文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > 工程科技 > 能源/化工 > Design of robust control and monitoring system for microgrid stabilization
侵权投诉

Design of robust control and monitoring system for microgrid stabilization

Design of robust control and monitoring system for microgrid stabilization

factorization is used to represent these unstructured KP_MTis,max, KI_MTis,max, KP_ESis,max, KI_ESis,max are the minimum uncertainties. A shaped plant Gs is expressed in form of and maximum parameters of PI controller in the side of normalized left coprime factor Gs = Ms-1Ns, when the islanding mode of MT output and ES load, respectively.

KP_MTug,min, KI_MTug,min, KP_ESug,min, KI_ESug,min and KP_MTug,max, perturbed plant GΔ is determined as follow:

KI_MTug,max, KP_ESug,max, KI_ESug,max are the minimum and

GΔ= (Ms+ΔMs) 1(Ns+ΔNs): ΔNsΔMs ≤1/γ (12) maximum parameters of PI controller in the side of

interconnected utility grid mode of MT output and ES load, ∞

where ΔMs and ΔNs are stable unknown transfer functions respectively. which represent unstructured uncertainties in the nominal

plant G. The H∞ robust stabilization problem can be E. PSO algorithm

The PSO algorithm [4] is explained as follows: established by GΔ and K as shown in Fig.5. The objective of

robust control design is to stabilize not only the nominal plant 1. Specify the parameters of PSO. Initialize a population of G but also the family of perturbed plant GΔ. In (12), 1/γ is the particles with random positions and velocities.

defined as the robust stability margin. The maximum stability 2. Evaluate the objective function in (18) for each particle. margin in the face of system uncertainties is given by the 3. Compare the fitness value of each particle with it’s the lowest achievable value of γ. Therefore, γmin implies the best position for particle (pbest). The best fitness value largest size of system uncertainties that can exist without among all the pbests is the best position of all particles in the destabilizing the closed-loop system in Fig.5. The value of group (gbest).

4. Update the velocity vi and position of particle xi by γmin can be easily calculated from

vi+1=w.vi+c1.rand1.(pbest xi)+c2.rand2.(gbest xi) (20) γmin=+λmaxXZ (13)

xi+1=xi+vi+1 (21) where λmax(XZ) denotes the maximum eigenvalue of XZ. For

w wminminimal state-space realization (A, B, C, D) of Gs, the values (22) w=wmax maxiter

of X and Z are unique positive solutions to the generalized itermaxcontrol algebraic Riccati equation where c and c are the cognitive and social acceleration

of range (0,1). w is the inertia weight factor. wmin and wmax are

the minimum and maximum of inertia weight factors,

T

A BS 1DTCZ+ZA BS 1DTC ZCTR 1CZ+BS 1BT=0 (15) respectively. iter and itermax are the iteration count and

maximum iteration, respectively.

where R = I + DDT and S = I + DT D. Note that no iteration on 5. When the maximum number of iterations is arrived, stop γ is needed to solve for γmin. To ensure the robust stability of the process. Otherwise go to process 2. the nominal plant, the weighting function is selected so that 1 ≤ γmin < 4.0 [5].

IV. SIMULATION REULTS

In this section, the weighting functions in (11) are

D. Generate the objective function

suitably selected as

K∞ in Fig.4 can be defined as

s+48.9961,s+46.9839,

WW=910.=806.K∞=W1 1K(s) MTisESis(16) s+96.9381s+93.9256

Because W2 = I. The necessary and sufficient condition of the s+41.9219,s+44.9197

WMTug=889.WESug=886.(23) robust controller K(s) [5] is s+90.9769s+95.9942

Hence, the shaped plant (Gs) can be established. Because I 1 (I GK)IG≤γ (17) s∞s K γmin is 2.5849. The searching parameters and PSO parameters

∞ ∞

are set as follows: KP_MTis∈[0.00011],KI_MTis∈[0.00011],

Substitute (16) in (17), the objective function is

KP_ESis∈[0.00011],KI_ESis∈[0.00011],KP_MTug∈[0.00010.5],

I 1 1

Minimize 1 I GsW1K(s) IGs (18) KI_MTug∈[0.00010.5],KP_ESug∈[0.00010.5],KI_ESug∈[0.00010.5],

∞ W1K(s)

PSO sizes = 50, maximum iterations = 100, c1= 2, c2= 2,

Subject to

wmin= 0.4 and wmin= 0.9. Therefore, the proposed MT and KP_MTis,min<KP_MTis<KP_MTis,max,

ES controllers (PMT & PES) in the sides of islanding and

KI_MTis,min<KI_MTis<KI_MTis,max, interconnected utility grid modes of RCMS are KP_ESis,min<KP_ESis<KP_ESis,max, 0.66070.8951

KPMTis=0.4300+,KPESis=0.4397+

ssKI_ESis,min<KI_ESis<KI_ESis,max,

0.39940.3999

KP_MTug,min<KP_MTug<KP_MTug,max, KPMTug=0.1493+ (24) , KPESug=0.1500+

ss

KI_MTug,min<KI_MTug<KI_MTug,max, In this simulation, the PMT & PES in RCMS is compared

with the CMT & CES in CMS [2-3]. It is supposed that the KP_ESug,min<KP_ESug<KP_ESug,max,

system is operated under the WP, PV and FC outputs and

KI_ESug,min<KI_ESug<KI_ESug,max (19) random load change in Figs.6–7. Fig.8 shows the frequency

where KP_MTis,min, KI_MTis,min, KP_ESis,min, KI_ESis,min and deviation with islanding operation. It is observed that the

(A BS

1T

DCX+XA BSDC XBSBX+CRC=0 (14) factors, respectively. rand1 and rand2 are the random numbers

1T

1T

T

1

)

T

()

12

and the generalized filtering algebraic Riccati equation

()()

()

上一页第3页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email