Group Formation Algorithms Based On LEACH For Multi-agent Systems

In the field of multi-agent systems, formation control is always a hot issue and is widely used in the field of unmanned aerial vehicles and intelligent transportation networks with important military value and civilian value. At present, under the restriction of only partial agents in the system having target information, most of the existing formation control algorithms are unable to guarantee the completion of formation to one hundred percent, and all these algorithms do not consider the system power consumption. This situation limits the practicality of formation control algorithms. In order to address this limitation, this thesis conducted an in-depth study about the issues of formation efficiency and system power consumption, and used the formation algorithm simulation software to verify the related research results. The main contents are as follows:1. Proposed a single target tracking group formation algorithm based on LEACH structure. In order to divide the multi-agent system into two collections: one collection consists of master agents having the ability to obtain target information and the other one consists of slave agents having no target information, the algorithm applies the classic LEACH structure into the system. Meanwhile, this thesis improves the master agent selection method to optimize the existing formation algorithms. The optimized algorithm can not only effectively reduce the target-detection energy consumption which allows the system a longer running time, but also reduce the dependence of the formation completion to the initial state of the system, the selection of master agents and the proportion rate of master agents, and what is more, the size of master agents collection can realize adaptive adjustment according to the degree of formation cohesion.2. Proposed a single target tracking obstacle avoidance group formation algorithm based on LEACH structure. This thesis extends the group formation algorithm based on LEACH by adding obstacle avoidance control entry to it, and makes the new improved formation algorithm applicable under environmental constraints. This algorithm can adaptively adjust the proportion rate of master agents in the system according to the environmental complexity and the system fragmentation, and make the multi-agent system avoid obstacles successfully and complete single target tracking tasks at the same time.3. Design and implement the formation algorithm simulation software. This simulation software can not only realize the above two group formation algorithms based on LEACH structure, but also do the comparative experiments against some existing formation algorithms to verify the effect of different formation algorithms.