College of Computer and Information Technology, China Three Gorges University, Yichang 443002, China.
Sensors (Basel). 2023 Mar 3;23(5):2804. doi: 10.3390/s23052804.
The problem of positioning wireless sensor networks is an important and challenging topic in all walks of life. Inspired by the evolution behavior of natural plant communities and traditional positioning algorithms, a novel positioning algorithm based on the behavior of artificial plant communities is designed and presented here. First, a mathematical model of the artificial plant community is established. Artificial plant communities survive in habitable places rich in water and nutrients, offering the best feasible solution to the problem of positioning a wireless sensor network; otherwise, they leave the non-habitable area, abandoning the feasible solution with poor fitness. Second, an artificial plant community algorithm is presented to solve the positioning problems encountered in a wireless sensor network. The artificial plant community algorithm includes three basic operations, namely seeding, growing, and fruiting. Unlike traditional artificial intelligence algorithms, which always have a fixed population size and only one fitness comparison per iteration, the artificial plant community algorithm has a variable population size and three fitness comparisons per iteration. After seeding by an original population size, the population size decreases during growth, as only the individuals with high fitness can survive, while the individuals with low fitness die. In fruiting, the population size recovers, and the individuals with higher fitness can learn from each other and produce more fruits. The optimal solution in each iterative computing process can be preserved as a parthenogenesis fruit for the next seeding operation. When seeding again, the fruits with high fitness can survive and be seeded, while the fruits with low fitness die, and a small number of new seeds are generated through random seeding. Through the continuous cycle of these three basic operations, the artificial plant community can use a fitness function to obtain accurate solutions to positioning problems in limited time. Third, experiments are conducted using different random networks, and the results verify that the proposed positioning algorithms can obtain good positioning accuracy with a small amount of computation, which is suitable for wireless sensor nodes with limited computing resources. Finally, the full text is summarized, and the technical deficiencies and future research directions are presented.
无线传感器网络定位问题是各行各业的一个重要且具有挑战性的课题。受自然植物群落的进化行为和传统定位算法的启发，本文设计并提出了一种基于人工植物群落行为的新型定位算法。首先，建立了人工植物群落的数学模型。人工植物群落生存在富含水分和养分的适宜生境中，为解决无线传感器网络的定位问题提供了最佳可行解；否则，它们会离开不适生境，放弃适应度较差的可行解。其次，提出了一种用于解决无线传感器网络中定位问题的人工植物群落算法。人工植物群落算法包括三个基本操作，即播种、生长和结果。与传统的人工智能算法不同，传统的人工智能算法每次迭代只有一个固定的种群大小和一次适应度比较，而人工植物群落算法每次迭代具有可变的种群大小和三次适应度比较。在初始种群进行播种后，种群大小在生长过程中会减少，因为只有高适应度的个体才能存活，而低适应度的个体则会死亡。在结果阶段，种群大小会恢复，高适应度的个体可以相互学习并产生更多的果实。每个迭代计算过程中的最优解可以作为部分生殖的果实保留下来，用于下一次的播种操作。再次播种时，高适应度的果实可以存活并进行播种，而低适应度的果实则会死亡，并且通过随机播种产生少量新的种子。通过这三个基本操作的不断循环，人工植物群落可以使用适应度函数在有限的时间内获得定位问题的精确解。第三，通过不同的随机网络进行实验，验证了所提出的定位算法可以在少量计算量的情况下获得良好的定位精度，适用于计算资源有限的无线传感器节点。最后，对全文进行总结，并提出技术上的不足和未来的研究方向。
