基于LS-SVM的一类非线性系统的性能估计

Performance Assessment for a Class of Non-Linear Systems Based on LS-SVM

控制系统性能估计的研究多数针对线性系统,但是非线性系统本质上更加复杂,用传统的方法进行估计存在局限性。对于一类可由非线性部分叠加线性干扰表示的非线性系统,首先分析了其反馈不变量的存在性,指出此类系统性能估计的关键在于构造超前预测模型。接着用最小二乘支持向量机辨识非线性模型,把最小方差性能估计问题转换成模型参数辨识问题,辨识参数的同时可得系统最小方差估计值。最后通过一个仿真实例把方法和前人所述方法进行了比较,仿真结果验证了方法的有效性。

The study of performance assessment mostly considers linear systems. Owing to the complexity of non-linear systems, assessing the performance using conventional methods has some insufficiency. For a class of non-linear processes that can be described by the superposition of a non-linear dynamic model and additive linear disturbance, the minimum variance feedback invariant has been proved to be existent. The emphasis of estimating minimum variance performance bound is to construct b-step ahead representation. The least squares support vector machine is used to identify the non-linear system model. Furthermore, the problem of performance assessment can be equal to a step of parametric model identification. The minimum variance can be estimated from the process of model identification. A simulation example indicates that this approach gives more credible estimates of minimum variance performance bound than other methods.