基于MATLAB的数据驱动模型仿真方法研究

Research on Data-driven Model Simulation Method Based on MATLAB

数据驱动模型的设计可以利用虚拟模型作为开发过程的核心,从而改进设计复杂系统的方式。使用MATLAB中Simulink进行基于模型的设计,可以缩短开发周期。文中基于Simulink研究了数据驱动下的互耦机械容器液位控制系统,在不同PID整定下的性能指标,根据仿真结果成功指导了系统设计。文中首先介绍了互耦机械容器模型,基于给定的数据集进行传递函数建模,并验证了模型的准确程度。然后对传递函数模型进行了频域分析,并利用几种不同的PID整定方法进行参数整定,通过仿真方法比较了这些方法的性能,最后得出了一种有效系统设计方案,验证了数据驱动模型仿真方法在开发过程中的有效性。

The design of data-driven models can utilize virtual models as the core of the development process,thereby improving the way complex systems are designed.Using Simulink in MATLAB for model based design can shorten the development cycle.This article is based on Simulink to study the performance indicators of a data-driven mutual coupling water tank level control system under different PID settings.Based on the simulation results,the system design was successfully guided.The article first introduces the mutual coupling water tank model,and then models the transfer function based on the given dataset,and verifies the accuracy of the model.Next,the article conducted frequency domain analysis of the transfer function model and utilized several different PID tuning methods for parameter tuning.The performance of these methods was compared through simulation methods,and an effective system design scheme was finally obtained to verify the effectiveness of the data-driven model simulation method in the development process.