论文部分内容阅读
为了控制随意运动,中枢神经系统需要解决不同层次的计算问题:定出视觉坐标上的希望轨线,并将它变换为身体的坐标,产生运动命令。本文在生理知识和以前模型的基础上,提出一个多级神经网络模型。在模型中,联合皮层为运动区提供身体坐标系中的希望轨线,由长程感觉反馈闭环计算出运动命令。在脊-小脑与红核大细胞系统中,由突触的可塑性,在神经内部一个肌肉-骨骼系统的动态模型被建立,并用于监控运动命令的执行。如实际运动有误差,则可通过动态模型的内部反馈机制更新运动命令。而大脑-小脑-红核小细胞系统中,有一
In order to control free movement, the CNS needs to address different levels of computational problems: setting the desired trajectory in the visual coordinates and transforming it into the body’s coordinates, producing the motion commands. Based on physiological knowledge and previous models, this paper proposes a multi-level neural network model. In the model, the co-cortex provides the locomotion area with the desired trajectory in the body coordinate system, and motion commands are calculated from the long-range sensory feedback loop. In the ridge-cerebellum and nucleus red blood cell system, a dynamic model of the musculoskeletal system is built up inside the nerve by synaptic plasticity and is used to monitor the execution of motor commands. If there is an error in the actual motion, the motion command can be updated through the internal feedback mechanism of the dynamic model. The brain - cerebellum - red nucleus small cell system, one