Purpose -The purpose of this paper is to generate a virtual knee angle reference to be followed by a knee prosthesis control, using an adaptive central pattern generator (CPG). Also, to study the feasibility of this approach to implement a continuous control strategy on the prosthesis. Design/methodology/approach -A CPG based on amplitude controlled phase oscillators (ACPOs) to track the current percentage of gait cycle on the prosthesis is proposed. Then, the virtual knee angle reference is generated along gait cycle, by interpolation with the corresponding angle of a sound knee. The structure and coupling of the CPG, as well as the control strategy are presented. Findings -The coupling of the CPG with real gait on the prosthesis was proven, regardless of gait speed. Also, it was found that the maximum knee angle reached during walking is proportional to gait speed. Finally, generation of virtual knee angle reference to be followed by a prosthesis is demonstrated.Research limitations/implications -As only one event detected along gait cycle was used to update the CPG phase, the response to gait speed changes might be slow. Updating the CPG with more events remains for a future work. Practical implications -The coupling of the CPG with real gait on the prosthesis results in a continuous gait cycle tracker, useful for any control strategy to be applied. Originality/value -It is the first time a bio-inspired concept as CPGs is applied to the prosthetic field. This could mean the beginning of a new era of cybernetic prostheses, which reproduce the lost limb and also the control functions of it.

No abstract

Velocity fields techniques are often used in the path planning and control problem in autonomous navigation. Velocity fields are compounded by vectors that can be expressed in polar form, i.e. Vx = iVl . cos(n) and Vy = IVI . sin(n). In the proposed technique, the problem is divided into two separate problems: the calculation of the orientation angle n, which was treated in a previous work, and the calculation of the linear velocity IVI. This paper addresses the calculation of IVI given a trajectory vector field that encodes n, thus generating the speed reference for a Velocity Field Controller. The linear velocity I V I is generated through a Mamdani-type Fuzzy Inference System, that considers the curl of the trajectory vector field, and an extension of the Voronoi Diagram. Results obtained in both simulations and tests with a real robot show that an adequate intelligent velocity behavior, and a smooth obstacle avoidance are achieved.

No abstract

Abstract. This paper describes a technique for statically stable gait synthesis for a quadruped robot using a simple Feed Forward Neural Networks (FFNN). A common approach for gait synthesis based on neural networks, is to use an implementation with Continuous Time Recurrent Neural Network (CTRNN) of arbitrary complex architecture as pattern generator for rhythmic limb motion. The preferred training method is implemented using genetic algorithms (GAs). However, to achieve the desired trajectory becomes an obstacle during the training process. This paper presents a much more simpler process converting a statically stable gait into actuator's space via inverse kinematics; the training of the network is done with those references. By doing so, the training problem becomes a spatio-temporal machine learning problem. It is described a solution for trajectory generation combining a simple oscillator model with a Multilayer Feedforward Neural Network (MFNN) to generate the desired trajectory.