Neuronal control of 6-legged locomotion

• Applicant: Professorin Dr. Silvia Daun
• Subject Area: Automation, Mechatronics, Control Systems, Intelligent Technical Systems, Robotics
• Term: from 2009 to 2016
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 100760332

Currently, little is known about the neuronal inter-leg coordination of insects during locomotion. Recent findings in the stick insect suggest that inter-leg coordination may not primarily result from the direct information exchange between the legs about their respective leg joint angles as has been inferred from behavioral studies. On the neuronal level evidence has been gathered indicating that weak inter-leg influences may modify local feedback and thereby affect motor pattern generation in a given leg. My goal is to create a mathematical model of the stick insect walking system in order to understand the neuronal mechanisms that underlie coordinating information transfer between legs. I propose to model the system of central pattern generators (CPG system), known to be present in the stick insect, in the following way: 1) the single CPGs will be modeled as two coupled and mutually inhibitory Hodgkin-Huxley type neurons forming a half-center oscillator; 2) the three CPGs of one leg will be coupled via sensory feedback signals; 3) the joint CPGs of the three legs on one side of the animal that are closest to the body will initially be described via a network of synaptically coupled half-center oscillators to represent synchronized rhythmic activity between these CPGs (in this case the coupling between opposite legs is known to be negligible) and 4) additional coupling of these oscillators via weighted sensory feedback signals will then be added, where the weighting will account for influences of different sensory signals. In order to understand the neuronal interactions between the functional elements of the stick insect walking system, I am going to embed the neuronal oscillators into the external world and analyze the control that the CPGs exert on a joint, on intra-leg and inter-leg movements, respectively. This will be done first, by using a very simple joint motor model and second, by including muscle properties and mechanical influences between opposite legs into the motor system. With this modeling study, I want to create a neuronal replication of Cruses behavioral rules of coordination between stepping legs (Cruse, 1990), and I plan to use genetic algorithms to propose different optimal CPG couplings responsible for coordinated inter-leg movements in different walking situations. If it turns out that the biologically known CPG couplings that I included into the CPG system so far are not sufficient to reproduce those rules, further CPG couplings have to be carefully identified and tested. The results and predictions of the simulations can then be used to plan and conduct new neurophysiological experiments.

• DFG Programme: Independent Junior Research Groups