Four-dimensional space translation and rotation-based mental rotation mechanism realization model

Abstract

The invention discloses a four-dimensional space translation and rotation-based mental rotation mechanism realization model, and aims at exploring application prospect of the model in mobile robot navigation. According to the model, the main brain region functions involved in the mental rotation process of primates are expressed in a quantification manner, such as a superior parietal cortex (SPC), a prefrontal cortex (PFC) and a premotor cortex (PMC). The model is implemented through the following steps of carrying out map construction and image expression by an SPC module; carrying out feature and posture matching by a PFC module, comparing image features obtained by the SPC with a specified threshold value so as to decide whether objects are the same or not and whether postures are the same or not; and carrying out simulated rotation on the current object by the PMC, comparing dot cloud feature maps of the current object and a target object, calculating a rotation matrix, carrying out simulated rotation on the current object, and repeating the step to ensure that the difference between the objects is minimum, so as to obtain a time-related four-dimensional rotation operator, wherein the rotation operator is used for instructing navigation tasks of a mobile robot. According to the model, the defects such as bad learning ability, strong dependency on environment and high training cost of the traditional navigation algorithms are overcome.

Description

technical field [0001] The invention relates to the modeling of the mental rotation mechanism in the cognitive process of the brain, in particular to the realization of a four-dimensional space translation and rotation model and its possible application in mobile robot navigation. Background technique [0002] Mental rotation is an important spatial cognition ability of primates. It can help us simulate the position state of rotating objects in the brain to solve a specific task, such as distinguishing the direction of two objects. Is it consistent. In the 1970s, Shepard et al.'s famous three-dimensional object psychological rotation experiment concluded that the reaction time for judging whether two objects match is linearly positively correlated with their angular difference. After the experiment was proposed, some new methods have also been used to study the biological neurological basis of mental rotation, such as: functional magnetic resonance imaging, positron emissio...

AI Technical Summary

Problems solved by technology

However, the traditional navigation algorithm also has some insurmountable disadvantages, such as weak adaptive ability and poor learning ability, etc.

Images