Nikola Georgiev Shakev

Comparative Results on Stabilization of the Quad-rotor Rotorcraft Using Bounded Feedback Controllers

In the recent years autonomous flying vehicles are being increasingly used in both civil and military areas. With the advancement of the technology it has become possible to test efficiently and cost-effectively different autonomous flight control concepts and design variations using small-scale aerial vehicles. In this paper the stabilization problem of the quad-rotor rotorcraft using bounded feedback controllers is investigated. Five different types of nonlinear feedback laws with saturation elements, previously proposed for global control of systems with multiple integrators, are applied and tested to control the quad-rotor rotorcraft roll and pitch angles. The results obtained from autonomous flight simulations and real time experiments with the Draganflyer V Ti four-rotor mini-rotorcraft are analyzed with respect to the structural simplicity of the control schemes and the transient performance of the closed-loop system.

Sliding mode incremental learning algorithm for interval type-2 Takagi–Sugeno–Kang fuzzy neural networks

Type-2 fuzzy logic systems are an area of growing interest over the last years. The ability to model uncertainties and to perform under noisy conditions in a better way than type-1 fuzzy logic systems increases their applicability. A new stable on-line learning algorithm for interval type-2 Takagi–Sugeno–Kang (TSK) fuzzy neural networks is proposed in this paper. Differently from the other recently proposed variable structure system theory-based on-line learning approaches for the type-2 TSK fuzzy neural nets, where the adopted consequent part of the fuzzy rules consists solely of a constant, the developed algorithm applies the complete structure of the Takagi–Sugeno type fuzzy if–then rule base (i.e. first order instead of zero order output function is implemented). In addition it is able to adapt the existing relation between the lower and the upper membership functions of the type-2 fuzzy systems. This allows managing of non-uniform uncertainties. Simulation results from the identification of a nonlinear system with uncertainties and a non-bounded-input bounded-output nonlinear plant with added output noise have demonstrated the better performance of the proposed algorithm in comparison with the previously reported in the literature sliding mode on-line learning algorithms for both type-1 and type-2 fuzzy neural structures.

Sliding mode algorithm for online learning in analog multilayer feedforward neural networks

A new dynamical sliding mode control algorithm is proposed for robust adaptive learning in analog multilayer feedforward networks with a scalar output. These type neural structures are widely used for modeling, identification and control of nonlinear dynamical systems. The zero level set of the learning error variable is considered as a sliding surface in the space of network learning parameters. The convergence of the algorithm is established and conditions are given. Its effectiveness is shown when applied to on-line learning of nonmonotonic function using a two-layered feedforward neural network.

Neural net tracking control of a mobile platform in robotized wireless sensor networks

In the recent years sensor networks have emerged as an effective tool for monitoring environments that are large in size and have complex topology. They have been successfully applied to solve various problems such as monitoring of different environmental indicators, detection of floods and fires, temperature control in office buildings, to collect information about the health status of hospitalized patients, to monitor the activity of some species and many others. Based-on the wireless communications the unification of the sensor modules can be implemented as a wireless sensor network (WSN). A newly observed trend is the inclusion of mobile robots into the WSN structure. The present paper describes a work-in-progress aiming to build laboratory prototype of a robotized hybrid WSN. The latter can be used for various distributed control experiments over the network structure and its efficiency. A robotized wireless sensor node has been designed and its functionality has been experimentally tested using an advanced trajectory tracking control algorithm. The proposed robotic node consists itself of a mobile robot platform integrated with a wireless sensor node. The latter has been implemented with a Tiva C Series TM4C1294NCPDT LaunchPad evaluation board equipped with TI Sensor Hub and WiFi Booster Pack. The mobile robot platform is based on the iRobot Create additionally upgraded with an on-board control system built in with gumstix verdex pro XL6P embedded microprocessor.

Trajectory control of unmanned aerial vehicle using neural nets with a stable learning algorithm

A neuro-adaptive trajectory control approach for unmanned aerial vehicles is proposed. The aerial robots altitude and latitude-longitude is controlled by three neuro-adaptive controllers that are used to track the desired altitude, airspeed and roll angle of the vehicle. Each intelligent control module consists of a conventional and a neural network feedback controller. The former is provided both to guarantee global asymptotic stability in compact space and as an inverse reference model of the response of the controlled system. Its output is used as an error signal by a stable on-line learning algorithm to update the parameters of the neurocontroller. In this way the latter is able to eliminate gradually the conventional controller from the control of the system. The proposed learning algorithm makes direct use of the variable structure systems theory and establishes a sliding motion in term of the neurocontroller parameters, leading the learning error toward zero. The performance of the proposed trajectory control scheme is evaluated with time based diagrams under MATLABs standard configuration and the Aeronautical Simulation Block Set.

A neuro-fuzzy adaptive sliding mode controller: Application to second-order chaotic system

To control complex dynamical systems, which are frequently coupled with unknown dynamics, modeling errors, nonlinearities, various sorts of disturbances, uncertainties and noise robust or model-free control methods should be employed. The features of a novel dynamical algorithm for robust adaptive learning in fuzzy rule-based neural networks of Takagi-Sugeno-Kang type with sigmoid membership functions and its application to the neuro-fuzzy adaptive nonlinear feedback control of systems with uncertain dynamics are presented. The proposed approach makes direct use of variable structure systems theory and the feedback-error-learning scheme. In the simulations, it has been tested on the control of Duffing oscillator and the analytical claims have been justified under the existence of uncertainty and large nonzero initial errors.

Trajectory control of manipulators using type-2 fuzzy neural friction and disturbance compensator

An incrementally tuned interval type-2 Takagi-Sugeno-Kang (TSK) fuzzy neural network implementing fuzzy if-then rule base with first order output functions is proposed for compensation of friction and disturbance effects during the trajectory tracking control of rigid robot manipulators. Friction and disturbances have an important influence on the robot manipulator dynamics. They are highly nonlinear terms that cannot be easily modeled. The proposed intelligent compensator makes use of a newly developed stable Variable Structure Systems theory-based on-line learning algorithm that is also able to adapt the existing relation between the lower and the upper membership functions of the type-2 fuzzy system. This allows managing of non-uniform uncertainties. Simulation results from the trajectory tracking control of two degrees of freedom RR planar robot manipulator using feedback linearization techniques and the proposed adaptive interval type-2 fuzzy neural compensator have shown that the joint positions are well controlled under wide variation of operation conditions and existing uncertainties.

An extended sliding mode learning algorithm for type-2 fuzzy neural networks

Type-2 fuzzy logic systems are an area of growing interest over the last years. The ability to model uncertainties in a better way than type-1 fuzzy logic systems increases their applicability. A new stable on-line learning algorithm for type-2 fuzzy neural networks is proposed in this paper. It can be considered as an extended version of the recently developed on-line learning approaches for type-2 fuzzy neural networks based on the Variable Structure System theory concepts. Simulation results from the identification of a nonlinear system with uncertainties have demonstrated the better performance of the proposed extended algorithm in comparison with the previously reported in the literature sliding mode learning algorithms for both type-1 and type-2 fuzzy neural structures.

Sliding Mode Algorithm for On-Line Learning in Fuzzy Rule-Based Neural Networks

Abstract A new, variable structure systems theory based, algorithm has been developed for on-line training of fuzzy-neural networks. Such computationally intelligent structures are widely used for modeling, identification and control of nonlinear dynamic systems. The algorithm is applicable to fuzzy rule-based neural nets of Takagi-Sugeno-Kang type with a scalar output. Its convergence is established and the conditions are given. Differently from other similar approaches which are limited to the adaptation of the parameters of the network defuzzification part only, the proposed algorithm tunes also the parameters of the implemented membership functions. The zero level set of the learning error variable is considered as a sliding surface in the space of network learning parameters. The effectiveness of the proposed algorithm is shown when applied to on-line learning of nonlinear functions approximation.

Neural network identification, predictive modeling and control with a sliding mode learning mechanism: an application to the robotic manipulators

The features of a novel adaptive PID-like neurocontrol scheme for nonlinear plants are presented. The controller tuning is based on an estimate of the command-error determined via one-step-ahead neural predictive model of the plant. An on-line learning sliding mode algorithm is applied to the model and to the controller as well. The control architecture developed has been simulated and its effect on the trajectory tracking performance of a simple two-degree-of-freedom robot manipulator has been evaluated. The results show that both learning structures, the neural predictive model and the controller, inherit some of the advantages of SMC: high speed of learning and robustness.