CAJ | 학술논문

针对无线传感器网络(WSNs)节点的定位误差较大的问题,提出了一种新的具有局部搜索能力强的多智能体蝙蝠算法.改进算法中对寻优蝙蝠个体融入多智能体技术,通过邻域竞争合作算子以及自学习过程提高了算法全局搜索能力,避免算法陷入局部最优,加快算法的收敛速度.通过对标准测试函数的仿真,改进算法相比于其他算法,寻优精度和进化效率得到了较大的提高.随后采用多智能体蝙蝠算法求解无线传感节点定位问题,仿真结果表明改进算法减少了测距误差对定位精度的影响,提高了未知节点定位的精度,为无线传感网络节点定位的实际应用提供理论参考.
침대무선전감기망락(WSNs)절점적정위오차교대적문제,제출료일충신적구유국부수색능력강적다지능체편복산법.개진산법중대심우편복개체융입다지능체기술,통과린역경쟁합작산자이급자학습과정제고료산법전국수색능력,피면산법함입국부최우,가쾌산법적수렴속도.통과대표준측시함수적방진,개진산법상비우기타산법,심우정도화진화효솔득도료교대적제고.수후채용다지능체편복산법구해무선전감절점정위문제,방진결과표명개진산법감소료측거오차대정위정도적영향,제고료미지절점정위적정도,위무선전감망락절점정위적실제응용제공이론삼고.
In order to solve the node location error in wireless sensor networks(WSNs),this paper proposes a new multi-agent bat algorithm,which possesses favorable local searching ability. In the proposed algorithm,the bat indi-vidual is a agent,which could compete and cooperate with its agent neighbor areas to improve the efficiency of local searching. In this way the multi-agent bat algorithm(MA-BA)could avoid the algorithm into a local optimum and in-crease the convergence speed. Simulation results for standard test functions indicate that the proposed algorithm re-markably improves the global optimizing ability and evolutionary efficiency compared to other algorithms. Through implementing the MA-BA to node location prediction,the precision of the unknown node location could be im-proved due to decreasing the ranging error and has a certain significance to practical application of wireless sensor network node localization.