基于预测控制的NCS时延补偿算法

针对网络控制系统(networked control system, NCS)中随机时延导致系统性能下降的问题，利用粒子群优化(particle swarm optimization, PSO)的最小二乘支持向量机(least square support vector machine, LS-SVM)建立NCS中随机时延预测模型，精确预测未来时刻的时延；同时利用该预测算法预测的时延通过快速隐式广义预测控制算法对NCS随机时延进行补偿。仿真结果表明，PSO优化的LS-SVM算法对随机时延具有较高的预测精度，同时快速隐式广义预测控制算法可使系统的输出很好地跟踪参考轨迹，保证系统良好的控制效果。

Networked control system time-delay compensation algorithm based on predictive control

According to the problem of random time-delay resulting in performance degradation of the networked control system (NCS), a particle swarm optimization (PSO) least square support vector machine (LS-SVM) method is used to establish an NCS time-delay prediction model and predict the future time-delay. At the same time, the fast implicit generalized prediction algorithm is used to compensate the random time-delay in the NCS.The simulation results show that the PSO optimized LS-SVM algorithm has a high prediction accuracy for the random time-delay. Meanwhile, fast implicit generalized prediction control (GPC) can make the output of the system track the reference trajectory well and ensure the good control effect of the system.