Sharif Digital Repository

This book provides a comprehensive presentation of issues and challenges faced by researchers and practicing engineers in motion planning and hybrid control of dynamical legged locomotion. The major features range from offline and online motion planning algorithms to generate desired feasible periodic walking and running motions and tow-level control schemes, including within-stride feedback laws, continuous time update laws and event-based update laws, to asymptotically stabilize the generated desired periodic orbits. This book describes the current state of the art and future directions across all domains of dynamical legged locomotion so that readers can extend proposed motion planning... 

In this part, similar to part I of this paper, a new approach for Goal Coordination is introduced that is more convergent than the previously suggested methods. This is mainly because of using the gradient of interaction errors; difference between the actual interactions and the optimum calculated interaction values, to improve the coordination parameters. The proposed scheme extremely improves the convergence rate of the solution in compare to the classical methods. The significance and applicability of the proposed approach is shown in controlling the excitation voltage of two coupled synchronous machines described by six coupled non-linear differential equations. © 2005 IEEE  

Coordination strategies in large-scale systems are mainly based on two principles; Interaction Prediction Principle and Interaction Balance Principle. In this paper, which consists of two parts, the concept of coordination is introduced within the framework of two-level large-scale systems, and two new approaches for coordination based on Interaction Prediction Principle and Interaction Balance Principle have been presented. The new strategies have much faster convergence rate than the previously suggested classical methods. They extremely reduce the number of iterations required for obtaining the overall optimal solutions. The efficacy and advantage of the proposed approaches, in compare to... 

In this paper, the concept of coordination is introduced within the framework of two-level large-scale systems and a new approach based on Interaction Prediction Principle is presented. The proposed approach is formulated in an intelligent manner in such a way that it provides a new strategy that can be used to synthesize an on-line supervisory controller for the overall two-level large-scale systems, extendable to multi-level control systems. By using the new methodology which is based on using neural network for modeling each sub-system, typical gradient method for optimization of first-level sub-problems, and the gradient of the interaction prediction errors related to the predicted... 

Although there are many works in which path planning of robots is studied, but path planning of the bevel-tip needles with highly flexible body is different and difficult due to unique properties of soft tissues. Real soft tissues are nonhomogeneously elastic and uncertainly deformable and hence, during needle motions the planned path changes unknowingly. In this paper, a novel adaptive path planning of bevel-tip needles inside the uncertain brain tissue is presented. The proposed approach is based on minimization of a Lyapanov energy function used as the cost function which consists of 6 partial costs: path length, number of changes in bevel direction, tissue deformation, horizontal and... 

The canonical problems in control of the biped robots arise from underactuation, impulsive nature of the impact with the environment and existence of the many degrees of freedom in their mechanism. Since biped walkers have fewer actuators than degrees of freedom, they are underactuated mechanical systems. In this paper according to the humans and animals locomotion algorithms, the stability of an underactuated biped walker with point feet is done by Central Pattern Generator (CPG) and feedback networks. For tuning the parameters of the CPG network, the control problem is defined as an optimization problem. This optimization problem is solved by using of Genetic algorithm. Also a new feedback... 

The canonical problems in control of the biped robots arise from underactuation, impulsive nature of the impact with the environment and existence of the many degrees of freedom in their mechanism. Since biped walkers have fewer actuators than degrees of freedom, they are underactuated mechanical systems. In this paper according to the humans and animals locomotion algorithms, the stability of an underactuated biped walker with point feet is done by central pattern generator and feedback networks. For tuning the parameters of the CPG network, the control problem is defined as an optimization problem. This optimization problem is solved by using of Genetic algorithm. Also a new feedback... 

In this paper, a novel approach for tuning the parameters of fuzzy wavelet systems which are used for modeling of nonlinear and complex systems is proposed. In fuzzy inference system, each fuzzy rule is analogous to a wavelet basis function multiplied by a coefficient. Using clustering techniques, the center of these basis functions are located in the detected center of clusters. In this way, not only the approximation accuracy is increased, but also the number of unknown parameters is decreased. The feasibility of the proposed method is shown by modeling two highly nonlinear functions. The comparison of the results using the proposed approach, with the previous schemes, shows the... 

In this paper, we propose a decentralized control scheme for combined motion/force control of nonredundant multi-manipulator robotic systems, cooperatively grasping a rigid body in a prespecifled position/force trajectory. The proposed control scheme uses the impedance method for modeling the internal force exerted by the end effectors to the object. The precise mathematical analysis of the overall dynamical system and the proposed controller are presented and the stability of the overall system is studied. The proposed controller, then, is applied to a cooperative system composed of two PUMA560 manipulators and the simulation results are presented to verify the proposed control scheme  

In this paper, an adaptive multi-model sliding mode controller for robotic manipulators is presented. By using the multiple models technique, the nominal part of the control signal is constructed according to the most appropriate model at different environments. Adaptive single-input-single-output (SISO) fuzzy systems or radial basis function (RBF) neural networks, regarding their functional equivalence property, are used to approximate the discontinuous part of control signal; control gain, in a classical sliding mode controller. The key feature of this scheme is that prior knowledge of the system uncertainties is not required to guarantee the stability. Also the chattering phenomenon in... 

In this paper, a new approach for optimal control of robot manipulators is presented. For this purpose, the system is considered as a two-level large-scale system where a gradient-based coordination strategy is used to coordinate the overall system. This is achieved within a decomposition-coordination framework, where the robot manipulator is first decomposed into several sub-systems at the first level, and then, the coordination is done at the second level. The solution is based on Interaction Prediction Principle, where in the first level, the optimization is done by a gradient method and in the second level, the coordination is done by a new method based on the gradient of interaction... 

In this paper, a new non-linear D/A converter (DAC) is presented. This D/A converter uses neural network to approximate a sine function. It is targeted for direct digital frequency synthesizers in which the phase to sine amplitude conversion is usually performed using a ROM. Utilizing the proposed DAC, the power hungry ROM is removed. A design example is presented and the performance of the DAC is compared to the conventional method. The impact of non-idealities on the performance of the system is also investigated. © 2006 IEEE  

The main objective of this paper is to design a fast,non-expensive and practical controller for towercranes. The controllers are designed to transfer theloads from point to point, as fast as possible, and at thesame time the load swing is kept small during thetransfer process. Moreover, variations of the systemparameters such as cable length are taken intoconsideration. It is shown how the state feedbackcontrollers along with the gain scheduling could beused for a time varying linear model with varyingparameters. For this reason, linear controllers aredesigned for a number of operating points, where theirparameters are interpolated for conditions in between.In this way, a globally nonlinear... 

In this paper, a new approach for path generation and control of multi-agent systems is proposed. This method is based on global optimization techniques used for solving the inverse kinematic problem of redundant manipulators. Convergence of any performance function such as obstacle avoidance, collision avoidance, and heading angle to its global optimum is guaranteed by introducing a control law which is based on the Pontryagin's Maximum Principle. The efficacy of the proposed algorithm is demonstrated through simulation experiments. © 2006 IEEE  

In this paper a neural network decentralized control for trajectory tracking of robot manipulators is developed. The proposed decentralized control allows the overall closed-loop system to be stabilized while making the tracking error to be uniformly ultimately bounded (UUB), without having any prior knowledge of the robot manipulator dynamics. The interconnections in the dynamic equations of each subsystem are considered with unknown nonlinear bounds. The RBF neural networks (RBFNNs) are proposed to model the unknown nonlinear dynamics of the robots and the interconnection terms. Using Lyapunov method, the stability of the overall system is investigated, © 2005 IEEE