1.3 Physical Anti-Collision
25
1.3 Physical Anti-Collision
1.3.1 Definition of Physical Collision Avoidance
The “physical anti-collision” proposed in this book is relative to the software anticollision. Research has found that multi-tag collisions that occur in the frequency
bands above UHF have both inherent algorithm design flaws and some tags that
cannot be identified due to various external physical interferences. Physical anticollision mainly solves the problem of low batch recognition success rate caused by
the latter. Therefore, physical collision avoidance is defined as the use of physical
means to solve the problem of multi-tag collision caused by non-software factors.
1.3.2 The Main Features of Physical Collision Avoidance
(1) Reflecting physicality in principle: The main technical methods of physical
collision prevention are based on physical principles such as radiophysics,
thermodynamic analysis, and electromagnetic analysis.
(2) The verification method reflects the physicality: the verification of physical
anti-collision mainly adopts the semi-physical verification method, that is, the
physical quantity sensor is used to simulate the actual signal to verify the
anti-collision performance of multiple tags in actual application scenarios.
(3) The realization of anti-collision reflects the physicality: the specific realization
of physical anti-collision mainly adopts the method of physical space optimization, predicts the optimal physical space distribution of multiple tags through
artificial intelligence, and finally optimizes the physical space distribution to
maximize the resistance to the outside Physical interference.
1.3.3 The Structure of the Physical Anti-Collision System
The front end of the physical anti-collision system uses physical means to collect
data, for example, image sensors are used to collect the geometric characteristics of multi-tag distribution, and the back-end uses neural network algorithms to
learn, train, and predict the physical optimal distribution structure, and then adjust
the tag arrangement through physical means, angle, to achieve the overall optimal
recognition performance of the tag group.
25
1.3 Physical Anti-Collision
1.3.1 Definition of Physical Collision Avoidance
The “physical anti-collision” proposed in this book is relative to the software anticollision. Research has found that multi-tag collisions that occur in the frequency
bands above UHF have both inherent algorithm design flaws and some tags that
cannot be identified due to various external physical interferences. Physical anticollision mainly solves the problem of low batch recognition success rate caused by
the latter. Therefore, physical collision avoidance is defined as the use of physical
means to solve the problem of multi-tag collision caused by non-software factors.
1.3.2 The Main Features of Physical Collision Avoidance
(1) Reflecting physicality in principle: The main technical methods of physical
collision prevention are based on physical principles such as radiophysics,
thermodynamic analysis, and electromagnetic analysis.
(2) The verification method reflects the physicality: the verification of physical
anti-collision mainly adopts the semi-physical verification method, that is, the
physical quantity sensor is used to simulate the actual signal to verify the
anti-collision performance of multiple tags in actual application scenarios.
(3) The realization of anti-collision reflects the physicality: the specific realization
of physical anti-collision mainly adopts the method of physical space optimization, predicts the optimal physical space distribution of multiple tags through
artificial intelligence, and finally optimizes the physical space distribution to
maximize the resistance to the outside Physical interference.
1.3.3 The Structure of the Physical Anti-Collision System
The front end of the physical anti-collision system uses physical means to collect
data, for example, image sensors are used to collect the geometric characteristics of multi-tag distribution, and the back-end uses neural network algorithms to
learn, train, and predict the physical optimal distribution structure, and then adjust
the tag arrangement through physical means, angle, to achieve the overall optimal
recognition performance of the tag group.