论文部分内容阅读
为了尽可能准确评估和预测网络安全状态,在研究量子粒子群优化算法(QPSO)的基础上,探索影响算法全局收敛性能的因素,尝试改变QPSO中局部吸引子的更新方式,充分运用了种群社会信息作为,提高全局收敛性能,形成一种基于进化策略的改进QPSO算法(evolutionary QPSO,eQPSO),并将其与成熟的灰色模型相结合,作为求解模型参数的可靠方式.同时,试验结果表明,所提出的改进算法收敛速度快,而且收敛精度优于PSO算法和QPSO算法,与灰色模型相结合后预测精度更高,可靠性更好.
In order to evaluate and predict the state of network security as accurately as possible, this paper explores the factors affecting the global convergence performance of QPSO based on the research of Quantum-behaved Particle Swarm Optimization (QPSO), tries to change the update way of local attractors in QPSO, makes full use of population society Information as a way to improve the global convergence performance and form an evolutionary QPSO (evolutionary QPSO, eQPSO), which is combined with a mature gray model as a reliable way to solve the model parameters.At the same time, the experimental results show that, The proposed algorithm has faster convergence rate and better convergence accuracy than PSO algorithm and QPSO algorithm. Combined with the gray model, the prediction accuracy is higher and the reliability is better.