6
1 Overview of RFID System Anti-Collision Technology
is not conducive to real-time control; in addition, in many unknown dynamic environments, this type of robot cannot adapt to the constantly changing environment
and complete autonomous control. Especially the robot’s autonomous orientation
problem cannot be solved by the general neural network method. Other theories such
as fuzzy control and genetic algorithm have also been introduced into the design of
robot behavior controllers or the study of behavior coordination and fusion strategies,
but in practical applications, there are generally problems such as poor reliability and
weak adaptive ability [29–32].
In 1998, the American scholar Arkin first proposed perception is not an independent process, but a unified whole that works with dynamic systems and control
systems [33]. Later researchers also generally believed that perception is an inseparable process from behavior, that is, the behavior needs to provide specific content for
the perception process, and perception provides the information needed for motion
control. In motion control, the perception ability of the robot can be improved by
placing multiple sensors in the appropriate position, which is conducive to the robot
selecting important information for specific tasks and discarding less important information. Traditional cognitive science divides the actual world into different categories
(perception types) as internal expressions (models) formed by facing the external
environment. Researchers of machine perception got inspiration from the perception
mechanism of this organism and began to study a new robot perception scheme.
Among them, the Swiss Federal Institute of Technology Verschoor and his collaborators proposed a new perception structure called “Distributed Adaptive Control”
(DAC) and applied it to the robot in the neural model [34, 35]. The DAC structure
is composed of three closely related control layers: Reactive Layer, Adaptive Layer,
and Contextual Layer. The reaction layer implements a series of conditioned reflex
behaviors based on low-level perception and unconditional input. The adaptive layer
associates the system with more complex external stimuli. The background layer
establishes a high-level expression through the memory storage structure.
Almost all adaptive behaviors require the fusion and processing of multi-sensory
information, and the transformation of this information into a series of purposeguided behaviors. Scientists have discovered in the study of certain animals in
nature that the entire perception process of animals is completed by external (environmental) stimuli and internal (animal body) feedback [36]. The cerebral cortex
processes information from recognized targets in the environment. This information
is obtained through long-term stimulation (training) of the neural unit receiver. In
this process, the overall dynamic behavior constitutes a neural response to external
excitation, and then a nonlinear dynamic connection is formed in the cerebral cortex.
Freeman, a famous biologist at the University of California, Berkeley, and his
collaborators have discovered the dynamics of animal cerebral cortex in the process
of perceptual processing through long-term experimental research, and put forward
the perceptual dynamics theory, which considers brain activity. It can be expressed by
the behavior of Chaotic Dynamics [37–39]. They did many interesting experiments
with rabbits, such as letting rabbits inhale various odors, and then observe them
through EEG scanners. They evaluated the electrical stimulation response of the
olfactory bulb in the rabbits’ brains. The electric waves of the olfactory bulb exhibit
1 Overview of RFID System Anti-Collision Technology
is not conducive to real-time control; in addition, in many unknown dynamic environments, this type of robot cannot adapt to the constantly changing environment
and complete autonomous control. Especially the robot’s autonomous orientation
problem cannot be solved by the general neural network method. Other theories such
as fuzzy control and genetic algorithm have also been introduced into the design of
robot behavior controllers or the study of behavior coordination and fusion strategies,
but in practical applications, there are generally problems such as poor reliability and
weak adaptive ability [29–32].
In 1998, the American scholar Arkin first proposed perception is not an independent process, but a unified whole that works with dynamic systems and control
systems [33]. Later researchers also generally believed that perception is an inseparable process from behavior, that is, the behavior needs to provide specific content for
the perception process, and perception provides the information needed for motion
control. In motion control, the perception ability of the robot can be improved by
placing multiple sensors in the appropriate position, which is conducive to the robot
selecting important information for specific tasks and discarding less important information. Traditional cognitive science divides the actual world into different categories
(perception types) as internal expressions (models) formed by facing the external
environment. Researchers of machine perception got inspiration from the perception
mechanism of this organism and began to study a new robot perception scheme.
Among them, the Swiss Federal Institute of Technology Verschoor and his collaborators proposed a new perception structure called “Distributed Adaptive Control”
(DAC) and applied it to the robot in the neural model [34, 35]. The DAC structure
is composed of three closely related control layers: Reactive Layer, Adaptive Layer,
and Contextual Layer. The reaction layer implements a series of conditioned reflex
behaviors based on low-level perception and unconditional input. The adaptive layer
associates the system with more complex external stimuli. The background layer
establishes a high-level expression through the memory storage structure.
Almost all adaptive behaviors require the fusion and processing of multi-sensory
information, and the transformation of this information into a series of purposeguided behaviors. Scientists have discovered in the study of certain animals in
nature that the entire perception process of animals is completed by external (environmental) stimuli and internal (animal body) feedback [36]. The cerebral cortex
processes information from recognized targets in the environment. This information
is obtained through long-term stimulation (training) of the neural unit receiver. In
this process, the overall dynamic behavior constitutes a neural response to external
excitation, and then a nonlinear dynamic connection is formed in the cerebral cortex.
Freeman, a famous biologist at the University of California, Berkeley, and his
collaborators have discovered the dynamics of animal cerebral cortex in the process
of perceptual processing through long-term experimental research, and put forward
the perceptual dynamics theory, which considers brain activity. It can be expressed by
the behavior of Chaotic Dynamics [37–39]. They did many interesting experiments
with rabbits, such as letting rabbits inhale various odors, and then observe them
through EEG scanners. They evaluated the electrical stimulation response of the
olfactory bulb in the rabbits’ brains. The electric waves of the olfactory bulb exhibit