158
5 Optimization Algorithm and RFID System Physical Anti-Collision
multiple inputs and outputs. Besides that the neural network has a strong adaptive
and self-learning ability. So, the neural network has been widely used in many situations. In [16], Wang built a wind power range prediction model based on the multiple
output property of backpropagation (BP) neural network. Then, the improved particle
swarm optimization (PSO) algorithm was used to optimize the model. In [17], Ding
proposed the genetic algorithm (GA) to optimize the BP algorithm. The genetic algorithm can overcome BP’s disadvantage of being easily stuck in a local minimum.
In [18], Wang presented a two-layer decomposition technique and then developed
a hybrid model based on fast ensemble empirical mode decomposition (FEEMD),
variational mode decomposition (VMD), and BP neural network. In [19], Doucoure
developed a prediction method for renewable energy sources to achieve an intelligent
management of a microgrid system. The proposed method was based on the multiresolution analysis of the time-series by means of wavelet decomposition and artificial neural networks. Sharma proposed a mixed wavelet neural network (WNN) for
short-term solar irradiance forecasting, with initial application in tropical Singapore
[20].
In order to optimize the reading performance of tags, this chapter mainly models
the nonlinear relationship between the coordinate data of multi-tag and the reading
distance through the following three aspects, and reads the optimal geometric distribution of tags. Therefore, the main contents of this chapter are as follows. Section 5.1
introduces the multi-tag optimization method based on particle swarm optimization (PSO). Section 5.2 introduces the multi-tag anti-collision optimization algorithm based on support vector machine (SVM). Section 5.3 introduces the multi-tag
anti-collision optimization algorithm based on wavelet. Finally, conclusion is given.
5.1 Physical Anti-Collision Based on Particle Swarm
Optimization (PSO)
5.1.1 Design and Application of Detection System
(1) Structure of detection system
In order to simulate the mobile product and the environment, a RFID detection system
has been designed (shown in Figs. 5.1 and 5.2). The RFID detection system mainly
consists of three parts, including the acquisition system, detection system, and control
system. The acquisition system is composed of a vertical CCD, a horizontal CCD,
an servo motor and a laser ranging sensor. The detection system includes a reader,
a certain number of antennas, and an antenna frame. The control system consists of
a pallet, a transportation device, and a control computer. The application inspection
items of the system include the test of tags’ reading range, anti-collision performance,
and location optimization.
5 Optimization Algorithm and RFID System Physical Anti-Collision
multiple inputs and outputs. Besides that the neural network has a strong adaptive
and self-learning ability. So, the neural network has been widely used in many situations. In [16], Wang built a wind power range prediction model based on the multiple
output property of backpropagation (BP) neural network. Then, the improved particle
swarm optimization (PSO) algorithm was used to optimize the model. In [17], Ding
proposed the genetic algorithm (GA) to optimize the BP algorithm. The genetic algorithm can overcome BP’s disadvantage of being easily stuck in a local minimum.
In [18], Wang presented a two-layer decomposition technique and then developed
a hybrid model based on fast ensemble empirical mode decomposition (FEEMD),
variational mode decomposition (VMD), and BP neural network. In [19], Doucoure
developed a prediction method for renewable energy sources to achieve an intelligent
management of a microgrid system. The proposed method was based on the multiresolution analysis of the time-series by means of wavelet decomposition and artificial neural networks. Sharma proposed a mixed wavelet neural network (WNN) for
short-term solar irradiance forecasting, with initial application in tropical Singapore
[20].
In order to optimize the reading performance of tags, this chapter mainly models
the nonlinear relationship between the coordinate data of multi-tag and the reading
distance through the following three aspects, and reads the optimal geometric distribution of tags. Therefore, the main contents of this chapter are as follows. Section 5.1
introduces the multi-tag optimization method based on particle swarm optimization (PSO). Section 5.2 introduces the multi-tag anti-collision optimization algorithm based on support vector machine (SVM). Section 5.3 introduces the multi-tag
anti-collision optimization algorithm based on wavelet. Finally, conclusion is given.
5.1 Physical Anti-Collision Based on Particle Swarm
Optimization (PSO)
5.1.1 Design and Application of Detection System
(1) Structure of detection system
In order to simulate the mobile product and the environment, a RFID detection system
has been designed (shown in Figs. 5.1 and 5.2). The RFID detection system mainly
consists of three parts, including the acquisition system, detection system, and control
system. The acquisition system is composed of a vertical CCD, a horizontal CCD,
an servo motor and a laser ranging sensor. The detection system includes a reader,
a certain number of antennas, and an antenna frame. The control system consists of
a pallet, a transportation device, and a control computer. The application inspection
items of the system include the test of tags’ reading range, anti-collision performance,
and location optimization.