An improved DECPSOHDV-Hop algorithm for node location of WSN in Cyber-Physical-Social-System
An improved DECPSOHDV-Hop algorithm for node location of WSN in Cyber-Physical-Social-System
复制标题
DOI：
10.1016/j.comcom.2022.05.008
复制
发表时间：
2022-05-26
影响因子：
6
通讯作者：
Zheng, Zeng
中科院分区：
文献类型：
--
作者：
Deng, Tan;Tang, Xiaoyong;Zheng, Zeng
信息物理社会系统是指信息物理系统与社交网络之间的相互作用。信息传输的准确性在信息物理系统中非常重要，而无线传感器网络节点定位的信息传输占据着重要的地位。针对节点定位中普遍存在的定位精度和资源分配问题，提出一种基于差分进化的混沌粒子群优化混合距离矢量跳变算法（DECPSOHDV-Hop）。该算法在无线传感器网络节点定位中采用了改进的算法并行机制，通过动态优化可以更准确地定位未知节点。另外，粒子群优化算法在求解高维函数优化问题时容易过早收敛，陷入局部极值。因此，本文基于（DE）/current-tobest/1算子，提出了混合并行混沌粒子群算法。对于差分算子中的粒子优化，使用高斯/正态分布进行平衡优化。混合算法根据概率选择基本粒子群更新策略，简化粒子群更新策略或差分优化算法对个体进行更新。并行算法利用改进后的差分进化模型更新差分进化算法的高斯变异算子，从而更新整个种群。在仿真实验结果中，对20个高维基准函数的测试结果表明，与PSO和CPSO相比，上述两种改进算法具有更好的搜索精度和更高的收敛速度，以及对测试函数的优化结果;在WSN节点定位中，DECPSOHDV-Hop算法的定位准确率高达90%，并从三个不同的标准验证了该算法良好的定位稳定性。
Cyber-Physical Social System (CPSS) refers to the interaction between cyber-physical systems and social networks. The accuracy of information transmission in Cyber-Physical System is very important, and the information transmission of WSN node positioning occupies an important position. Aiming at the ubiquitous positioning accuracy and resource allocation problems in the node positioning, we propose an improved Chaotic Particle Swarm Optimization Hybrid Distance Vector Hopping Algorithm Based on Differential Evolution (DECPSOHDV-Hop) algorithm. This algorithm uses an improved algorithm parallel mechanism in the WSN node positioning, which can be more accurately positioned unknown node through dynamic optimization. In addition, when solving high-dimensional function optimization problems, particle swarm optimization (PSO) algorithms tend to converge prematurely and fall into local extremes. Therefore, based on the (DE)/current-tobest/1 operator, this paper proposes chaotic PSO algorithms of hybrid and parallel. For particle optimization in the difference operator, Gaussian/normal distribution is used for balance optimization. The hybrid method selects the basic particle swarm update strategy according to the probability, simplified the particle swarm update strategy or differential optimization to update the individual. The parallel method uses the improved model to update the Gaussian mutation operator of DE to update the entire population. In the simulation experiment results, the results of testing 20 high-dimensional benchmark functions show that compared with PSO and CPSO, the above two improved algorithms have better search accuracy and higher convergence speed, as well as the optimization results of the test function; In WSN node positioning, the positioning accuracy of the DECPSOHDV-Hop algorithm is as high as 90%, and the excellent positioning stability of the algorithm is verified from three different standards.