Hindawi Publishing Corporation Mathematical Problems in Engineering Volume 2013, Article ID 240929, 8 pages http://dx.doi.org/10.1155/2013/240929
Research Article Recursive Least-Squares Estimation for Hammerstein Nonlinear Systems with Nonuniform Sampling Xiangli Li,1,2 Lincheng Zhou,1 Ruifeng Ding,3 and Jie Sheng4 1
Key Laboratory of Advanced Process Control for Light Industry (Ministry of Education), Jiangnan University, Wuxi 214122, China Jiangsu College of Information Technology, Wuxi 214153, China 3 School of Internet of Things Engineering, Jiangnan University, Wuxi 214122, China 4 Institute of Technology, University of Washington, Tacoma, WA 98402-3100, USA 2
Correspondence should be addressed to Ruifeng Ding;
[email protected] Received 17 June 2013; Accepted 2 September 2013 Academic Editor: Victoria Vampa Copyright © 2013 Xiangli Li et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. This paper focuses on the identification problem of Hammerstein nonlinear systems with nonuniform sampling. Using the key-term separation principle, we present a discrete identification model with nonuniform sampling input and output data based on the frame period. To estimate parameters of the presented model, an auxiliary model-based recursive least-squares algorithm is derived by replacing the unmeasurable variables in the information vector with their corresponding recursive estimates. The simulation results show the effectiveness of the proposed algorithm.
1. Introduction In actual industrial processes, there exist widely nonlinear systems which are described by block-oriented nonlinear systems [1–3]. Block-oriented nonlinear models are in general divided into Hammerstein systems and Wiener systems [4]. A Hammerstein system, which consists of a static nonlinear subsystem followed by a linear dynamic subsystem, can represent some nonlinear systems [5]. Many publications have been reported for the identification of the Hammerstein systems [6, 7]. For example, Chen et al. studied identification problems for the Hammerstein systems with saturation and dead-zone nonlinearities by choosing an appropriate switching function [8]; Ding et al. presented the projection, the stochastic gradient, and the Newton recursive and the Newton iterative identification algorithms for the Hammerstein nonlinear systems, and then they analyzed and compared the performances of these approaches by numerical examples [9]. Li et al. derived a least-squares based iterative algorithm for the Hammerstein output error systems with nonuniform sampling by using the overparameterization model [10]. The different input-output updating period (or called multirate sampling) is inevitable in discrete-time systems
[11–13]. The identification of multirate sampled systems have attracted much attention of many researchers. Recently, Liu et al. proposed a novel hierarchical least-squares algorithm for a class of nonuniformly sampled systems based on the hierarchical identification principle [14]. Shi et al. presented a crosstalk identification algorithm for multirate xDSL FIR systems [15]. Han et al. gave state-space models for multirate multi-input sampled-data systems and derived an auxiliary model-based recursive least-squares algorithm for identifying the parameters of multirate systems [16]. The recursive least-squares algorithm is a class of basic parameter estimation approaches which are suitable for online applications. In this literature, Wang adopted a filtering auxiliary model-based recursive least-squares identification algorithm for output error moving average systems [17]. Differing from the work in [14, 16], this paper discusses the parameter estimation problem for nonuniformly sampled Hammerstein nonlinear systems. The basic idea is, to combine the auxiliary model identification idea [18– 24] and the key-term separation principle to derive the auxiliary model-based recursive least-squares algorithm for the Hammerstein nonlinear systems with nonuniform sampling.
2
Mathematical Problems in Engineering Suppose that 𝑃𝑐 has the following state-space representation:
𝜐(kT) u(kT + ti )
H𝜏
u(t)
u(t) P f(·) c
y(t)
ST
y(kT)
⊕
y1 (kT)
ẋ (𝑡) = A𝑐 x (𝑡) + B𝑐 𝑢 (𝑡) ,
Figure 1: Hammerstein systems with nonuniform sampling.
𝑦 (𝑡) = Cx (𝑡) + 𝐷𝑢 (𝑡) ,
The rest of this paper is organized as follows. Section 2 establishes the identification model of the Hammerstein nonlinear systems with nonuniform sampling. Section 3 derives a recursive least-squares parameter estimation algorithm based on the auxiliary model identification idea. Section 4 provides an example to illustrate the effectiveness of the proposed algorithm. The conclusions of the paper are summarized in Section 5.
(3)
where x(𝑡) ∈ R𝑛 is the state vector, 𝑢(𝑡) and 𝑦(𝑡) are the input and output of the continuous-time process, respectively, and A𝑐 , B𝑐 , C, and 𝐷 are matrices of appropriate sizes. Referring to [25] and discretizing (3) with the frame period 𝑇, we have x (𝑘𝑇 + 𝑇) = 𝑒A𝑐 𝑇 x (𝑘𝑇) 𝑟
𝜏𝑖
+ ∑𝑒A𝑐 (𝑇−𝑡𝑖 ) ∫ 𝑒A𝑐 𝑡 d𝑡B𝑐 𝑢 (𝑘𝑇 + 𝑡𝑖−1 ) 0
𝑖=1
(4)
𝑟
= Ax (𝑘𝑇) + ∑B𝑖 𝑢 (𝑘𝑇 + 𝑡𝑖−1 ) , 𝑖=1
2. The Identification Model Let us introduce some notations. The superscript 𝑇 denotes the matrix transpose; I stands for an identity matrix of appropriate sizes; 1𝑛 represents an 𝑛-dimensional column vector whose elements are 1; “X := A” stands for “A is defined as X”; and 𝑧−1 is a unit backward shift operator; that is, 𝑧−1 𝑥(𝑡) = 𝑥(𝑡 − 1). Consider a Hammerstein nonlinear system with nonuniform sampling shown in Figure 1, where 𝐻𝜏 is a nonuniform zero-order hold with irregularly updating intervals {𝜏1 , 𝜏2 , . . . , 𝜏𝑟 }, dealing with a discrete-time signal 𝑢(𝑘𝑇 + 𝑡𝑖 ) and producing the input 𝑢(𝑡) of the nonlinear subsystem 𝑓(⋅); 𝑢(𝑡) is the output of the nonlinear subsystem; 𝑃𝑐 is a continuous-time process; 𝑦(𝑡) is the true output of 𝑃𝑐 but is unmeasurable; 𝑆𝑇 is a sampler that produces a discrete-time signal 𝑦(𝑘𝑇) with period 𝑇 = 𝜏1 +𝜏2 +⋅ ⋅ ⋅+𝜏𝑟 ; and 𝑦1 (𝑘𝑇) is the system output but is corrupted by the additive noise V(𝑘𝑇). Assuming that the input 𝑢(𝑡) has the updating intervals {𝜏1 , 𝜏2 , . . . , 𝜏𝑟 }, we have [11, 25]
𝑢 (𝑘𝑇) , 𝑘𝑇 ⩽ 𝑡 < 𝑘𝑇 + 𝑡1 , { { { { { {𝑢 (𝑘𝑇 + 𝑡1 ) , 𝑘𝑇 + 𝑡1 ⩽ 𝑡 < 𝑘𝑇 + 𝑡2 , { 𝑢 (𝑡) = { { .. { { . { { { {𝑢 (𝑘𝑇 + 𝑡𝑟−1 ) , 𝑘𝑇 + 𝑡𝑟−1 ⩽ 𝑡 < (𝑘 + 1) 𝑇,
where A := 𝑒Ac 𝑇 ∈ R𝑛×𝑛 , 𝜏𝑖
B𝑖 := 𝑒Ac (𝑇−𝑡𝑖 ) ∫ 𝑒Ac 𝑡 d𝑡B𝑐 ∈ R𝑛×𝑛 . 0
The output 𝑦(𝑡) at the sampling instant 𝑡 = 𝑘𝑇 can be expressed as 𝑦 (𝑘𝑇) = Cx (𝑘𝑇) + 𝐷𝑢 (𝑘𝑇) .
𝑦1 (𝑘𝑇) = 𝑦 (𝑘𝑇) + V (𝑘𝑇) .
𝑟
𝑧−𝑛 C adj [𝑧I − A] B𝑖 −𝑛 𝑖=1 𝑧 det [𝑧I − A]
𝑦 (𝑘𝑇) = ∑
=
(8)
1 𝑟 ∑𝛽 (𝑧) 𝑢 (𝑘𝑇 + 𝑡𝑖−1 ) , 𝛼 (𝑧) 𝑖=1 𝑖
where 𝛼 (𝑧) := 𝑧−𝑛 det [𝑧I − A] = 1 + 𝛼1 𝑧−1 + 𝛼2 𝑧−2 + ⋅ ⋅ ⋅ + 𝛼𝑛 𝑧−𝑛 , 𝛽1 (𝑧) := 𝑧−𝑛 C adj [𝑧I − A] B1 + 𝐷𝛼 (𝑧) = 𝛽10 + 𝛽11 𝑧−1 + 𝛽12 𝑧−2 + ⋅ ⋅ ⋅ + 𝛽1𝑛 𝑧−𝑛 , 𝛽𝑖 (𝑧) := 𝑧−𝑛 C adj [𝑧I − A] B𝑖 = 𝛽𝑖1 𝑧−1 + 𝛽𝑖2 𝑧−2 + ⋅ ⋅ ⋅ + 𝛽𝑖𝑛 𝑧−𝑛 ,
where 𝑛𝛾 is the polynomial order.
(7)
Referring to [26] and from (4) and (6), we have
(1)
𝑢 (𝑡) = 𝑓 (𝑢 (𝑡)) = 𝛾1 𝑢 (𝑡) + 𝛾2 𝑢2 (𝑡) + ⋅ ⋅ ⋅ + 𝛾𝑛𝛾 𝑢𝑛𝛾 (𝑡) , (2)
(6)
Hence, the system output 𝑦1 (𝑘𝑇) is written as
× 𝑢 (𝑘𝑇 + 𝑡𝑖−1 ) + 𝐷𝑢 (𝑘𝑇)
where 𝑇 := 𝜏1 + 𝜏2 + ⋅ ⋅ ⋅ + 𝜏𝑟 is the frame period. The nonlinear subsystem 𝑓(⋅) in the Hammerstein nonlinear system is a polynomial of a known order:
(5)
𝑖 = 1, 2, 3, . . . , 𝑟.
(9)
Mathematical Problems in Engineering
3
Equation (8) can be transformed into
Define the information vectors and the parameter vectors as 𝜑𝑦 (𝑘𝑇)
𝑟
𝑦 (𝑘𝑇) = [1 − 𝛼 (𝑧)] 𝑦 (𝑘𝑇) + ∑𝛽𝑖 (𝑧) 𝑢 (𝑘𝑇 + 𝑡𝑖−1 ) .
(10)
𝑖=1
:= [−𝑦 (𝑘𝑇 − 𝑇) , −𝑦 (𝑘𝑇 − 2𝑇) , . . . , −𝑦 (𝑘𝑇 − 𝑛𝑇)] ∈ R𝑛 , 𝜑𝑢 (𝑘𝑇)
Substituting (10) into (7), the system output 𝑦1 (𝑘𝑇) can be expressed as
:= [𝑢 (𝑘𝑇 − 𝑇) , 𝑢 (𝑘𝑇 − 2𝑇) , . . . , 𝑢 (𝑘𝑇 − 𝑛𝑇) , 𝑢 (𝑘𝑇 − 𝑇 + 𝑡1 ) , 𝑢 (𝑘𝑇 − 2𝑇 + 𝑡1 ) , . . . , 𝑢 (𝑘𝑇 − 𝑛𝑇 + 𝑡1 ) , . . . , 𝑢 (𝑘𝑇 − 𝑇 + 𝑡𝑟−1 ) ,
𝑦1 (𝑘𝑇) = [1 − 𝛼 (𝑧)] 𝑦 (𝑘𝑇)
𝑢 (𝑘𝑇 − 2𝑇 + 𝑡𝑟−1 ) , . . . , 𝑢 (𝑘𝑇 − 𝑛𝑇 + 𝑡𝑟−1 )] ∈ R𝑟𝑛 , T
𝑟
+ ∑𝛽𝑖 (𝑧) 𝑢 (𝑘𝑇 + 𝑡𝑖−1 ) + V (𝑘𝑇) .
(11)
T
𝜑𝛾 (𝑘𝑇) := [𝑢 (𝑘𝑇) , 𝑢2 (𝑘𝑇) , . . . , 𝑢𝑛𝛾 (𝑘𝑇)] ∈ R𝑛𝛾 ,
𝑖=1
𝜑𝑦 (𝑘𝑇) 𝜑 (𝑘𝑇) := [𝜑𝑢 (𝑘𝑇)] ∈ R𝑛0 , [𝜑𝛾 (𝑘𝑇)]
Equation (11) can be rewritten equivalently as
𝑛0 := (𝑟 + 1) 𝑛 + 𝑛𝛾 ,
𝜃𝑦 := [𝛼1 , 𝛼2 , . . . , 𝛼𝑛 ] ∈ R𝑛 , T
𝑛
𝜃𝑢 := [𝛽11 , 𝛽12 , . . . , 𝛽1𝑛 , 𝛽21 , 𝛽22 , . . . , 𝛽2𝑛 , . . . ,
𝑦1 (𝑘𝑇) = ∑ 𝛼𝑗 𝑦 (𝑘𝑇 − 𝑗𝑇) + 𝛽10 𝑢 (𝑘𝑇) 𝑗=1
𝛽𝑟1 , 𝛽𝑟2 , . . . , 𝛽𝑟𝑛 ] ∈ R𝑟𝑛 , T
𝑛
(12)
𝑟
T
𝜃𝛾 := [𝛾1 , 𝛾2 , . . . , 𝛾𝑛𝛾 ] ∈ R𝑛𝛾 ,
+ ∑ ∑𝛽𝑖𝑗 𝑢 (𝑘𝑇 + 𝑡𝑖−1 − 𝑗𝑇) + V (𝑘𝑇) . 𝑗=1 𝑖=1
Here, substituting (2) into (12) results in a complex expression containing the products of parameters. To solve this problem, we use the key-term separation principle presented in [27], and let 𝛽10 = 1. Then, the identification model of the proposed system is as follows:
𝜃𝑦 𝜃 := [𝜃𝑢 ] ∈ R𝑛0 . [ 𝜃𝛾 ] (14) Equation (13) can be written in a regressive form as 𝑦1 (𝑘𝑇) = 𝜑T (𝑘𝑇) 𝜃 + V (𝑘𝑇) .
𝑛
𝑛𝛾
𝑗=1
𝑖=1
Define a quadratic criterion function as 𝑖
𝑦1 (𝑘𝑇) = ∑ 𝛼𝑗 𝑦 (𝑘𝑇 − 𝑗𝑇) + ∑𝛾𝑖 𝑢 (𝑘𝑇) 𝑛
(15)
𝑟
𝑘
(13)
+ ∑ ∑𝛽𝑖𝑗 𝑢 (𝑘𝑇 + 𝑡𝑖−1 − 𝑗𝑇) + V (𝑘𝑇) . 𝑗=1 𝑖=1
The objective of this paper is to develop a recursive leastsquares algorithm for estimating the parameters of the nonuniformly sampled Hammerstein systems by using the auxiliary model identification idea in [11].
3. The Recursive Least-Squares Algorithm In this section, we derive the recursive least-squares estimation algorithm for the Hammerstein nonlinear systems with nonuniform sampling, referring to the method in [1].
2
𝐽 (𝜃) := ∑[𝑦1 (𝑖𝑇) − 𝜑T (𝑖𝑇) 𝜃] .
(16)
𝑖=1
̂ Let 𝜃(𝑘𝑇) be the estimate of 𝜃 at time 𝑘𝑇. Minimizing 𝐽(𝜃) gives the following recursive least-squares algorithm: 𝜃̂ (𝑘𝑇) = 𝜃̂ (𝑘𝑇 − 𝑇) + P (𝑘𝑇) 𝜑 (𝑘𝑇) × [𝑦1 (𝑘𝑇) − 𝜑𝑇 (𝑘𝑇) 𝜃̂ (𝑘𝑇 − 𝑇)] ,
(17)
P (𝑘𝑇) = P (𝑘𝑇 − 𝑇) −
P (𝑘𝑇 − 𝑇) 𝜑 (𝑘𝑇) 𝜑𝑇 (𝑘𝑇) P (𝑘𝑇 − 𝑞) . 1 + 𝜑𝑇 (𝑘𝑇) P (𝑘𝑇 − 𝑇) 𝜑 (𝑘𝑇)
(18)
Note that the information vector 𝜑(𝑘𝑇) in (17) contains unknown inner variables 𝑦(𝑘𝑇 − 𝑗𝑇) and 𝑢(𝑘𝑇 + 𝑡𝑖−1 − 𝑗𝑇); the parameter vector 𝜃 cannot be estimated by the standard
4
Mathematical Problems in Engineering
least-squares method. The solution is based on the auxiliary model identification idea [11]: to replace the unmeasurable term 𝑦(𝑘𝑇 − 𝑗𝑇) in 𝜑(𝑘𝑇) with its estimate ̂ T (𝑘𝑇) 𝜃̂ (𝑘𝑇) . 𝑦̂ (𝑘𝑇) = 𝜑
Equations (19) to (22) form the AM-RLS algorithm for the not uniformly sampled Hammerstein nonlinear systems, which can be summarized as 𝜃̂ (𝑘𝑇) = 𝜃̂ (𝑘𝑇 − 𝑇) ̂ T (𝑘𝑇) 𝜃̂ (𝑘𝑇 − 𝑇)] , ̂ (𝑘𝑇) [𝑦1 (𝑘𝑇) − 𝜑 + P (𝑘𝑇) 𝜑
(19)
P (𝑘𝑇) = P (𝑘𝑇 − 𝑇) Replacing 𝛾𝑖 in (2) with its estimate 𝛾̂𝑖 (𝑘𝑇), we can obtain the estimate 𝑢̂(𝑘𝑇 + 𝑡𝑖−1 ) of 𝑢(𝑘𝑇 + 𝑡𝑖−1 ) as follows:
−
̂ 𝑦 (𝑘𝑇) 𝜑 [ 𝜑 ̂ 𝜑 (𝑘𝑇) = ̂ 𝑢 (𝑘𝑇)] , ̂ 𝛾 (𝑘𝑇)] [𝜑
𝑢̂ (𝑘𝑇 + 𝑡𝑖−1 ) = 𝛾̂1 (𝑘𝑇) 𝑢 (𝑘𝑇 + 𝑡𝑖−1 ) + 𝛾̂2 (𝑘𝑇) 𝑢2 (𝑘𝑇 + 𝑡𝑖−1 ) + ⋅ ⋅ ⋅ + 𝛾̂𝑛𝛾 (𝑘𝑇) 𝑢𝑛𝛾 (𝑘𝑇 + 𝑡𝑖−1 ) ,
̂ (𝑘𝑇) 𝜑 ̂ T (𝑘𝑇) P (𝑘𝑇 − 𝑇) P (𝑘𝑇 − 𝑇) 𝜑 , T ̂ (𝑘𝑇) P (𝑘𝑇 − 𝑇) 𝜑 ̂ (𝑘𝑇) 1+𝜑
(20)
̂ 𝑦 (𝑘𝑇) 𝜑 T
= [−𝑦̂ (𝑘𝑇 − 𝑇) , −𝑦̂ (𝑘𝑇 − 2𝑇) , . . . , −𝑦̂ (𝑘𝑇 − 𝑛𝑇)] ,
𝑖 = 1, 2, . . . , 𝑟.
̂ 𝑢 (𝑘𝑇) = [̂ 𝜑 𝑢 (𝑘𝑇 − 𝑇) , 𝑢̂ (𝑘𝑇 − 2𝑇) , . . . , 𝑢̂ (𝑘𝑇 − 𝑛𝑇) , 𝑢̂ (𝑘𝑇 − 𝑇 + 𝑡1 ) ,
Define the estimate of 𝜑(𝑘𝑇) as
𝑢̂ (𝑘𝑇 − 2𝑇 + 𝑡1 ) , . . . , 𝑢̂ (𝑘𝑇 − 𝑛𝑇 + 𝑡1 ) , . . . , 𝑢̂ (𝑘𝑇 − 𝑇 + 𝑡𝑟−1 ) , 𝑢̂ (𝑘𝑇 − 2𝑇 + 𝑡𝑟−1 ) , . . . ,
̂ 𝑦 (𝑘𝑇) 𝜑 ̂ 𝑢 (𝑘𝑇)] ∈ R𝑛0 , ̂ (𝑘𝑇) := [𝜑 𝜑 ̂ 𝛾 (𝑘𝑇)] [𝜑
T
𝑢̂(𝑘𝑇 − 𝑛𝑇 + 𝑡𝑟−1 )] , T
𝜑𝛾 (𝑘𝑇) = [𝑢 (𝑘𝑇) , 𝑢2 (𝑘𝑇) , . . . , 𝑢𝑛𝛾 (𝑘𝑇)] ,
̂ 𝑦 (𝑘𝑇) := [−𝑦̂ (𝑘𝑇 − 𝑇) , −𝑦̂ (𝑘𝑇 − 2𝑇) , . . . , 𝜑
𝑢̂ (𝑘𝑇 + 𝑡𝑖−1 ) = 𝛾̂1 (𝑘𝑇) 𝑢 (𝑘𝑇 + 𝑡𝑖−1 )
−𝑦̂ (𝑘𝑇 − 𝑛𝑇)] ∈ R𝑛 ,
+ 𝛾̂2 (𝑘𝑇) 𝑢2 (𝑘𝑇 + 𝑡𝑖−1 )
̂ 𝑢 (𝑘𝑇) := [̂ 𝜑 𝑢 (𝑘𝑇 − 𝑇) , 𝑢̂ (𝑘𝑇 − 2𝑇) , . . . , 𝑢̂ (𝑘𝑇 − 𝑛𝑇) ,
+ ⋅ ⋅ ⋅ + 𝛾̂𝑛𝛾 (𝑘𝑇) 𝑢𝑛𝛾 (𝑘𝑇 + 𝑡𝑖−1 ) ,
𝑢̂ (𝑘𝑇 − 𝑇 + 𝑡1 ) , 𝑢̂ (𝑘𝑇 − 2𝑇 + 𝑡1 ) , . . . ,
(23)
̂ T (𝑘𝑇) 𝜃̂ (𝑘𝑇) . 𝑦̂ (𝑘𝑇) = 𝜑
𝑢̂ (𝑘𝑇 − 𝑛𝑇 + 𝑡1 ) , . . . , 𝑢̂ (𝑘𝑇 − 𝑇 + 𝑡𝑟−1 ) ,
̂ To initialize the algorithm, we take 𝜃(0) to be a small real ̂ vector; for example, 𝜃(0) = 1𝑛0 /𝑝0 and P(0) = 𝑝0 I with 𝑝0 normally a large positive number (e.g., 𝑝0 = 106 ).
𝑢̂ (𝑘𝑇 − 2𝑇 + 𝑡𝑟−1 ) , . . . , 𝑢̂(𝑘𝑇 − 𝑛𝑇 + 𝑡𝑟−1 )] ∈ R𝑟𝑛 . T
(21)
An example is given to demonstrate the feasibility of the proposed algorithm. Assume that the dynamical linear subsystem 𝑃𝑐 has the following state-space representation:
̂ (𝑘𝑇) in place of 𝜑(𝑘𝑇) in (17) and (18), we have Using 𝜑
−0.3 −0.2 1 ẋ (𝑡) = [ ] x (𝑡) + [ ] 𝑢 (𝑡) , 1 0 0
̂ (𝑘𝑇) 𝜃̂ (𝑘𝑇) = 𝜃̂ (𝑘𝑇 − 𝑇) + P (𝑘𝑇) 𝜑
and the static nonlinear subsystem is denoted by (22)
P (𝑘𝑇) = P (𝑘𝑇 − 𝑇) −
̂ T (𝑘𝑇) P (𝑘𝑇 − 𝑇) 𝜑 ̂ (𝑘𝑇) 1+𝜑
(24)
𝑦 (𝑡) = [−0.4, 0.2] x (𝑡) + 𝑢 (𝑡) ,
̂ T (𝑘𝑇) 𝜃̂ (𝑘𝑇 − 𝑞)] , × [𝑦1 (𝑘𝑇) − 𝜑
̂ (𝑘𝑇) 𝜑 ̂ T (𝑘𝑇) P (𝑘𝑇 − 𝑇) P (𝑘𝑇 − 𝑇) 𝜑
4. Example
.
𝑢 (𝑡) = 𝑓 (𝑢 (𝑡)) = 𝛾1 𝑢 (𝑡) + 𝛾2 𝑢2 (𝑡) + 𝛾3 𝑢3 (𝑡) = 𝑢 (𝑡) + 0.5𝑢2 (𝑡) + 0.25𝑢3 (𝑡) .
(25)
Mathematical Problems in Engineering
5
Table 1: The parameter estimates and their errors (𝜎2 = 2.002 ). 𝑘 100 500 1000 2000 3000 4000 5000 6000 True values
𝛼1 −0.03122 −0.53614 −0.55071 −0.57852 −0.60057 −0.61945 −0.66253 −0.67482 −0.68000
𝛼2 −0.05592 0.31821 0.44816 0.42159 0.39569 0.41573 0.42857 0.46019 0.47241
𝛽1 0.02344 −0.43136 −0.41426 −0.43335 −0.44560 −0.44904 −0.49385 −0.52100 −0.52674
𝛽2 0.39192 0.63245 0.78764 0.71714 0.68205 0.68292 0.68309 0.71556 0.73948
𝛽3 −0.08169 −0.30871 −0.24332 −0.22310 −0.23745 −0.24136 −0.26702 −0.26552 −0.25070
𝛽4 0.14407 0.53689 0.62087 0.63026 0.60417 0.60799 0.62299 0.64442 0.66221
𝛾1 1.52981 1.30985 1.12565 1.04567 1.09320 1.03387 1.00550 0.99935 1.00000
𝛾2 0.68962 0.69341 0.56690 0.49863 0.49337 0.50662 0.50873 0.53599 0.50000
𝛾3 0.02254 0.10358 0.22360 0.23534 0.20626 0.24914 0.25535 0.26184 0.25000
𝛿 (%) 73.36523 26.81256 12.86475 8.87717 10.44562 7.82199 5.03493 2.87761
𝛾2 0.54701 0.54706 0.51876 0.50079 0.49893 0.50179 0.50215 0.50881 0.50000
𝛾3 0.18587 0.22836 0.25274 0.25083 0.24231 0.25233 0.25341 0.25474 0.25000
𝛿 (%) 13.23523 7.30117 4.13831 2.89926 3.07166 2.68214 2.01651 1.45453
Table 2: The parameter estimates and their errors (𝜎2 = 0.502 ). 𝑘 100 500 1000 2000 3000 4000 5000 6000 True values
𝛼1 −0.69605 −0.68503 −0.65503 −0.64233 −0.64733 −0.65262 −0.66319 −0.66697 −0.68000
𝛼2 0.48092 0.51969 0.49678 0.46233 0.45069 0.45294 0.45510 0.46155 0.47241
𝛽1 −0.59484 −0.54542 −0.50768 −0.49229 −0.49445 −0.49564 −0.50654 −0.51409 −0.52674
𝛽2 0.83269 0.81494 0.79029 0.74295 0.72782 0.72445 0.72291 0.72921 0.73948
1
𝛽3 −0.30638 −0.30331 −0.26915 −0.25140 −0.25286 −0.25257 −0.25789 −0.25706 −0.25070
𝛽4 0.59981 0.68632 0.68531 0.66911 0.65764 0.65608 0.65842 0.66243 0.66221
𝛾1 1.17677 1.05813 1.01708 1.00467 1.01820 1.00431 0.99803 0.99700 1.00000
Thus, the corresponding input-output expression is given by 𝑦1 (𝑘𝑇) = (−0.68000𝑧−1 + 0.47241𝑧−2 )
0.6
× 𝑦 (𝑘𝑇) + 𝑢 (𝑘𝑇)
𝛿
0.8
0.4
+ (−0.52674𝑧−1 + 0.73948𝑧−2 ) 𝑢 (𝑘𝑇)
𝜎2 = 0.502
+ (−0.25070𝑧−1 + 0.66221𝑧−2 )
0.2 0
(27)
𝜎2 = 2.002 0
1000
2000
3000
4000
5000
6000
k
Figure 2: The estimation errors 𝛿 versus 𝑘.
Let 𝑟 = 2, 𝜏1 = 1s, and 𝜏2 = 2s; that is, 𝑡1 = 𝜏1 = 1s and 𝑡2 = 𝑡1 + 𝜏2 = 𝑇 = 3s. Discretizing 𝑃𝑐 with the frame period 𝑇, we obtain
𝑥 (𝑘𝑇 + 𝑇) = [
0.1274 −0.2835 ] x (𝑘𝑇) 1.4177 0.5527
+[
𝑢 (𝑘𝑇) 0.2981 1.1196 ], ][ 1.2978 0.9388 𝑢 (𝑘𝑇 + 𝑡1 )
𝑦 (𝑘𝑇) = [0.50, 0.21] x (𝑘𝑇) .
(26)
× 𝑢 (𝑘𝑇 + 𝑡1 ) + V (𝑘𝑇) . In simulation, the inputs {𝑢(𝑘𝑇)} and {𝑢(𝑘𝑇 + 𝑡1 )} are taken as persistent excitation signal sequences with zero mean and unit variance; {V(𝑘𝑇)} is a white noise with zero mean and variance 𝜎2 . Applying the proposed algorithm to estimate the parameters of this system, the estimates of 𝜃 and their errors with different noise variances are shown in Tables 1 and 2, and the parameter estimation errors 𝛿 := ̂ ‖𝜃(𝑘𝑇) − 𝜃‖/‖𝜃‖ versus 𝑘 are shown in Figure 2. When 𝜎2 = 2.002 and 𝜎2 = 0.502 , the corresponding signal-tonoise ratios (the square root of the ratio of output and noise variances) are SNR = 1.0342 and SNR = 4.1367, respectively. From Tables 1 and 2 and Figure 2, we can draw the following conclusions. (i) The parameter estimation errors of the AM-RLS algorithm become (generally) smaller as 𝑘 increases; see the estimation errors of the last columns of Tables 1 and 2 and Figure 2.
6
Mathematical Problems in Engineering (ii) Under different noise levels, the parameter estimates can converge to the true value, and a lower noise level results in a faster convergence rate of the parameter estimates to the true parameters; see the error curves in Figure 2 and the estimation errors in Tables 1 and 2. (iii) The proposed recursive algorithm differs from the iterative identification approach in [10] and can be used as an online identification.
5. Conclusions In this paper, we have established the identification model of the Hammerstein nonlinear systems with nonuniform sampling by using the key-term separation principle. To estimate the parameters of the proposed model, the recursive leastsquares parameter estimation algorithm is derived based on the auxiliary model identification idea. The proposed algorithm can simultaneously estimate the parameters of the linear and nonlinear subsystems of the Hammerstein nonlinear systems with nonuniform sampling. The simulation results show that the parameters of the Hammerstein systems with nonuniform sampling can be estimated effectively by the proposed algorithm. Although the algorithm is presented for a class of nonuniformly sampled Hammerstein nonlinear systems, the basic idea can also be extended to identify other linear and nonlinear systems [28, 29] and can combine the hierarchical identification methods [30–34], the multi-innovation identification methods [35–44], and other identification methods [45–58] to present new identification algorithms for linear or nonlinear and scalar or multivariable systems [59].
Acknowledgments This work was supported by the Fundamental Research Funds for the Central Universities (JUDCF11042 and JUDCF12031) and the PAPD of Jiangsu Higher Education Institutions and the 111 Project (B12018).
References [1] F. Ding, System Identification—New Theory and Methods, Science Press, Beijing, China, 2013. [2] L. C. Zhou, X. L. Li, and F. Pan, “Gradient-based iterative identification for MISO Wiener nonlinear systems: application to a glutamate fermentation process,” Applied Mathematics Letters, vol. 26, no. 8, pp. 886–892, 2013. [3] D. Wang and F. Ding, “Least squares based and gradient based iterative identification for Wiener nonlinear systems,” Signal Processing, vol. 91, no. 5, pp. 1182–1189, 2011. [4] I. Cherif, S. Abid, and F. Fnaiech, “Nonlinear blind identification with three-dimensional tensor analysis,” Mathematical Problems in Engineering, vol. 2012, Article ID 284815, 22 pages, 2012. [5] D. Wang, F. Ding, and Y. Chu, “Data filtering based recursive least squares algorithm for Hammerstein systems using the keyterm separation principle,” Information Sciences, vol. 222, pp. 203–212, 2013.
[6] D. Q. Wang and F. Ding, “Hierarchical least squares estimation algorithm for Hammerstein-Wiener systems,” IEEE Signal Processing Letters, vol. 19, no. 12, pp. 825–828, 2012. [7] F. Ding, “Hierarchical multi-innovation stochastic gradient algorithm for Hammerstein nonlinear system modeling,” Applied Mathematical Modelling, vol. 37, no. 4, pp. 1694–1704, 2013. [8] J. Chen, X. Wang, and R. Ding, “Gradient based estimation algorithm for Hammerstein systems with saturation and deadzone nonlinearities,” Applied Mathematical Modelling, vol. 36, no. 1, pp. 238–243, 2012. [9] F. Ding, X. P. Liu, and G. Liu, “Identification methods for Hammerstein nonlinear systems,” Digital Signal Processing, vol. 21, no. 2, pp. 215–238, 2011. [10] X. L. Li, R. F. Ding, and L. C. Zhou, “Least-squares-based iterative identification algorithm for Hammerstein nonlinear systems with non-uniform sampling,” International Journal of Computer Mathematics, vol. 90, no. 7, pp. 1524–1534, 2013. [11] Y. Liu, L. Xie, and F. Ding, “An auxiliary model based on a recursive least-squares parameter estimation algorithm for non-uniformly sampled multirate systems,” Proceedings of the Institution of Mechanical Engineers. Part I: Journal of Systems and Control Engineering, vol. 223, no. 4, pp. 445–454, 2009. [12] X. G. Liu and J. Lu, “Least squares based iterative identification for a class of multirate systems,” Automatica, vol. 46, no. 3, pp. 549–554, 2010. [13] Y. Wu, Y. Liu, and W. Zhang, “A discrete-time chattering free sliding mode control with multirate sampling method for flight simulator,” Mathematical Problems in Engineering, vol. 2013, Article ID 865493, 8 pages, 2013. [14] Y. Liu, F. Ding, and Y. Shi, “Least squares estimation for a class of non-uniformly sampled systems based on the hierarchical identification principle,” Circuits, Systems, and Signal Processing, vol. 31, no. 6, pp. 1985–2000, 2012. [15] Y. Shi, F. Ding, and T. Chen, “Multirate crosstalk identification in xDSL systems,” IEEE Transactions on Communications, vol. 54, no. 10, pp. 1878–1886, 2006. [16] L. Han, J. Sheng, F. Ding, and Y. Shi, “Auxiliary model identification method for multirate multi-input systems based on least squares,” Mathematical and Computer Modelling, vol. 50, no. 78, pp. 1100–1106, 2009. [17] D. Q. Wang, “Least squares-based recursive and iterative estimation for output error moving average systems using data filtering,” IET Control Theory & Applications, vol. 5, no. 14, pp. 1648–1657, 2011. [18] F. Ding and T. Chen, “Combined parameter and output estimation of dual-rate systems using an auxiliary model,” Automatica, vol. 40, no. 10, pp. 1739–1748, 2004. [19] F. Ding and T. Chen, “Parameter estimation of dual-rate stochastic systems by using an output error method,” IEEE Transactions on Automatic Control, vol. 50, no. 9, pp. 1436–1441, 2005. [20] F. Ding, Y. Shi, and T. Chen, “Auxiliary model-based leastsquares identification methods for Hammerstein output-error systems,” Systems & Control Letters, vol. 56, no. 5, pp. 373–380, 2007. [21] F. Ding and J. Ding, “Least-squares parameter estimation for systems with irregularly missing data,” International Journal of Adaptive Control and Signal Processing, vol. 24, no. 7, pp. 540– 553, 2010.
Mathematical Problems in Engineering [22] F. Ding and T. Chen, “Identification of dual-rate systems based on finite impulse response models,” International Journal of Adaptive Control and Signal Processing, vol. 18, no. 7, pp. 589– 598, 2004. [23] F. Ding and Y. Gu, “Performance analysis of the auxiliary modelbased least-squares identification algorithm for one-step statedelay systems,” International Journal of Computer Mathematics, vol. 89, no. 15, pp. 2019–2028, 2012. [24] F. Ding and Y. Gu, “Performance analysis of the auxiliary modelbased stochastic gradient parameter estimation algorithm for state-space systems with one-step state delay,” Circuits, Systems, and Signal Processing, vol. 32, no. 2, pp. 585–599, 2013. [25] F. Ding, L. Qiu, and T. Chen, “Reconstruction of continuoustime systems from their non-uniformly sampled discrete-time systems,” Automatica, vol. 45, no. 2, pp. 324–332, 2009. [26] L. Xie, Y. J. Liu, H. Z. Yang, and F. Ding, “Modelling and identification for non-uniformly periodically sampled-data systems,” IET Control Theory & Applications, vol. 4, no. 5, pp. 784–794, 2010. [27] J. V¨or¨os, “Identification of Hammerstein systems with timevarying piecewise-linear characteristics,” IEEE Transactions on Circuits and Systems II, vol. 52, no. 12, pp. 865–869, 2005. [28] L. C. Zhou, X. L. Li, and F. Pan, “Gradient based iterative parameter identification for Wiener nonlinear systems,” Applied Mathematical Modelling, vol. 37, no. 16-17, pp. 8203–8209, 2013. [29] L. Zhou, X. Li, and F. Pan, “Least-squares-based iterative identification algorithm for Wiener nonlinear systems,” Journal of Applied Mathematics, vol. 2013, Article ID 565841, 6 pages, 2013. [30] F. Ding and T. Chen, “Hierarchical gradient-based identification of multivariable discrete-time systems,” Automatica, vol. 41, no. 2, pp. 315–325, 2005. [31] F. Ding and T. Chen, “Hierarchical least squares identification methods for multivariable systems,” IEEE Transactions on Automatic Control, vol. 50, no. 3, pp. 397–402, 2005. [32] F. Ding and T. Chen, “Hierarchical identification of lifted statespace models for general dual-rate systems,” IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 52, no. 6, pp. 1179– 1187, 2005. [33] J. Ding, F. Ding, X. P. Liu, and G. Liu, “Hierarchical least squares identification for linear SISO systems with dual-rate sampleddata,” IEEE Transactions on Automatic Control, vol. 56, no. 11, pp. 2677–2683, 2011. [34] D. Wang, R. Ding, and X. Dong, “Iterative parameter estimation for a class of multivariable systems based on the hierarchical identification principle and the gradient search,” Circuits, Systems, and Signal Processing, vol. 31, no. 6, pp. 2167–2177, 2012. [35] F. Ding and T. Chen, “Performance analysis of multi-innovation gradient type identification methods,” Automatica, vol. 43, no. 1, pp. 1–14, 2007. [36] F. Ding, P. X. Liu, and G. Liu, “Multiinnovation least-squares identification for system modeling,” IEEE Transactions on Systems, Man, and Cybernetics B, vol. 40, no. 3, pp. 767–778, 2010. [37] F. Ding, P. X. Liu, and G. Liu, “Auxiliary model based multiinnovation extended stochastic gradient parameter estimation with colored measurement noises,” Signal Processing, vol. 89, no. 10, pp. 1883–1890, 2009. [38] F. Ding, “Several multi-innovation identification methods,” Digital Signal Processing, vol. 20, no. 4, pp. 1027–1039, 2010.
7 [39] F. Ding, H. Chen, and M. Li, “Multi-innovation least squares identification methods based on the auxiliary model for MISO systems,” Applied Mathematics and Computation, vol. 187, no. 2, pp. 658–668, 2007. [40] L. Han and F. Ding, “Multi-innovation stochastic gradient algorithms for multi-input multi-output systems,” Digital Signal Processing, vol. 19, no. 4, pp. 545–554, 2009. [41] D. Wang and F. Ding, “Performance analysis of the auxiliary models based multi-innovation stochastic gradient estimation algorithm for output error systems,” Digital Signal Processing, vol. 20, no. 3, pp. 750–762, 2010. [42] Y. Liu, L. Yu, and F. Ding, “Multi-innovation extended stochastic gradient algorithm and its performance analysis,” Circuits, Systems, and Signal Processing, vol. 29, no. 4, pp. 649–667, 2010. [43] Y. Liu, Y. Xiao, and X. Zhao, “Multi-innovation stochastic gradient algorithm for multiple-input single-output systems using the auxiliary model,” Applied Mathematics and Computation, vol. 215, no. 4, pp. 1477–1483, 2009. [44] L. Han and F. Ding, “Identification for multirate multi-input systems using the multi-innovation identification theory,” Computers & Mathematics with Applications, vol. 57, no. 9, pp. 1438– 1449, 2009. [45] F. Ding, G. Liu, and X. P. Liu, “Partially coupled stochastic gradient identification methods for non-uniformly sampled systems,” IEEE Transactions on Automatic Control, vol. 55, no. 8, pp. 1976–1981, 2010. [46] F. Ding, “Coupled-least-squares identification for multivariable systems,” IET Control Theory & Applications, vol. 7, no. 1, pp. 68–79, 2013. [47] F. Ding and T. Chen, “Performance bounds of forgetting factor least-squares algorithms for time-varying systems with finite meaurement data,” IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 52, no. 3, pp. 555–566, 2005. [48] Y. Liu, J. Sheng, and R. Ding, “Convergence of stochastic gradient estimation algorithm for multivariable ARX-like systems,” Computers & Mathematics with Applications, vol. 59, no. 8, pp. 2615–2627, 2010. [49] F. Ding, G. Liu, and X. P. Liu, “Parameter estimation with scarce measurements,” Automatica, vol. 47, no. 8, pp. 1646–1655, 2011. [50] F. Ding, T. Chen, and L. Qiu, “Bias compensation based recursive least-squares identification algorithm for MISO systems,” IEEE Transactions on Circuits and Systems II:Express Briefs, vol. 53, no. 5, pp. 349–353, 2006. [51] F. Ding, H. Yang, and F. Liu, “Performance analysis of stochastic gradient algorithms under weak conditions,” Science in China F, vol. 51, no. 9, pp. 1269–1280, 2008. [52] F. Ding, P. X. Liu, and H. Yang, “Parameter identification and intersample output estimation for dual-rate systems,” IEEE Transactions on Systems, Man, and Cybernetics A, vol. 38, no. 4, pp. 966–975, 2008. [53] W. Wang, F. Ding, and J. Dai, “Maximum likelihood least squares identification for systems with autoregressive moving average noise,” Applied Mathematical Modelling, vol. 36, no. 5, pp. 1842–1853, 2012. [54] F. Ding, X. G. Liu, and J. Chu, “Gradient-based and leastsquares-based iterative algorithms for Hammerstein systems using the hierarchical identification principle,” IET Control Theory and Applications, vol. 7, pp. 176–184, 2013. [55] F. Ding, Y. Liu, and B. Bao, “Gradient-based and least-squaresbased iterative estimation algorithms for multi-input multioutput systems,” Proceedings of the Institution of Mechanical
8
[56]
[57]
[58]
[59]
Mathematical Problems in Engineering Engineers. Part I: Journal of Systems and Control Engineering, vol. 226, no. 1, pp. 43–55, 2012. F. Ding, P. X. Liu, and G. Liu, “Gradient based and leastsquares based iterative identification methods for OE and OEMA systems,” Digital Signal Processing, vol. 20, no. 3, pp. 664–677, 2010. F. Ding, “Decomposition based fast least squares algorithm for output error systems,” Signal Processing, vol. 93, no. 5, pp. 1235– 1242, 2013. F. Ding, “Two-stage least squares based iterative estimation algorithm for CARARMA system modeling,” Applied Mathematical Modelling, vol. 37, no. 7, pp. 4798–4808, 2013. F. Ding and H. H. Duan, “Two-stage parameter estimation algorithms for Box-Jenkins systems,” IET Signal Processing, vol. 7, no. 8, pp. 646–654, 2013.
Advances in
Operations Research Hindawi Publishing Corporation http://www.hindawi.com
Volume 2014
Advances in
Decision Sciences Hindawi Publishing Corporation http://www.hindawi.com
Volume 2014
Journal of
Applied Mathematics
Algebra
Hindawi Publishing Corporation http://www.hindawi.com
Hindawi Publishing Corporation http://www.hindawi.com
Volume 2014
Journal of
Probability and Statistics Volume 2014
The Scientific World Journal Hindawi Publishing Corporation http://www.hindawi.com
Hindawi Publishing Corporation http://www.hindawi.com
Volume 2014
International Journal of
Differential Equations Hindawi Publishing Corporation http://www.hindawi.com
Volume 2014
Volume 2014
Submit your manuscripts at http://www.hindawi.com International Journal of
Advances in
Combinatorics Hindawi Publishing Corporation http://www.hindawi.com
Mathematical Physics Hindawi Publishing Corporation http://www.hindawi.com
Volume 2014
Journal of
Complex Analysis Hindawi Publishing Corporation http://www.hindawi.com
Volume 2014
International Journal of Mathematics and Mathematical Sciences
Mathematical Problems in Engineering
Journal of
Mathematics Hindawi Publishing Corporation http://www.hindawi.com
Volume 2014
Hindawi Publishing Corporation http://www.hindawi.com
Volume 2014
Volume 2014
Hindawi Publishing Corporation http://www.hindawi.com
Volume 2014
Discrete Mathematics
Journal of
Volume 2014
Hindawi Publishing Corporation http://www.hindawi.com
Discrete Dynamics in Nature and Society
Journal of
Function Spaces Hindawi Publishing Corporation http://www.hindawi.com
Abstract and Applied Analysis
Volume 2014
Hindawi Publishing Corporation http://www.hindawi.com
Volume 2014
Hindawi Publishing Corporation http://www.hindawi.com
Volume 2014
International Journal of
Journal of
Stochastic Analysis
Optimization
Hindawi Publishing Corporation http://www.hindawi.com
Hindawi Publishing Corporation http://www.hindawi.com
Volume 2014
Volume 2014