-
A novel approach for multi-agent cooperative pursuit to capture grouped evaders
Authors:
Muhammad Zuhair Qadir,
Songhao Piao,
Haiyang Jiang,
Mohammed El Habib Souidi
Abstract:
An approach of mobile multi-agent pursuit based on application of self-organizing feature map (SOFM) and along with that reinforcement learning based on agent group role membership function (AGRMF) model is proposed. This method promotes dynamic organization of the pursuers' groups and also makes pursuers' group evader according to their desire based on SOFM and AGRMF techniques. This helps to ove…
▽ More
An approach of mobile multi-agent pursuit based on application of self-organizing feature map (SOFM) and along with that reinforcement learning based on agent group role membership function (AGRMF) model is proposed. This method promotes dynamic organization of the pursuers' groups and also makes pursuers' group evader according to their desire based on SOFM and AGRMF techniques. This helps to overcome the shortcomings of the pursuers that they cannot fully reorganize when the goal is too independent in process of AGRMF models operation. Besides, we also discuss a new reward function. After the formation of the group, reinforcement learning is applied to get the optimal solution for each agent. The results of each step in capturing process will finally affect the AGR membership function to speed up the convergence of the competitive neural network. The experiments result shows that this approach is more effective for the mobile agents to capture evaders.
△ Less
Submitted 27 June, 2020; v1 submitted 1 June, 2020;
originally announced June 2020.
-
Coalition formation for Multi-agent Pursuit based on Neural Network and AGRMF Model
Authors:
Zhaoyi Pei,
Songhao Piao,
Mohammed Ei Souidi
Abstract:
An approach for coalition formation of multi-agent pursuit based on neural network and AGRMF model is proposed.This paper constructs a novel neural work called AGRMF-ANN which consists of feature extraction part and group generation part. On one hand,The convolutional layers of feature extraction part can abstract the features of agent group role membership function(AGRMF) for all of the groups,on…
▽ More
An approach for coalition formation of multi-agent pursuit based on neural network and AGRMF model is proposed.This paper constructs a novel neural work called AGRMF-ANN which consists of feature extraction part and group generation part. On one hand,The convolutional layers of feature extraction part can abstract the features of agent group role membership function(AGRMF) for all of the groups,on the other hand,those features will be fed to the group generation part based on self-organizing map(SOM) layer which is used to group the pursuers with similar features in the same group. Besides, we also come up the group attractiveness function(GAF) to evaluate the quality of groups and the pursuers contribution in order to adjust the main ability indicators of AGRMF and other weight of all neural network. The simulation experiment showed that this proposal can improve the effectiveness of coalition formation for multi-agent pursuit and ability to adopt pursuit-evasion problem with the scale of pursuer team growing.
△ Less
Submitted 17 July, 2017;
originally announced July 2017.
-
Compact Yb$^+$ optical atomic clock project: design principle and current status
Authors:
Clément Lacroûte,
Maël Souidi,
Pierre-Yves Bourgeois,
Jacques Millo,
Khaldoun Saleh,
Emmanuel Bigler,
Rodolphe Boudot,
Vincent Giordano,
Yann Kersalé
Abstract:
We present the design of a compact optical clock based on the $^2S_{1/2} \rightarrow ^2D_{3/2}$ 435.5 nm transition in $^{171}$Yb$^+$. The ion trap will be based on a micro-fabricated circuit, with surface electrodes generating a trapping potential to localize a single Yb ion a few hundred $μ$m from the electrodes. We present our trap design as well as simulations of the resulting trapping pseudo-…
▽ More
We present the design of a compact optical clock based on the $^2S_{1/2} \rightarrow ^2D_{3/2}$ 435.5 nm transition in $^{171}$Yb$^+$. The ion trap will be based on a micro-fabricated circuit, with surface electrodes generating a trapping potential to localize a single Yb ion a few hundred $μ$m from the electrodes. We present our trap design as well as simulations of the resulting trapping pseudo-potential. We also present a compact, multi-channel wavelength meter that will permit the frequency stabilization of the cooling, repumping and clear-out lasers at 369.5 nm, 935.2 nm and 638.6 nm needed to cool the ion. We use this wavelength meter to characterize and stabilize the frequency of extended cavity diode lasers at 369.5 nm and 638.6 nm.
△ Less
Submitted 18 March, 2016;
originally announced March 2016.
