---

In the present article, an improved Learning Classifier Systems (LCS) is proposed to control the balance of a moving unmanned bicycle. Significant characteristics of learning classifier systems is that they can learn through a set of system actions in the real world (similar to intelligent creatures) while no dynamic model of the system is needed. Contrary to studies reported in the literature where action domain of the controller is discrete and accordingly such controller cannot be used in real world applications, in the present study efficacy of the classifier system is enhanced by definition of continuous domain for the outputs, and then is used to control the balance of unmanned bicycle. A scheme based upon fuzzy membership functions is proposed which makes it possible for the domain of actions to be continuous. The proposed LCS features a dynamic reward assignment mechanism which is invented to cope with the bicycle’s delayed response due to its mass inertias. This allows the rapid calculation of the reward and hence enables the controller to be used in such real time applications as the balance control of unmanned vehicles. A standard 2 degree of freedom (2-DOF) bicycle model is employed to demonstrate the efficiency of the enhanced LCS. Simulation results show that the proposed classifier system outperforms traditional classifier system as well as some of the more common balance-control strategies reported in the literature.

---

Due to necessity of increasing performance in new generations of the humanoid robots, in this paper, a novel power transmission mechanism to actuate the ankle joint of a humanoid robot is presented in order to increase the motion speed of SURENAIII humanoid robot. Also, the energy consumption of the proposed and the previous mechanisms are studied. In the proposed mechanism, the actuators of the ankle joint are located in the shank link. Then, a combined timing belt-pulley and a harmonic drive module are exploited for power transmission for the pitch joint. Also, the roll joint drive has employed a roller screw. In order to validate the design procedure, the simulation results of the robot are compared with the experimental data. The results reveal that the dynamic model is fairly matched to the real behavior of the robot. Also, the revolutionary genetic algorithm is employed to optimize the effective path planning parameters with respect to the minimum knee joint torque. This optimization procedure which is employed in robot walking on flat terrains consist of straight motion, ensures the robot's stability. As a result, the optimal path planning parameters for proposed mechanism are obtained in such a way that has decreased the actuating torques of lower-body of SURENAIII. Also, the proposed mechanism can achieve using lighter motors and getting the robot faster by means of mass reduction of foot.

---

In this paper a push recovery controller for balancing humanoid robot under severe pushes for situation that contact surface is small is presented. Human response to progressively increasing disturbances can be categorized into three strategies: ankle strategy, hip strategy and stepping strategy. The reaction of human to external disturbances in the situations that contact surface is small or stepping is not possible is generating upper body angular momentum. In this way in this paper, a single model predictive controller scheme is employed to controlling the capture point by modulating zero moment point and centroidal moment pivot. The proposed algorithm is capable of recovering balance of humanoid robot under severe pushes without stepping in situation that contact surface is shrunked to a strip. The goal of the proposed controller is to control the capture point, employing the centroidal moment pivot when the capture point is out of the support polygon, and/or the zero moment point when the capture point is inside the support polygon. The merit of proposed algorithm is shown successfully in different simulation scenarios using characteristic of SURENA III humanoid robot.

---

“Tensegrity” refers to a class of discrete structures with two force members (bars and cables) wherein bars only take tensile loads and cables only take compressive loads. The pre stressed members are interconnected so as to form a self equilibrium structure. Compared to a truss supporting the same external loading, a tensegrity structure has fewer members and could weigh less. Determining the stable topology (member connectivities), form (node coordinates) and size (cross sectional areas of members) of a tensegrity structure for weight minimization is a challenging task, as the governing equations are nonlinear and the conventional matrix analysis methods cannot be used. This article addresses the weight minimization of a class one tensegrity structure with a given number of bars and cables, anchored at certain nodes and supporting given load(s) at certain node(s). Member connectivities and their cross sectional areas and force densities are taken as design variables, whereas the members’ strength and buckling requirements and maximum nodal displacements constitute the constraints, along with the coordinates of the floating nodes to make the structure symmetric. Constraints are evaluated through the nonlinear shape design of the self equilibrium structure and the linear analysis of the loaded structure, assuming small displacements. Using a novel approach, optimization is simultaneously performed in multiple promising areas of the solution space, resulting in multiple, optimum solutions. The diversity of the solutions is demonstrated by applying the proposed approach to a number of structural design problem.