This paper proposes an explicit adaptive controller to damp oscillations and to enhance the single machine infinite bus SMIB stability. Owing to the increasing requests for renewable energy and operating conditions, the identification for power systems has been increased recently. Changes in the power system parameters cause to use an explicit self-tuning control. The controller structure consists of combined on-line identifier and a feedback controller as PID and a radial basis function neural network (RBFNN) which acts as an adaptive power system stabilizer for SMIB. An adaptive linear neural network (ADALINE) depending on the input and output of open loop system is employed as on-line model identification to mimic on-line the SMIB output. The difference between SMIB and the identified model responses is used to adjust the ADALANN model weights on-line depending on a recursive least squares principle RLS and a recursive least square with adaptive directional forgetting RLSMadf. The particles swarm optimization (PSO) beside RLS and RLSMadf assess the weights of (RBFNN) and coefficients of PID controllers depending on the on-line ADALINE model weights. The proposed controller is validated with several operating conditions under various disturbances. The simulation results show the proposed controller whose parameters depend on on-line tuning techniques provides better performance than a conventional PID controller.

本文提出了一种显式自适应控制器来抑制振荡并增强单机无限总线 SMIB 的稳定性。由于对可再生能源和运行条件的要求不断增加，最近增加了对电力系统的识别。电力系统参数的变化导致使用显式自整定控制。控制器结构由组合在线识别器和作为 PID 的反馈控制器以及作为 SMIB 自适应电力系统稳定器的径向基函数神经网络 (RBFNN) 组成。依赖于开环系统的输入和输出的自适应线性神经网络 (ADALINE) 被用作在线模型识别以在线模拟 SMIB 输出。SMIB 和识别的模型响应之间的差异用于根据递归最小二乘原理 RLS 和具有自适应方向遗忘 RLSMadf 的递归最小二乘法在线调整 ADALANN 模型权重。RLS 和 RLSMadf 旁边的粒子群优化 (PSO) 根据在线 ADALINE 模型权重评估 (RBFNN) 的权重和 PID 控制器的系数。建议的控制器在各种干扰下的几种操作条件下进行了验证。仿真结果表明，所提出的控制器的参数取决于在线调谐技术，其性能优于传统的 PID 控制器。RLS 和 RLSMadf 旁边的粒子群优化 (PSO) 根据在线 ADALINE 模型权重评估 (RBFNN) 的权重和 PID 控制器的系数。建议的控制器在各种干扰下的几种操作条件下进行了验证。仿真结果表明，所提出的控制器的参数取决于在线调谐技术，其性能优于传统的 PID 控制器。RLS 和 RLSMadf 旁边的粒子群优化 (PSO) 根据在线 ADALINE 模型权重评估 (RBFNN) 的权重和 PID 控制器的系数。建议的控制器在各种干扰下的几种操作条件下进行了验证。仿真结果表明，所提出的控制器的参数取决于在线调谐技术，其性能优于传统的 PID 控制器。