论文部分内容阅读
提出了一种基于改进差分进化算法的驾驶员诱发振荡(PIO)预测评估方法。针对基本差分进化算法收敛慢、易陷入局部最优的问题,提出了基于混沌理论与高斯扰动的改进型差分进化算法。在此基础上,构建了基于改进差分进化算法的PIO预测评估流程与实施步骤,建立了基于速率限制的人机闭环系统模型。实例计算结果表明,该方法克服了传统评估方法的缺陷,可在多参数摄动情况下对PIO进行快速准确的评估。
A driver induced oscillation (PIO) prediction and assessment method based on improved differential evolution algorithm is proposed. Aiming at the problem that the basic differential evolution algorithm converges slowly and falls into local optimum, an improved differential evolution algorithm based on chaos theory and Gaussian perturbation is proposed. On this basis, the PIO predictive evaluation process and implementation steps based on improved differential evolution algorithm are constructed, and a human-machine closed-loop system model based on rate limitation is established. The numerical results show that this method overcomes the shortcomings of traditional methods and can be used to evaluate PIO quickly and accurately under multi-parameter perturbation.