzallah

2015

Neural Networks

View full text Add to dashboard Cite

Robust controllers for nonlinear stochastic systems with functional uncertainties can be consistently designed using probabilistic control methods. In this paper a generalised probabilistic controller design for the minimisation of the Kullback-Leibler divergence between the actual joint probability density function (pdf) of the closed loop control system, and an ideal joint pdf is presented emphasising how the uncertainty can be systematically incorporated in the absence of reliable systems models. To achieve this objective all probabilistic models of the system are estimated from process data using mixture density networks (MDNs) where all the parameters of the estimated pdfs are taken to be state and control input dependent. Based on this dependency of the density parameters on the input values, explicit formulations to the construction of optimal generalised probabilistic controllers are obtained through the techniques of dynamic programming and adaptive critic methods. Using the proposed generalised probabilistic controller, the conditional joint pdfs can be made to follow the ideal ones. A simulation example is used to demonstrate the implementation of the algorithm and encouraging results are obtained.

show abstract

Section: Generalised Nonlinear Probabilistic Control Algorithmmentioning

confidence: 99%

Fully probabilistic control for stochastic nonlinear control systems with input dependent noise

Herzallah

2015

Neural Networks

View full text Add to dashboard Cite

show abstract

“…Similarly, the control law can be calculated from (35) once the critic network and system dynamic models become available. Although the probabilistic DHP adaptive critic approach involves multiple computation levels, its implementation can be made by means of modular approach constituting of functional modules and algorithmic modules [8], [20], [12]. The key functional modules are the action and critic networks.…”

Section: Nonlinear Randomized Control Algorithm Based On Probabilmentioning

confidence: 99%

“…The proof of convergence given in [16], [8] is directly applicable to the probabilistic adaptive critic design in this paper. A nonlinear control problem example to demonstrate the convergence of the proposed probabilistic critic network is given in Section V. Further discussion on the convergence and speed of convergence of adaptive critic designs can be found in [20]. Moreover, empirical evidence on the convergence of the adaptive critic design can be found in [3], [11], [19], [21].…”

Section: Nonlinear Randomized Control Algorithm Based On Probabilmentioning

confidence: 99%

Probabilistic DHP adaptive critic for nonlinear stochastic control systems

Herzallah

2013

Neural Networks

View full text Add to dashboard Cite

Following the recently developed algorithms for fully probabilistic control design for general dynamic stochastic systems [15], [18], this paper presents the solution to the probabilistic dual heuristic programming (DHP) adaptive critic method [15] and randomized control algorithm for stochastic nonlinear dynamical systems. The purpose of the randomized control input design is to make the joint probability density function of the closed loop system as close as possible to a predetermined ideal joint probability density function. This paper completes the previous work [15], [18] by formulating and solving the fully probabilistic control design problem on the more general case of nonlinear stochastic discrete time systems. A simulated example is used to demonstrate the use of the algorithm and encouraging results have been obtained. Index Termsnonlinear stochastic systems; fully probabilistic design; nonlinear randomized control input design; adaptive critic.

show abstract

“…In [81,185] a RFC for the NASA F-18/HARV based on a QFT compensator and an adaptive filter is used. Flight control based on reinforcement learning is the subject of [89,90,126]. Indirect adaptive control using a moving window/batch estimation for partial loss of the horizontal tail surface is studied in [157].…”

Section: Adaptive Controlmentioning

confidence: 99%

Adaptive Backstepping Flight Control for Modern Fighter Aircraft

Sonneveldt¹

2011

Advances in Flight Control Systems

View full text Add to dashboard Cite

Proposed framework for applying adaptive critics in real-time realm

Cited by 9 publications

References 30 publications

Fully probabilistic control for stochastic nonlinear control systems with input dependent noise

Fully probabilistic control for stochastic nonlinear control systems with input dependent noise

Probabilistic DHP adaptive critic for nonlinear stochastic control systems

Adaptive Backstepping Flight Control for Modern Fighter Aircraft

Contact Info

Product

Resources

About