A multi-robot patrol strategy for any security degree environment monitoring

Multi-robot system has become the hot issue in the last decades because of its efficiency, robustness, scalability and global optimality. This work proposes a multi-robot patrol strategy for environment monitoring. First, we propose the function of patrol nodes generating under the circumstances of known probability distribution of emergency. Then we put forward the hierarchy immune clustering algorithm to divide interested patrol nodes to each robot and finally produce the patrol strategy. Thus multi-robot patrol system can be suitable for any needs of security degree with balance of cost and benefit, and detect emergency with the maximal probability. The proposed solutions are verified through simulations and show satisfied performance.

1 Introduction

Environment monitoring is very important in realizing security and high efficiency of production and living. With the development of modern society the requirement for environment monitoring is getting higher. Many scholars have made a lot of research for different environments and a wide variety of monitoring applications. Sung et al. [1] established an intelligent environmental monitoring system that consists of a smart grid, bridges or machine tilt and temperature/humidity monitoring. Holman et al. [2] applied networked sensors for monitoring environmentally sensitive beaches and near shore coastal oceans. Ni et al. [3] proposed a novel abrupt event monitoring approach based on kernel principal component analysis (KPCA) and support vector machines, which was combined with the physical redundancy method. Lee et al. [4] introduced Activity Travel Pattern (ATP) monitoring in a large-scale city environment. And some researchers even can monitor the environment from a global-scale view [5]. In indoor environment monitoring many researchers focus on wireless sensor network with the shortcomings like energy consumption, less flexibility and time delay. So in this paper we consider to propose a multi-robot patrol strategy for any security degree environment monitoring.