CN104950675A - Adaptive control method and adaptive control device for multi-working-condition power system - Google Patents
Adaptive control method and adaptive control device for multi-working-condition power system Download PDFInfo
- Publication number
- CN104950675A CN104950675A CN201510325481.8A CN201510325481A CN104950675A CN 104950675 A CN104950675 A CN 104950675A CN 201510325481 A CN201510325481 A CN 201510325481A CN 104950675 A CN104950675 A CN 104950675A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- mtd
- msubsup
- math
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 61
- 230000003044 adaptive effect Effects 0.000 title claims abstract description 30
- 239000013598 vector Substances 0.000 claims abstract description 47
- 238000011217 control strategy Methods 0.000 claims abstract description 11
- 239000011159 matrix material Substances 0.000 claims description 43
- 230000008569 process Effects 0.000 claims description 8
- 238000010276 construction Methods 0.000 claims description 6
- 230000007704 transition Effects 0.000 description 7
- 230000010355 oscillation Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 230000004044 response Effects 0.000 description 5
- 230000008859 change Effects 0.000 description 3
- 230000014509 gene expression Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000006641 stabilisation Effects 0.000 description 1
- 238000011105 stabilization Methods 0.000 description 1
Landscapes
- Feedback Control In General (AREA)
Abstract
The invention discloses an adaptive control method and an adaptive control device for a multi-working-condition power system. The adaptive control device comprises a data acquisition module, a multi-working-condition power system model constructing module, an adaptive control strategy generation module, a controller combination module and a result outputting module, wherein the data acquisition module is used for acquiring network structure parameters, a power grid state vector and a control output vector; the multi-working-condition power system model constructing module is used for constructing a multi-working-condition power system model and determining the stability criterion; the adaptive control strategy generation module is used for solving all secondary working condition matching controllers of the system; the controller combination module is used for selecting the combination mode of the controllers by using a weighting method; the result outputting module is used for outputting the combination mode of the controllers. A time-varying power system model and a random stochastic gradient online identification method can be combined by using the adaptive control method and the adaptive control device for the multi-working-condition power system, and the adaptive control is realized by using a weighting coefficient.
Description
Technical Field
The invention relates to the technical field of stability analysis of power systems, in particular to the technical field of multi-working-condition power system control.
Background
With the rapid increase of power demand and the gradual complexity of grid structures, the problem of multi-working-condition change becomes one of the most critical problems influencing the safe and stable operation of a power system, so that the research on the self-adaptive control strategy aiming at the multi-working-condition change of the power system has extremely important practical significance.
At present, the design of a controller related to a multi-condition power system is mainly based on a robust control method, and multi-condition factors are used as uncertainty of the system and are integrated into H-infinity control and μ control, wherein the H-infinity control and the μ control are methods for designing a multivariable input and output (MIMO) robust control system in a modern control theory. Due to the working condition variation, the external interference and the modeling error, an accurate model of the actual industrial process is difficult to obtain, and various faults of the system cause the uncertainty of the model, so that the uncertainty of the model is widely existed in the control system. How to design a fixed controller to make the object with uncertainty meet the control quality is the technical effect to be achieved by robust control.
However, when the system operating condition changes in a large range, the robust control may not achieve the predetermined control effect. Aiming at the problem, a multi-cell control system is adopted in the prior art to solve the problem, the multi-cell control takes a plurality of sub-working conditions of the system as vertexes of the multi-cell, and a controller meeting all the sub-working conditions is solved.
Disclosure of Invention
In view of this, the present invention aims to overcome the problem in the prior art that a design scene is difficult to match with a current working condition due to the limitation of a fixed fault set, and provides a self-adaptive control method and device for a multi-working condition power system considering a discrete markov model. In the control method, firstly, a Lyapunov functional containing a discrete Markov multi-condition power system model is constructed, and a robust stochastic stability criterion meeting an H infinity norm boundary gamma is deduced by utilizing an iterative method. On the basis, a Linear Matrix Inequality (LMI) meeting the minimum variance constraint of each sub-working condition of the system and an H infinity norm boundary gamma is obtained by using a Schur supplementary theorem. Then, solving each sub-working condition matching controller of the system by utilizing the minimization problem of the linear objective function, and selecting a proper combination mode of the sub-working condition matching controllers according to the weighting coefficient calculated by the random gradient online identification method. In a specific embodiment, the correctness and the effectiveness of the self-adaptive control method and the self-adaptive control device for the multi-working-condition power system are verified by the time domain simulation of the multi-working-condition power system based on cascading failures.
In order to achieve the purpose, the invention adopts the following technical scheme.
A self-adaptive control device of a multi-working condition power system comprises a data acquisition module, a multi-working condition power system model construction module, a self-adaptive control strategy generation module, a controller combination module and a result output module which are sequentially connected, wherein,
the data acquisition module is used for acquiring network structure parameters, power grid state vectors and control output vectors and sending the acquired data to the multi-working-condition power system model construction module;
the multi-working-condition power system model building module is used for building a multi-working-condition power system model according to the collected data and determining a stability criterion according to the multi-working-condition power system model;
the adaptive control strategy generation module utilizes a minimization strategy of a linear objective function to solve each sub-working condition matching controller of the system;
the controller combination module is used for selecting the combination mode of each sub-working condition matching controller by using a weighting method;
and the result output module is used for outputting the combination mode of each sub-working condition matching controller.
And the controller combination module determines the weighting coefficient according to a random gradient online identification method, and selects a combination mode of each sub-working condition matching controller by using the weighting coefficient.
A multi-condition power system adaptive control method, comprising the steps of:
A. collecting network structure parameters, power grid state vectors and control output vectors;
B. constructing a multi-working-condition power system model according to the acquired data, and determining a stability criterion according to the multi-working-condition power system model;
C. solving each sub-working condition matching controller of the system by using a minimization strategy of a linear objective function;
D. selecting a combination mode of each sub-working condition matching controller by using a weighting method;
E. and outputting the combination mode of each sub-working condition matching controller.
Wherein the multi-working-condition power system model in the step B is as follows:
wherein xk∈RnIn the form of a state vector, the state vector,
uk∈Rpin order to control the input vector,
zk∈Rrin order to control the output vector,
process noise omegakIn the form of a zero-mean noise sequence,
{ S (t), t ≧ 0} is a Markov chain of values in finite space S ═ {1,2, …, l }, corresponding to each possible operating condition of cascading failure, its state probability pijComprises the following steps:
ΔAk(sk) For uncertain parameters, satisfy: delta Ak(sk)＝HiFk(i)Mi，
Wherein HiAnd MiFor a known matrix, a real matrix Fk(i) The uncertain parameter structure information of the system is reflected, and the conditions are met: <math>
<mrow>
<msubsup>
<mi>F</mi>
<mi>k</mi>
<mi>T</mi>
</msubsup>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<msub>
<mi>F</mi>
<mi>k</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mo>≤</mo>
<mi>I</mi>
<mo>.</mo>
</mrow>
</math>
and B, determining a stability criterion according to the multi-working-condition power system model as follows:
when u isk＝0，ωkIf all allowed uncertainties Δ a are 0iSatisfies the following conditions:
the multi-condition power system is robust and randomly stable,
wherein xk∈RnIn the form of a state vector, the state vector,
uk∈Rpin order to control the input vector,
zk∈Rrin order to control the output vector,
process noise omegakIn the form of a zero-mean noise sequence,
{ S (t), t ≧ 0} is a Markov chain that takes values in finite space S ═ {1,2, …, l } corresponding to each operating condition that a cascading failure may exist.
The minimization strategy using the linear objective function in step C is:
if there is a constanti>0，ξi>0 and a positive definite symmetric matrix Qi>0 and matrix Yi>0, then robust control satisfying the minimum variance constraint can be represented as
The linear objective function is:
wherein, <math>
<mrow>
<msub>
<mi>Ω</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mfenced open = '[' close = ']'>
<mtable>
<mtr>
<mtd>
<mrow>
<msqrt>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mn>1</mn>
</mrow>
</msub>
</msqrt>
<mi>I</mi>
</mrow>
</mtd>
<mtd>
<mrow>
<msqrt>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mn>2</mn>
</mrow>
</msub>
</msqrt>
<mi>I</mi>
</mrow>
</mtd>
<mtd>
<mo>...</mo>
</mtd>
<mtd>
<mrow>
<msqrt>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mi>l</mi>
</mrow>
</msub>
</msqrt>
<mi>I</mi>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>,</mo>
</mrow>
</math>
ψ＝diag(Q1,Q2,…,Ql)，
for each S (t) · i ∈ S, note a (S)k)、B(sk)、G(sk)、C(sk)、D(sk)、L(sk) Are respectively Ai、Bi、Gi、Ci、Di、Li，
Gamma is the degree of attenuation of the disturbance,
σijfor the variable to be determined,
Kiin order to feedback-control the gain of the gain,
βijare weight coefficients.
In step D, the combination mode of selecting the sub-working condition matching controller by using the weighting method is as follows:
K＝α1K1+α2K2+…+αlKl，
weighting factor alpha at time kiExpressed as:
wherein theta is an estimated parameter vector,
y is the output of the system and,
e is the output error and is the error,
alpha, beta and lambda are undetermined parameters,
j is the loss function.
By adopting the self-adaptive control method and the self-adaptive control device for the multi-working-condition power system, the time-varying power system model and the random gradient on-line identification method can be combined, the self-adaptive control is realized by utilizing the weighting coefficient, the problem that the design scene is difficult to match with the current working condition due to the limitation of a fixed fault set is effectively solved, and the multi-working-condition power system can be effectively controlled. When the system working condition is not matched with the sub-working condition, the combined controller can be put into use to effectively inhibit the system oscillation.
Drawings
Fig. 1 is a schematic structural diagram of an adaptive control device for a multi-condition power system in an embodiment of the present invention.
Fig. 2 is a schematic diagram of a 16-machine 68 node grid architecture, which is an exemplary application scenario for illustrating a multi-condition power system adaptive control method and apparatus according to an embodiment of the present invention.
Fig. 3 is a list of branch line state transition probabilities after line 1-31 outages in the grid configuration of fig. 2.
FIG. 4 is a listing of a set of system operating conditions after a line 1-31 fault shutdown in the grid configuration of FIG. 2.
FIG. 5 is a table listing sub-regime matching controller weighting coefficients after disconnection of lines 30-31 in the grid architecture of FIG. 2.
Fig. 6 is a diagram of dynamic response of the relative power angles of the generators G1-G8 and G1-G13 after different controllers are put into the adaptive control method and apparatus for a multi-operating-condition power system according to the embodiment of the present invention.
Fig. 7 is a diagram of dynamic response of the relative power angles of the generators G1-G8 and G1-G13 after different controllers are put into the adaptive control method and apparatus for a multi-operating-condition power system according to the embodiment of the present invention.
FIG. 8 is a listing of sub-regime matching controller weighting coefficients after disconnection of lines 35-34 in the grid architecture of FIG. 2.
Detailed Description
The present invention will be described in detail below with reference to the accompanying drawings.
Detailed exemplary embodiments are disclosed below. However, specific structural and functional details disclosed herein are merely for purposes of describing example embodiments.
It should be understood, however, that the intention is not to limit the invention to the particular exemplary embodiments disclosed, but to cover all modifications, equivalents, and alternatives falling within the scope of the disclosure. Like reference numerals refer to like elements throughout the description of the figures.
It will also be understood that the term "and/or" as used herein includes any and all combinations of one or more of the associated listed items. It will be further understood that when an element or unit is referred to as being "connected" or "coupled" to another element or unit, it can be directly connected or coupled to the other element or unit or intervening elements or units may also be present. Moreover, other words used to describe the relationship between components or elements should be understood in the same manner (e.g., "between" versus "directly between," "adjacent" versus "directly adjacent," etc.).
Before the specific implementation of the present invention is introduced, the principle of the adaptive control method and apparatus for a multi-condition power system of the present invention is introduced, and analysis and calculation are performed in combination with the technical solution of the present invention.
First, the multi-condition power system model may be represented by a discrete Markov system as:
wherein x isk∈RnIs a state vector; u. ofk∈RpIs a control input vector; z is a radical ofk∈RrIs a control output vector; process noise omegakIs a zero mean noise sequence.
{s(t),t≥0}Is a Markov chain which takes values in a finite space S ═ {1,2, …, l }, corresponds to each possible operation condition of a cascading failure, describes the evolution process of the system condition along with the failure, and has a state probability pijExpressed as:
for each S (t) · i ∈ S, note a (S)k)、B(sk)、G(sk)、C(sk)、D(sk)、L(sk) Are respectively Ai、Bi、Gi、Ci、Di、LiAnd the uncertain parameters in the formula (1) meet the matching condition:
ΔAk(sk)＝ΔAk(i)＝HiFk(i)Mi (3)
wherein HiAnd MiFor a known matrix, a real matrix F (i, t) reflects the structural information of uncertain parameters of the system and meets the following conditions:
the discrete Markov multi-working-condition power system models are established by the formulas (1) to (4), and the stability criterion of the system is further determined according to the models and the adaptive control strategy is made.
The robust stochastic stability of the discrete markov based multi-regime power system model is illustrated below.
For the multi-condition power system represented by formula (1), when u isk＝0，ωkIf all allowed uncertainties Δ a are 0iSatisfies the following conditions:
then the discrete markov random system, as in equation (1), is robust and stochastic stable.
In addition, let D, E, F be an adaptive real matrix, and matrix P > I, | F | ≦ 1, then:
(1) for any purpose>0 and the vector x, y ∈ Rn，2xTDFEy≤-1xTDDTx+yTEETy。
(2)DFE+ETFTDT≤DDT+-1ETE。
Wherein I is an identity matrix.
Precondition 1: for a given normal number gamma>0, if there is a set of positive definite symmetric matrices Qi>0, i ∈ S, such that the following set of matrix inequalities holds:
ψ＝diag(Q1,Q2,…,Ql)(8)
then when u (t) is equal to 0, the system described by the formula (1) is robust and stable randomly and meets the requirement of the disturbance attenuation degree gamma, namely:
from equation (6), for i ∈ S, the following linear matrix inequality LMI holds:
therefore, the system described in the following formula (1) is randomly stable only by determining the condition that the linear matrix inequality (10) is established, and the random stability criterion can be established.
Order toMultiplying the left and right of equation (10) by the matrix diag (P)1And I) has:
wherein,
applying the schuler's theorem, it is found from equation (11) that there is a scalar >0, so that the following equation holds:
defining a state dependence matrix Mk(sk) Comprises the following steps:
the Lyapunov function was constructed as:
Vk(Mk(sk),sk)＝tr[Mk(sk)P(sk)](P(sk)>0) (15)
then for i ∈ S, there are:
therefore, from formula (16), one can obtain:
wherein,
because:
thus, α satisfies the condition 1> α >0, and:
E{Vk+1(Mk+1(sk+1),sk+1)|xk,sk}
<αVk(Mk(sk),sk) (20)
iterating equation (20) from 0 to k, there is:
E{Vk+1(Mk+1(sk+1),sk+1)|x0,s0}<αkV0(M0(s0),s0) (21)
from formula (21):
from the formula (22):
recording:
comprises the following steps:
therefore, the system (1) is randomly stable as shown in the formulas (10) to (25). It is explained below that when ω (t) ≠ 0, the system has a degree of disturbance attenuation γ.
Pair formula (6) simultaneous left and right multiplication matrix diag (P)iI, I), the following inequality can be obtained:
using the schulk's theorem, equation (6) can be equivalent to:
the Lyapunov function was constructed as:
defining the index function as:
then, the following equations (27) to (29) can be obtained:
for i ∈ S, known from equation (22):
namely:
that is, the system (1) has the disturbance attenuation degree γ as can be seen from the expressions (26) to (33).
For the power system discrete markov multi-regime model, it is assumed that for all the allowable uncertainties the following conditions are met:
(1) presence of feedback control matrix KiMaking the system robust and stable randomly.
(2) For a given scalar y>0, initial state x00 and all ωkNot equal to 0, control output zkSatisfies the following formula:
(3) each operating condition steady state variance satisfies the following constraints:
in the formula:an upper acceptable variance bound to meet actual demand.
From equations (3) and (6), the following equations can be obtained:
it can be seen that equation (37) has the following inequality:
let Yi＝QiKiDiscrete Markov multi-condition system can be obtained by using the Schur complement theoremThe linear matrix inequality of the feedback control matrix under the robust random stabilization condition. Meanwhile, the premise 1 is utilized to explain that the system meets the similar method of the disturbance attenuation degree gamma, and the robust random stable control condition of the discrete Markov multi-working-condition system based on the allowable uncertainty condition can be obtained.
Precondition 2: given normal number gamma>0，If there is a constanti>0 and a positive definite symmetric matrix Qi>0 and matrix Yi>0, such that the following set of matrix inequalities holds:
then there is a feedback control gainThe system formula (1) is made robust and random stable.
Inequalities (41) to (44) relate to a constant variable σij、iAnd matrix variable Qi、YiThe linear matrix inequalities of (1) to (44), and the matrix variables and the constant variables satisfying the matrix inequalities of (41) to (44) constitute a convex set. In order to make the linear matrix inequality more satisfactory, the controller meeting the specific requirement can be designed by using the tolerance uncertainty condition, so that the minimum variance robust controller shown in the premise 3 can be obtained.
Precondition 3: given normal number gamma>0 and a set of weight coefficients betaij>0(i ═ 1, …, l; j ═ 1, …, n) andthe following advantagesSolving the problems:
s.t. (7)-(8)、(41)-(44)(45)
if there is a solution, thenIs a least variance robust controller for the system (1).
The optimization problem (45) is solved using a minimization problem solving method for a linear objective function, if constants existi>0，ξi>0 and a positive definite symmetric matrix Qi>0 and matrix Yi>0, then robust control satisfying the minimum variance constraint can be represented as
In actual operation, the current operation condition needs to be identified on line, and a proper controller combination is selected according to the weighting index for effective control.
The system multivariate recognition model is represented as follows:
the loss function is defined as:
let μ (t) denote a step length, and μ (t) ═ 1/r (t), r (t) ═ r (t-1) + | | Φ (t) | a2And minimizing J (theta) by using a negative gradient search method to obtain a random gradient method of the estimated parameter vector theta:
r(t)＝r(t-1)+||Φ(t)||2,r(0)＝1 (49)
the controller for defining the actual operation investment of the system can be represented as a combination form of each sub-working condition matching controller, and then the expression of the combination controller is as follows:
K＝α1K1+α2K2+…+αlKl (50)
wherein the weighting factor alpha at the time kiExpressed as:
as can be seen from the expressions (50) - (53), the multi-working-condition power system estimates the output error of each sub-working-condition model and the controlled object at each moment, adaptively corrects and updates the controller combination, and the higher the matching degree of the working condition of the controlled object and the ith working-condition model is, the higher the weighting coefficient alpha isiThe greater the weight occupied.
Therefore, as shown in fig. 1, the embodiment of the present invention discloses an adaptive control device for a multi-condition power system, which comprises a data acquisition module, a multi-condition power system model construction module, an adaptive control strategy generation module, a controller combination module and a result output module, which are connected in sequence, wherein,
the data acquisition module is used for acquiring network structure parameters, power grid state vectors and control output vectors and sending the acquired data to the multi-working-condition power system model construction module;
the multi-working-condition power system model building module is used for building a multi-working-condition power system model according to the collected data and determining a stability criterion according to the multi-working-condition power system model;
the adaptive control strategy generation module utilizes a minimization strategy of a linear objective function to solve each sub-working condition matching controller of the system;
the controller combination module is used for selecting the combination mode of each sub-working condition matching controller by using a weighting method;
and the result output module is used for outputting the combination mode of each sub-working condition matching controller.
Therefore, by using the self-adaptive control device for the multi-working-condition power system, a time-varying power system model can be combined with a random gradient on-line identification method, and self-adaptive control is realized by using the weighting coefficient, so that the problem that a design scene is difficult to match with the current working condition due to the limitation of a fixed fault set is effectively solved.
And the controller combination module determines the weighting coefficient according to a random gradient online identification method and selects a controller combination mode by using the weighting coefficient.
The invention also discloses a self-adaptive control method of the multi-working condition power system, which is matched with the self-adaptive control device of the multi-working condition power system and comprises the following steps:
A. collecting network structure parameters, power grid state vectors and control output vectors;
B. constructing a multi-working-condition power system model according to the acquired data, and determining a stability criterion according to the multi-working-condition power system model;
C. solving each sub-working condition matching controller of the system by using a minimization strategy of a linear objective function;
D. selecting a combination mode of each sub-working condition matching controller by using a weighting method;
E. and outputting the combination mode of each sub-working condition matching controller.
Wherein, the multi-working-condition power system model in the step B is as follows:
wherein xk∈RnIn the form of a state vector, the state vector,
uk∈Rpin order to control the input vector,
zk∈Rrin order to control the output vector,
process noise omegakIn the form of a zero-mean noise sequence,
{ S (t), t ≧ 0} is a Markov chain of values in finite space S ═ {1,2, …, l }, corresponding to each possible operating condition of cascading failure, its state probability pijComprises the following steps:
ΔAk(sk) For uncertain parameters, satisfy: delta Ak(sk)＝HiFk(i)Mi，
Wherein HiAnd MiFor a known matrix, a real matrix Fk(i) The uncertain parameter structure information of the system is reflected, and the conditions are met: <math>
<mrow>
<msubsup>
<mi>F</mi>
<mi>k</mi>
<mi>T</mi>
</msubsup>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<msub>
<mi>F</mi>
<mi>k</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mo>≤</mo>
<mi>I</mi>
<mo>.</mo>
</mrow>
</math>
A(sk)、B(sk)、G(sk)、C(sk)、D(sk)、L(sk) Is a parameter matrix of the system model, the values of which are determined from the system parameters, which can be determined in a manner known to the person skilled in the art, and is described hereAnd will not be described in detail.
And B, determining a stability criterion according to the multi-working-condition power system model as follows:
when u isk＝0，ωkIf all allowed uncertainties Δ a are 0iSatisfies the following conditions:
the multi-condition power system is robust and randomly stable,
wherein xk∈RnIn the form of a state vector, the state vector,
uk∈Rpin order to control the input vector,
zk∈Rrin order to control the output vector,
process noise omegakIn the form of a zero-mean noise sequence,
{ S (t), t ≧ 0} is a Markov chain that takes values in finite space S ═ {1,2, …, l } corresponding to each operating condition that a cascading failure may exist.
In addition, the minimization strategy using the linear objective function in step C is:
if there is a constanti>0，ξi>0 and a positive definite symmetric matrix Qi>0 and matrix Yi>0, then robust control satisfying the minimum variance constraint can be represented as
The linear objective function is:
wherein, <math>
<mrow>
<msub>
<mi>Ω</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mfenced open = '[' close = ']'>
<mtable>
<mtr>
<mtd>
<mrow>
<msqrt>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mn>1</mn>
</mrow>
</msub>
</msqrt>
<mi>I</mi>
</mrow>
</mtd>
<mtd>
<mrow>
<msqrt>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mn>2</mn>
</mrow>
</msub>
</msqrt>
<mi>I</mi>
</mrow>
</mtd>
<mtd>
<mo>...</mo>
</mtd>
<mtd>
<mrow>
<msqrt>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mi>l</mi>
</mrow>
</msub>
</msqrt>
<mi>I</mi>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>,</mo>
</mrow>
</math>
ψ＝diag(Q1,Q2,…,Ql)，
for each S (t) · i ∈ S, note a (S)k)、B(sk)、G(sk)、C(sk)、D(sk)、L(sk) Are respectively Ai、Bi、Gi、Ci、Di、Li，
Gamma is the degree of attenuation of the disturbance,
σijfor the variable to be determined,
Kiin order to feedback-control the gain of the gain,
βijare weight coefficients.
In step D, the selecting of the controller combination by the weighting method is:
K＝α1K1+α2K2+…+αlKl，
weighting factor alpha at time kiExpressed as:
wherein theta is an estimated parameter vector,
y is the output of the system and,
e is the output error and is the error,
alpha, beta and lambda are undetermined parameters.
J is the loss function.
In particular, the method for minimizing J (θ) by using the negative gradient search method to obtain the random gradient of the estimated parameter vector θ is as follows:
where μ (t) denotes the step size, and μ (t) ═ 1/r (t).
Therefore, the multi-working-condition power system can be effectively controlled by utilizing the self-adaptive control method of the multi-working-condition power system. When the system working condition is not matched with the sub-working condition, the combined controller can be put into use to effectively inhibit the system oscillation.
The technical effects of the adaptive control method and device for a multi-condition power system according to the embodiments of the present invention are described below with a more specific application scenario.
The present embodiment is explained based on the IEEE16 new england-new york interconnection system with 68 nodes, and the structure of the power system is shown in fig. 2.
The correctness and the effectiveness of the method are verified by taking the case that the cascading failure causes the multi-working-condition change of the system. Assuming that the initial trigger fault is the fault of the line 1-31 and exits the operation, the load is transferred to the rest lines, and the comprehensive state transition probability p of the rest lines is calculatedijFig. 3 shows a state transition probability list of the branch line after the line 1-31 is out of operation in the power system. Considering that the higher the transition probability is, the higher the probability of the line becoming the next-stage cascading failure line is when the preceding failure paths are consistent, therefore, the system behavior determined by the partial line failure with the higher state transition probability is selected as the state set of the discrete markov system, and the lower state transition probability is regarded as the wireless line failure. A list of the set of system conditions in the power system after a line 1-31 fault shutdown is shown in fig. 4.
Let S be {1,2,3,4,5}, and the probability density matrix under each sub-condition is:
firstly, the state matrix under the corresponding sub-working condition of the power system is reduced by using an SMA method, the reduced state matrix and the state transition probability matrix shown in the formula (45) are substituted, the disturbance attenuation degree gamma is taken as 0.1, and the control feedback response matrix corresponding to each model in a system model library is obtained. And according to whether the system working condition is matched with the sub-working condition model or not, the method is divided into two conditions.
After the lines 1-31 are out of operation due to faults, the lines 30-31 are disconnected, the power system is changed to the sub-condition 4, and the control effect of the control strategy is analyzed in the situation. The system working condition is identified by using a random gradient method, the weighting coefficients corresponding to the sub-working condition matching controllers are calculated by the formulas (50) - (53), the weighting coefficients of the sub-working condition matching controllers (hereinafter referred to as sub-controllers) after the system is changed are obtained as shown in fig. 6, it can be seen that the weighting coefficients of the other sub-controllers except the sub-controller 4 are very small and can be ignored because the system working condition is consistent with the characteristic model of the sub-working condition 4, fig. 6 shows the dynamic response of the power angle difference between the generators G1-G8 and G1-G13 when only the sub-controller 4, the combined controller and the non-matching controller are put into the system under the condition, wherein the dotted line shows the condition of no control addition, and the solid line shows the condition of control addition. As can be seen from fig. 6, when the system operating condition matches the characteristic model of the sub-operating condition 4, the control effect of only the input sub-controller 4 and the input weighting controller is nearly the same; when a mismatched controller is introduced, the system may not be effectively controlled and system oscillations may even be exacerbated.
After the lines 1-31 are in fault and quit operation, the lines 35-34 are disconnected, the system is not matched with any sub-working condition at the moment, the working condition of the system is identified by using a random gradient method, the weighting coefficients corresponding to all sub-controllers are calculated by the formulas (50) - (53), the weighting coefficients of the sub-controllers after the system is changed are obtained as shown in fig. 8, and it can be seen that the system needs to use the weighting coefficients to realize combination control to form a combination controller at the moment. Fig. 7 shows a dynamic response diagram of the power angle difference between the generators G1-G8, G1-G13 when the combined controller is switched in and the non-matching controller is switched in, wherein the dotted line shows the case of no control and the solid line shows the case of control. Fig. 7 shows that when the system operating condition is not matched with a sub-operating condition, the input combined controller can effectively suppress system oscillation, and as can be seen from fig. 7, when the input unmatched controller is input, the system may not be effectively controlled, and even system oscillation may be aggravated, but when the input combined controller is input, the system oscillation is effectively suppressed.
Therefore, the time-varying power system model and the random gradient online identification method are combined, the weighting coefficient is used for realizing the self-adaptive control, and the problem that the design scene is difficult to match with the current working condition due to the limitation of the fixed fault set can be effectively solved.
It should be noted that the above-mentioned embodiments are only preferred embodiments of the present invention, and should not be construed as limiting the scope of the present invention, and any minor changes and modifications to the present invention are within the scope of the present invention without departing from the spirit of the present invention.
Claims (7)
1. A self-adaptive control device of a multi-working condition power system comprises a data acquisition module, a multi-working condition power system model construction module, a self-adaptive control strategy generation module, a controller combination module and a result output module which are sequentially connected, wherein,
the data acquisition module is used for acquiring network structure parameters, power grid state vectors and control output vectors and sending the acquired data to the multi-working-condition power system model construction module;
the multi-working-condition power system model building module is used for building a multi-working-condition power system model according to the collected data and determining a stability criterion according to the multi-working-condition power system model;
the adaptive control strategy generation module utilizes a minimization strategy of a linear objective function to solve each sub-working condition matching controller of the system;
the controller combination module is used for selecting the combination mode of each sub-working condition matching controller by using a weighting method;
and the result output module is used for outputting the combination mode of each sub-working condition matching controller.
2. The adaptive control device for a multi-operating-condition power system according to claim 1, wherein the controller combination module determines the weighting coefficients according to a random gradient online identification method, and selects the combination mode of each sub-operating-condition matching controller by using the weighting coefficients.
3. A multi-condition power system adaptive control method, comprising the steps of:
A. collecting network structure parameters, power grid state vectors and control output vectors;
B. constructing a multi-working-condition power system model according to the acquired data, and determining a stability criterion according to the multi-working-condition power system model;
C. solving each sub-working condition matching controller of the system by using a minimization strategy of a linear objective function;
D. selecting a combination mode of each sub-working condition matching controller by using a weighting method;
E. and outputting the combination mode of each sub-working condition matching controller.
4. The adaptive control method for a multi-condition power system according to claim 3, wherein the multi-condition power system model in the step B is:
wherein xk∈RnIn the form of a state vector, the state vector,
uk∈Rpin order to control the input vector,
zk∈Rrin order to control the output vector,
process noise omegakIn the form of a zero-mean noise sequence,
{ S (t), t ≧ 0} is a Markov chain of values in finite space S ═ {1,2, …, l }, corresponding to each possible operating condition of cascading failure, its state probability pijComprises the following steps:
ΔAk(sk) For uncertain parameters, satisfy: delta Ak(sk)＝HiFk(i)Mi，
Wherein HiAnd MiFor a known matrix, a real matrix Fk(i) The uncertain parameter structure information of the system is reflected, and the conditions are met: <math>
<mrow>
<msubsup>
<mi>F</mi>
<mi>k</mi>
<mi>T</mi>
</msubsup>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<msub>
<mi>F</mi>
<mi>k</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
<mo>≤</mo>
<mi>I</mi>
<mo>.</mo>
</mrow>
</math>
5. the multi-condition power system adaptive control method according to claim 4, wherein the determining of the stability criterion according to the multi-condition power system model in the step B is:
when u isk＝0，ωkIf all allowed uncertainties Δ a are 0iSatisfies the following conditions:
the multi-condition power system is robust and randomly stable,
wherein xk∈RnIn the form of a state vector, the state vector,
uk∈Rpin order to control the input vector,
zk∈Rrin order to control the output vector,
process noise omegakIn the form of a zero-mean noise sequence,
{ S (t), t ≧ 0} is a Markov chain that takes values in finite space S ═ {1,2, …, l } corresponding to each operating condition that a cascading failure may exist.
6. The multi-condition power system adaptive control method according to claim 4, wherein the minimization strategy using the linear objective function in step C is:
if there is a constanti>0，ξi>0 and a positive definite symmetric matrix Qi>0 and matrix Yi>0, then robust control satisfying the minimum variance constraint can be represented as
The linear objective function is:
wherein, <math>
<mrow>
<msub>
<mi>Ω</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mfenced open = '[' close = ']'>
<mtable>
<mtr>
<mtd>
<mrow>
<msqrt>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mn>1</mn>
</mrow>
</msub>
</msqrt>
<mi>I</mi>
</mrow>
</mtd>
<mtd>
<mrow>
<msqrt>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mn>2</mn>
</mrow>
</msub>
</msqrt>
<mi>I</mi>
</mrow>
</mtd>
<mtd>
<mo>...</mo>
</mtd>
<mtd>
<mrow>
<msqrt>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mi>l</mi>
</mrow>
</msub>
</msqrt>
<mi>I</mi>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>,</mo>
</mrow>
</math>
ψ＝diag(Q1,Q2,…,Ql)，
for each S (t) · i ∈ S, note a (S)k)、B(sk)、G(sk)、C(sk)、D(sk)、L(sk) Are respectively Ai、Bi、Gi、Ci、Di、Li，
Gamma is the degree of attenuation of the disturbance,
σijfor the variable to be determined,
Kiin order to feedback-control the gain of the gain,
βijare weight coefficients.
7. The adaptive control method for a multi-condition power system according to claim 4, wherein in the step D, the combination mode of selecting the sub-condition matching controllers by using the weighting method is as follows:
K＝α1K1+α2K2+…+αlKl，
weighting factor alpha at time kiExpressed as:
wherein theta is an estimated parameter vector,
y is the output of the system and,
e is the output error and is the error,
alpha, beta and lambda are undetermined parameters,
j is the loss function.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510325481.8A CN104950675B (en) | 2015-06-12 | 2015-06-12 | Multi-state power system self-adaptation control method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510325481.8A CN104950675B (en) | 2015-06-12 | 2015-06-12 | Multi-state power system self-adaptation control method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104950675A true CN104950675A (en) | 2015-09-30 |
| CN104950675B CN104950675B (en) | 2018-01-23 |
Family
ID=54165411
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510325481.8A Active CN104950675B (en) | 2015-06-12 | 2015-06-12 | Multi-state power system self-adaptation control method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104950675B (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101282039A (en) * | 2008-05-09 | 2008-10-08 | 天津大学 | Method of polymerization of induction motor load dynamic parameter taking account of slip and load factor |
| CN101867191A (en) * | 2010-06-11 | 2010-10-20 | 中国南方电网有限责任公司超高压输电公司 | Method for adaptively regulating and controlling reference voltage value |
| CN101907685A (en) * | 2010-06-18 | 2010-12-08 | 华北电力大学 | Method for checking correctness of operating parameters of generator set capable of adapting to change of operating conditions |
| CN102664580A (en) * | 2012-05-16 | 2012-09-12 | 重庆交通大学 | Mixed smart control method of power system multi-stabilizer |
| CN103390198A (en) * | 2013-07-15 | 2013-11-13 | 温州大学 | Community self-organizing detection method for power network fault diagnosis |
| CN103661355A (en) * | 2013-12-04 | 2014-03-26 | 安徽工业大学 | Intelligent control system of hybrid electric vehicle power assembly |
| CN103779852A (en) * | 2014-03-01 | 2014-05-07 | 华北电力大学 | Regional protection system based on adaptive current protection associated factor and method thereof |
| CN103824137A (en) * | 2014-03-20 | 2014-05-28 | 北京信息科技大学 | Multi-condition fault prediction method for complex mechanical equipment |
| WO2014138896A1 (en) * | 2013-03-13 | 2014-09-18 | University Of Saskatchewan | Adaptive control technique for transmission line control devices to handle large disturbance scenarios |
| US20140365225A1 (en) * | 2013-06-05 | 2014-12-11 | DSP Group | Ultra-low-power adaptive, user independent, voice triggering schemes |
| CN104571099A (en) * | 2015-01-26 | 2015-04-29 | 北京国能日新系统控制技术有限公司 | Photovoltaic fault diagnosis system and method based on theoretical calculation and data analysis |
| CN104657915A (en) * | 2015-03-10 | 2015-05-27 | 国家电网公司 | Dynamic self-adaptive power system terminal security threat evaluation method |
-
2015
- 2015-06-12 CN CN201510325481.8A patent/CN104950675B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101282039A (en) * | 2008-05-09 | 2008-10-08 | 天津大学 | Method of polymerization of induction motor load dynamic parameter taking account of slip and load factor |
| CN101867191A (en) * | 2010-06-11 | 2010-10-20 | 中国南方电网有限责任公司超高压输电公司 | Method for adaptively regulating and controlling reference voltage value |
| CN101907685A (en) * | 2010-06-18 | 2010-12-08 | 华北电力大学 | Method for checking correctness of operating parameters of generator set capable of adapting to change of operating conditions |
| CN102664580A (en) * | 2012-05-16 | 2012-09-12 | 重庆交通大学 | Mixed smart control method of power system multi-stabilizer |
| WO2014138896A1 (en) * | 2013-03-13 | 2014-09-18 | University Of Saskatchewan | Adaptive control technique for transmission line control devices to handle large disturbance scenarios |
| US20140365225A1 (en) * | 2013-06-05 | 2014-12-11 | DSP Group | Ultra-low-power adaptive, user independent, voice triggering schemes |
| CN103390198A (en) * | 2013-07-15 | 2013-11-13 | 温州大学 | Community self-organizing detection method for power network fault diagnosis |
| CN103661355A (en) * | 2013-12-04 | 2014-03-26 | 安徽工业大学 | Intelligent control system of hybrid electric vehicle power assembly |
| CN103779852A (en) * | 2014-03-01 | 2014-05-07 | 华北电力大学 | Regional protection system based on adaptive current protection associated factor and method thereof |
| CN103824137A (en) * | 2014-03-20 | 2014-05-28 | 北京信息科技大学 | Multi-condition fault prediction method for complex mechanical equipment |
| CN104571099A (en) * | 2015-01-26 | 2015-04-29 | 北京国能日新系统控制技术有限公司 | Photovoltaic fault diagnosis system and method based on theoretical calculation and data analysis |
| CN104657915A (en) * | 2015-03-10 | 2015-05-27 | 国家电网公司 | Dynamic self-adaptive power system terminal security threat evaluation method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104950675B (en) | 2018-01-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen et al. | Resilient adaptive and H∞ controls of multi-agent systems under sensor and actuator faults | |
| Yoo | Distributed consensus tracking for multiple uncertain nonlinear strict-feedback systems under a directed graph | |
| Wang et al. | Hierarchical decomposition based consensus tracking for uncertain interconnected systems via distributed adaptive output feedback control | |
| Wang et al. | Robust adaptive neural tracking control for a class of stochastic nonlinear interconnected systems | |
| Chen et al. | Globally stable adaptive backstepping fuzzy control for output-feedback systems with unknown high-frequency gain sign | |
| CN106257873B (en) | A kind of uncatalyzed coking H ∞ fault tolerant control method of nonlinear network networked control systems | |
| Zhao et al. | Finite-time consensus for second-order stochastic multi-agent systems with nonlinear dynamics | |
| Wu et al. | Output synchronization of nonidentical linear multiagent systems | |
| Zhang et al. | Consensus acceleration in a class of predictive networks | |
| Wang et al. | Limited-budget consensus design and analysis for multiagent systems with switching topologies and intermittent communications | |
| Sakthivel et al. | Disturbance rejection for singular semi-Markov jump neural networks with input saturation | |
| Duan et al. | A distributed optimization scheme for state estimation of nonlinear networks with norm-bounded uncertainties | |
| Wen et al. | Distributed robust optimization for networked agent systems with unknown nonlinearities | |
| Xiong et al. | Event-triggered distributed online convex optimization with delayed bandit feedback | |
| CN104950676A (en) | Time-varying electric power system self-adaption control method and device | |
| CN113325717B (en) | Optimal fault-tolerant control method, system, processing equipment and storage medium based on interconnected large-scale system | |
| Montijano et al. | Step size analysis in discrete-time dynamic average consensus | |
| Ren et al. | Nonfragile H∞ fault detection for fuzzy discrete systems under stochastic communication protocol | |
| Mukherjee et al. | On robust model-free reduced-dimensional reinforcement learning control for singularly perturbed systems | |
| Yang et al. | Distributed estimation of graph spectrum | |
| CN104950675A (en) | Adaptive control method and adaptive control device for multi-working-condition power system | |
| Ding et al. | H∞ fault detection for a class of TS fuzzy model-based nonlinear networked control systems | |
| Li et al. | Further results on containment control for multi-agent systems with variable communication delay | |
| Tang et al. | Impulsive Synchronization of Complex Dynamical Networks | |
| Khalili et al. | Distributed adaptive fault-tolerant control of a class of high-order nonlinear uncertain multi-agent systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |