This paper presents a general control architecture for bipedal walking which is based on a divide-and-conquer approach. Based on the architecture, the sagittal-plane motion-control algorithm is formulated using a control approach known as Virtual Model Control. A reinforcment learning algorithm is used to learn the key parameter of the swing leg control task so that speed control can be achieved. The control algorithm is applied to two simulated bipedal robots. The simulation analyses demonstrate that the local speed control mechanism based on the stance ankle is effective in reducing the learning time. The algorithm is also demonstrated to be general in that it is applicable across bipedal robots that have different length and mass parameters.

Parametric formulation is a systematic method that ties the constraint functions to the resulting joint profiles for bipedal robots. Acceptable joint profiles for continuous and steady walking must satisfy the additional condition of being repeatable. This extra condition imposes restrictions on the selection of constraint functions, gait parameters and the initial posture of each step, and can make the synthesis of joint profiles extremely challenging. This paper presents a correlation between the gait parameters used in a class of specific constraint functions and the initial posture that will create repeatable gait. Furthermore, a general necessary and sufficient condition for repeatable gait is proposed. This condition provides a guideline for selecting constraint functions and their associated gait parameters in the context of producing repeatable gait. This research can not only provide valuable insights into bipedal walking, but also serve as a stepping stone for employing parametric formulation with various constraint functions.

A software tool for data mining (DMiner-I) is introduced, which integrates pattern recognition (PCA, Fisher, clustering, HyperEnvelop, regression), artificial intelligence (knowledge representation, decision trees), statistical learning (rough set, support vector machine), and computational intelligence (neural network, genetic algorithm, fuzzy systems). It consists of nine function models: pattern recognition, decision trees, association rule, fuzzy rule, neural network, genetic algorithm, HyperEnvelop, support vector machine and visualization. The principle, algorithms and knowledge representation of some function models of data mining are described. Nonmonotony in data mining is dealt with by concept hierarchy and layered mining. The software tool of data mining is realized by Visual C++ under Windows 2000. The software tool of data mining has been satisfactorily applied in the prediction of regularities of the formation of ternary intermetallic compounds in alloy systems, and diagnosis of brain glioma.

This paper discusses a technique for identifying the joint parameters of a modular robot in order to study the dynamic characteristics of the whole structure and to realise dynamic control. A method for identifying the joint parameters of the structure applying fuzzy logic combined with a genetic algorithm has been studied using a 9-DOF modular redundant robot. A Genetic Algorithm was used in the fuzzy optimisation, which helped to avoid converging to locally optimal solutions and made the results identified much more reasonable. The joint parameters of a 9-DOF modular redundant robot have been identified.

Planar parallel manipulators, with potential applications in high speed, high acceleration tasks such as electronic component placement, would be subject to mechanical vibration due to high inertial forces acting on the linkages and other components. To achieve high throughput capability, such motion induced vibration would have to be damped quickly, to reduce settling time of the platform position and orientation. This paper develops a two-time scale dynamic model of a three-degree-of-freedom planar parallel manipulator with structurally flexible linkages. Based on the two-time scale model, a composite controller, consisting of a computed torque controller for the slow time-scale or rigid body subsystem dynamics, and a linear-quadratic state-feedback regulator for the fast time-scale flexible dynamic subsystem, is designed. Simulation results show that the composite control scheme permits the parallel manipulator platform to follow a given desired trajectory, while damping structural vibration arising due to excitation from inertial forces.

This paper proposes a new orientation representation of planar manipulators by resorting to polar coordinates. Connecting the end-effector point to the first joint of a manipulator with a virtual adjustable link, the length of the adjustable link corresponds to a workspace point and is related to the orientation of the end effector link by a virtual angle in the form of a transcendental equation. Plotting this link length against the virtual angle in polar coordinates, the orientation of a manipulator can be represented in a compact form, and the range of partial dexterity can be identified. The characteristics of the new representation is hence revealed and related to the assessment of an orientation capability of a manipulator. Based on this representation, a desirable task can be presented and a manipulator can be synthesised with the required orientation.

A conservatism-reduced design of a gain scheduled output feedback H∞ controller for an n-joint rigid robotic manipulator, which integrates the varying-parameter rate without their feedback, is proposed. The robotic system is reduced to a 1inear parameter varying (LPV) form, which depends on the varying-parameter. By using a parameter-dependent Lyapunov function, the design of a controller, which satisfies the closed-loop H∞ performance, is reduced to a solution of the parameterized linear matrix inequalities (LMIs) of parameter matrices. With a use of the concept of “multi-convexity”, the solution of the infinite LMIs in the varying-parameter and its rate space is reduced to a solution of the finite LMIs for the vertex set. The proposed controller eliminates the feedback of the varying-parameter rate and fixes its upper boundary so that the conservatism of the controller design is reduced. Experimental results verify the effectiveness of the proposed design.

In this paper, a lumped parameter model of a robotic fibre placement system consisting of a Motoman SK-120 robot, a force/torque sensor, a pneumatic actuator and a stiff workpiece holder is developed and experimentally verified for the purpose of predicting and characterising the dynamic behaviour of the fibre placement system. Special attention has been given to the dynamics of the actuator which is represented as a mass confined to move between two non-linear springs and dampers. The overall model containing manipulator, force sensor, pneumatic actuator and the workpiece holder dynamics is of the tenth order. Step response experiments were conducted to verify the model and to determine the approximate values of the parameters in the mathematical model. The results prove that the established model is accurate enough to explain the dynamic behaviour of the fibre placement system and it can be employed to quantify the influence of the dynamics of the pneumatic actuator on the constant force-based fibre placement. The well-known fact that the dynamics of the pneumatic actuator varies with the piston position has also been experimentally demonstrated.

This paper presents a modified input shaping method to reduce the motion-induced vibration of a linear time-varying system after a rest-to-rest motion. Shaping parameters were obtained using the concept of modal-filtered impulse response. The conventional shaping method can be said just a special case of the proposed shaping method. The effectiveness of this proposing method was checked using some examples of both moderate and considerably fast time-varying systems. With a rest-to-rest motion control of a two-link flexible manipulator, this study also demonstrates that this method can be expanded to nonlinear cases.

This article focuses on research at the MEC, giving three examples of areas of concerted activities at the Centre: intelligent manufacturing, concurrent engineering and rapid manufacturing.