As a nonlinear control approach, a database-driven PID (DD-PID) controller has been proposed that uses distance calculations of an initial database to tune PID parameters. However, sorting operations are required in extracting neighborhood data from the database, which imposes a large computational load. Therefore, it is impossible to finish this computation within a restricted sampling period in control systems with fast response time and insufficient memory capacity. On the other hand, the cerebellar model articulation controller (CMAC) has advantages for these problems and has similar constructs to the DD approach. Therefore, this paper combines advantages of these two approaches, and proposes a CMAC-PID controller design based on off-line optimization of a database. Finally, a comparison with conventional off-line CMAC-PID controller and DD-PID controller through simulations are considered.

中文翻译：

---

使用CMAC存储器实现数据库驱动的PID控制系统

作为一种非线性控制方法，数据库驱动的 PID (DD-PID) 控制器被提出，它使用初始数据库的距离计算来调整 PID 参数。然而，从数据库中提取邻域数据时需要进行排序操作，这会带来很大的计算量。因此，在响应时间快、内存容量不足的控制系统中，不可能在有限的采样周期内完成这种计算。另一方面，小脑模型关节控制器 (CMAC) 在解决这些问题方面具有优势，并且具有与 DD 方法类似的结构。因此，本文结合这两种方法的优点，提出了一种基于数据库离线优化的CMAC-PID控制器设计。最后，