CAJ | 학술논문

针对二阶非线性系统,提出了一种用高斯基函数作为神经元激励函数的 PID(Proportion-Integral-Derivative)控制方法.该方法用高斯基函数模拟 PID 参数随误差变化的曲线,用神经网络算法在线调整各模拟曲线的系数,从而构造出具有非线性特征的 PID 控制策略,实现了基于高斯基神经网络的非线性 PID 智能控制方法.计算机仿真结果表明,该方法具有良好的非线性控制效果,因此在工业领域具有广泛的应用前景.
침대이계비선성계통,제출료일충용고사기함수작위신경원격려함수적 PID(Proportion-Integral-Derivative)공제방법.해방법용고사기함수모의 PID 삼수수오차변화적곡선,용신경망락산법재선조정각모의곡선적계수,종이구조출구유비선성특정적 PID 공제책략,실현료기우고사기신경망락적비선성 PID 지능공제방법.계산궤방진결과표명,해방법구유량호적비선성공제효과,인차재공업영역구유엄범적응용전경.
Aiming at second-order nonlinear systems, a PID neural-network control approach using the Gaussian basis functions as a neurons excitation functions is proposed. This method uses the Gaussian basis functions to simulate the curves of the PID parameters following error changes, and uses neural network algorithm online to adjust the coefficient of the simulation curves. Thereby the nonlinear control strategy is obtained. The intelligent control method on the nonlinear PID is implemented based on the Gaussian basis neural-network. The simulation results show that the approach achieves a good nonlinear control effect, so it has a broad prospect of applications in the industrial fields.