The Role of Chaotic Dynamics in Motor Pattern Generation

混沌动力学在运动模式生成中的作用PI：Peter F.Rowat，Co-PI：R.C.Elson动物如何产生和控制重复的行为，如行走、游泳或咀嚼？在多变或恶劣的环境中，对新的感觉信息做出快速、适当和灵活的反应的能力是至关重要的。产生动作的骨骼和肌肉的物理系统是由神经信号驱动的，神经信号是平行的动作电位序列，从被称为中央模式发生器(CPGs)的潜在生物回路传递到肌肉。研究的问题是理解CPG产生这些信号或运动模式的基本原理。现在，人们已经知道了CPG用来产生重复动作的许多神经“构件”。尚不清楚的是，如何将灵活性和响应性融入中央人民政府。被称为神经调节剂的生物物质可以在相对较慢的时间尺度上直接改变神经回路的属性。CPG电路的动力学特性能否直接影响到系统的灵活性和环境响应性？理论研究表明，混沌系统具有内在的特性，这对于产生可变的环境响应性输出是非常有用的。它的输出是高度可变的，但以一种结构化的方式，控制混沌系统的技术现在是众所周知的。同时，对生物CPG的实验研究表明，总是存在大量的可变性或抖动。这个项目一方面通过实验研究生物的可变性，另一方面开发了如何使用混沌系统来赋予电机系统响应的灵活性的理论模型。在实验数据收集和模型建立之间存在着持续的交互作用。将分析数据中的可变性，以确定噪声和确定性--混沌动力学--的相对贡献。使用来自龙虾口胃系统的两个众所周知的CPG，并由这些CPG的小子电路进行记录。最近发展起来的非线性分析算法被用于这种分析。使用相同的分析算法，研究了CPG内突触连接对固有神经元动力学的控制。将在真实的生物神经元和基于计算机的模型神经元之间建立模拟的突触连接。口胃CPG的动态控制将通过使用来自已识别的感觉神经元的模拟感觉反馈来研究。这里，通过再传入电路从CPG到外周再返回的反馈回路由基于计算机的模型闭合。在这些实验和数据分析的同时，将在模拟中开发一个基于混沌的机器人电机系统控制模型系统。它是围绕着一个健壮的、混沌的核心和一个利用最新的混沌控制算法的动态感应器控制器而设计的。这种设计与生物感觉运动系统的特性有许多相似之处。该项目将扩大我们对生物变异性的作用以及灵活机器人控制器设计的了解。

IBN-9975490The role of chaotic dynamics in motor pattern generationPI: Peter F. Rowat, Co-PI: R.C.ElsonHow do animals produce and control repetitive behaviors such as walking, swimming, or chewing? In a variable or hostile environment, the ability to produce fast, appropriate, and flexible responses to new sensory information is of vital importance. The physical system of bones and muscles that produces movements is driven by neural signals - parallel trains of action potentials -- passing from underlying biological circuits known as central pattern generators (CPGs) to the muscles. The research problem is to understand the basic principles of production of these signals, or motor patterns, by a CPG. Many neural "building blocks" with which a CPG produces repetitive movements are now known. What is not well understood is how flexibility and responsiveness is built into a CPG. Biological substances known as neuromodulators can directly alter the properties of neural circuits, on a relatively slow time-scale. Could dynamical properties of CPG circuits directly contribute to flexibility and environmental responsiveness?From theoretical studies, it is known that a chaotic system has intrinsic properties, which could be very useful for the production of variable and environmentally responsive output. Its output is highly variable but in a structured way and techniques for the control of a chaotic system are now well known. At the same time, experimental studies of biological CPGs show that a significant amount of variability or jitter is always present. This project experimentally investigates the biological variability, on the one hand, and on the other, develops theoretical models of how a chaotic system can be used to impart flexibility of response into a motor system. There is a continual interaction between experimental data collection and model building. The variability in the data will be analyzed to determine the relative contributions of noise and deterministic -- chaotic - dynamics. Two well-known CPGs from the lobster stomatogastric system are used and recordings are made from small sub-circuits of these CPGs. Recently developed nonlinear analysis algorithms are used for this analysis. The control of intrinsic neuronal dynamics by synaptic connections within a CPG are studied using the same analysis algorithms. Simulated synaptic connections will be constructed between real biological neurons and computer-based model neurons. The dynamical control of a stomatogastric CPG will be studied by using simulated sensory feedback from an identified sensory neuron. Here the feedback loop from CPG to periphery and back via re-afferent circuits is closed by a computer-based model. In parallel with these experiments and data analysis, a chaos-based model system for the control of a robotic motor system, will be developed in simulations. It is designed around a robust, chaotic core and a dynamic, sensorimotor controller that utilizes recent chaos-control algorithms. There are many parallels between this design and properties of biological sensorimotor systems. The project will extend our knowledge about the role of biological variability and also the design of flexible robotic controllers.