基于遗传算法整定的PID网络流量控制

结合经典控制理论和优化控制理论设计了基于遗传算法的线性PID控制器,并与经典的PID控制器进行了比较．此控制器较好地控制了交换机缓冲器的队列长度,动态性能明显优于一般PID控制器．并得到了仿真结果的验证．

Self-Tuned by Genetic Algorithm with PID Network Flow Control

The rate-based flow control is one of the important means that avoid congestion and guarantee high throughput for available bit rate(ABR)communication in asynchronous transfer mode(ATM)net- works.The network time delay has a great adverse effect in the rate-based ABR flow control in ATM networks.But the general PID flow control algorithms are easy to produce huge oscillations,thus lead- ing buffer to overflow.This paper combines classic control and optimization control theories to design the linear PID controller.The controller parameters are self-tuned by the genetic algorithm compared with classical PID controller.The controller can control the queue length in the buffer of ATM switch and is much better than the general PID controller on the dynamic performance.The superiority is veri- fied by simulation.