This chapter presents the theoretical foundations and methodology to develop a bioinspired hybrid control architecture for a biped robotic device that reproduces gait and human motor control strategies with the ability to adapt the trajectory to environmental conditions. The objective is to design robotic devices (such as exoskeletons), through the functional integration of hybrid dynamic system modeling (event-driven and continuous dynamics) with efficient and robust conventional control techniques and bioinspired control algorithms, with a near-natural human gait pattern. The human gait cycle is modeled as a hybrid dynamic using a finite state machine (FSM). The gait trajectories are to be generated in such a way that they will be capable of adapting to disturbances in the path followed by the robotic device; this will be achieved using a neuronal oscillator that simulates the behavior of a central pattern generator (CPG).
Top1. Introduction
Anthropomorphic robotics is an interdisciplinary field of study that integrates concepts of biomechanics, anatomy and human cognition, together with traditional robotic areas such as mechatronics and control theory. Current interest in this field presents attractive challenges, particularly in bipedal robotics, to design robots with a walking pattern that is as close to natural dynamics as possible.
In general, considering an application, for example, a lower limb exoskeleton, the device is required to operate in parallel to the human body or to mimic with high precision its movements, which generates the need to consider control techniques to achieve these goals.
Control in bipedal robots is challenging because they are performed interacting with dynamic environment and exhibit time variant dynamics. Bipedal robot locomotion requires the development of new technologies and more sophisticated techniques to overcome wide range of issues, such as motion control, torque control, stability, trajectory generation, generation of different gait patterns depending on the environment conditions, sensing and perception (Habib, Liu, Watanabe, & Izumi, 2007). The design of biped robots has traditionally been based on constant torque control of the joints with a local stability criterion, usually based on the Zero Moment Point (ZMP) approach, which lead to patterns of slow and jerky movement, and therefore do not reproduce the trajectories of joints exhibited in normal gait, the which makes the non-biomimetic gait (Batista, 2017).
Anthropomorphic robots can theoretically operate in many human environments, and to take advantage of their potential, robots must be able to operate and navigate in complex and unstructured environments. This requires a flexible walking system that can adapt quickly to new situations (Buschmann, Ewald, Ulbrich, & Buschges, 2012). The progress of humanoid robotics, therefore, requires the study and understanding of the human neuromusculoskeletal system (NMS) along with human motor control paradigms, and propose bioinspired control strategies to generate better performance rhythmic movements, and adaptable to the environmental conditions.
Findings from neurobiology suggest that human gait is a dynamic, partly self-stabilizing, process based on the interaction of the biomechanical structure with the neural control. The coordination of this process is a very complex problem and has been proposed in several kinds of research, involving a hierarchy of levels, where the lowest, for example, interactions between the muscles and the spinal cord, are mostly autonomous and where the top-level, cortical control, acts timely, as needed. This requires an architecture of several nested sensorimotor loops, where the gait process provides feedback signals to the person's sensory systems, which can be used to coordinate their movements (Manoonpong, Geng, Kulvicius, Porr, & Wörgötter, 2007).
Therefore, neurobiology plays a crucial role in hypothesizing engineering-inspired biological models. The integration between neuroscience and applications in robotics has developed innovations in rhythmic movement generation research by neural oscillators acting as Central Pattern Generator (CPG), enhancing temporal synchronization with the environment, and the online correction of spatial errors for robots. As an example, there is the Darpa Robotics Challenge, which was created to stimulate research on bipedal humanoid robots capable of intervening in dangerous areas to perform emergency maneuvers instead of humans (Suekichi, 2017). Control architectures based on neural oscillators, such as the Matsuoka oscillator, generate almost sinusoidal trajectories to perform stable rhythmic movements and adapted by sensorimotor couplings. Conventionally, the oscillator output is implemented in the framework of feedforward control, where it is directly regarded as any manipulated quantity, such as torque, rate of angle, angle, etc., for each active joint in a robot (Habib, Watanabe, & Izumi, 2009).
Incorporating these new bioinspired control laws with an event-based control system that triggers phase transitions in gait cycle based on the sensed walking state instead of only relying on time-based reference trajectories, increases the robustness of biped robots on uneven or unstructured terrain (Buschmann et al., 2012).