An evolutionary computing technique for solving initial value problems in Ordinary Differential Equations is proposed in this paper. Neural network is used as a universal approximator while the adaptive parameters of neural networks are optimized by genetic algorithm. The solution is achieved on the continuous grid of time instead of discrete as in other numerical techniques. The comparison is ...

A fuzzy processor is programmed to provide an optimum output for solving a given problem. It could theoretically solve any problem (from a static point of view) if it is an universal approximator. This paper addresses the design of fuzzy processors aiming at a twofold objective: efficient adaptive approximation of different and even dynamically changing surfaces and hardware simplicity. Adequat...

This paper focuses on the development of a robust fuzzy sliding-mode scheme for controlling a vehicle motion system by continuously adjusting the brake torque, Fuzzy logic known for its properties of universal approximator and sliding mode control for its robustness in the presence of parameter variations and the disturbances are employed to control the wheel slip rate in emergency braking mane...

In this work we propose an approach for generalization in continuous domain Reinforcement Learning that, instead of using a single function approximator, tries many different function approximators in parallel, each one defined in a different region of the domain. Associated with each approximator is a relevance function that locally quantifies the quality of its approximation, so that, at each...

This paper proposes a methodology for analysis of the dynamic behavior of a robotic manipulator in continuous time. Initially this system (nonlinear system) will be decomposed into linear submodels and analyzed in the context of the Linear and Parameter Varying (LPV) Systems. The obtained linear submodels, which represent the local dynamic behavior of the robotic manipulator in some operating p...

The objective of transfer reinforcement learning is to generalize from a set of previous tasks to unseen new tasks. In this work, we focus on the transfer scenario where the dynamics among tasks are the same, but their goals differ. Although general value function (Sutton et al., 2011) has been shown to be useful for knowledge transfer, learning a universal value function can be challenging in ...

This paper proves that supervised learning algorithms used to train recurrent neural networks have an equilibrium point when the network implements a Maximum A Posteriori Probability (MAP) classiier. The result holds as a limit when the size of the training set goes to innnity. The result is general, since it stems as a property of cost minimizing algorithms, but to prove it we implicitly assum...

This paper connects two thoroughly investigated universal approximator techniques to each other. Recently, it has been shown that the input-output function of the general fuzzy KH interpolation method [1, 2] as well as its modification [3] are stable in the mathematical sense, or in other words, they can be considered as universal approximators with respect to the Lp (p ∈ [1,∞]) norm in the spa...

This study proposes a novel adaptive control approach using a feedforward Takagi-Sugeno (TS) fuzzy approximator for a class of highly unknown multi-input multi-output (MIMO) nonlinear plants. First of all, the design concept, namely, feedforward fuzzy approximator (FFA) based control, is introduced to compensate the unknown feedforward terms required during steady state via a forward TS fuzzy s...

With this paper we describe a well suited, scalable problem for reinforcement learning approaches in the field of mobile robots. We show a suitable representation of the problem for a reinforcement approach and present our results with a model based standard algorithm. Two different approximators for the value function are used, a grid based approximator and a neural network based approximator.