CN104300568A - Frequency stabilization control method for alternating-current/direct-current series-parallel system - Google Patents

Abstract

The invention discloses a frequency stabilization control method for an alternating-current/direct-current series-parallel system, which comprises the steps of designing an under-frequency load shedding optimization model considering emergency power generation control and high-voltage direct current power support, providing an objective function giving consideration to the frequency recovery performance and the load shedding amount, selecting the action frequency of an under-frequency load shedding program and the load shedding amount of each turn as control variables, carrying out optimization on a scheme by using a cloud adaptive particle swarm algorithm containing a dimension mutation operator, and realizing intelligent control for the frequency of the alternating-current/direct-current series-parallel system. The control method disclosed by the invention solves a problem of coordination and optimization between the system instability possibly caused by frequency oscillation after the alternating-current/direct-current series-parallel system breaks down and the economy of system frequency recovery; an under-frequency load shedding scheme is optimized, an optimal setting scheme is searched intelligently, the frequency recovery performance is considered, the load shedding amount is minimized, and the transient-state performance and the steady-state performance of the alternating-current/direct-current series-parallel system are improved.

Description

A kind of frequency stabilization control method of alternating current-direct current combined hybrid system

Technical field

The present invention relates to modern power systems protection and control field, particularly a kind of frequency stabilization control method of alternating current-direct current combined hybrid system.

Background technology

Frequency is as one of important indicator weighing the quality of power supply, and the exception of frequency can bring many bad impacts to electric power system, can endanger system stability time serious, the economy that influential system runs.

When occurring in electric power system gaining merit more greatly vacancy, UFLS (UFLS) is one of most important means recovering frequency stabilization.At present, low frequency deloading method mainly contains traditional low-frequency off-load and self adaptation UFLS two class.UFLS scheme traditional both at home and abroad, mainly by the operation experience of operations staff according to real system, simulation analysis of computer conclusion is adjusted.Tradition UFLS scheme takes constant time lag to fix resection, and off-line setting calculation, real-time is poor, may cause and cut or owe to cut.Get the self adaptation UFLS of frequency change rate, although can arrange operate time and load resection neatly, great majority do not consider the frequency modulation characteristic of system topology, generator and the frequency characteristic of load, may cause unnecessary load loss.In addition, when the frequency of occurrences is vibrated, frequency instantaneous rate of change is only adopted may to cause the inaccurate of measurement as signal.

Along with the development of extra-high voltage grid and super high voltage direct current electricity transmission, the grid structure become increasingly complex and changeable operational mode will make FREQUENCY CONTROL more complicated.The FREQUENCY CONTROL effect of the model shorter mention DC transmission system that the improvement project of recent domestic is studied, and consider that it controls with Emergency electric generation to coordinate UFLS when (EAGC) coordinates mutually, to strengthen system frequency stability further.For this reason, need a kind of to consider AC/DC interconnected system Emergency electric generation control action, consider that the UFLS optimization coordinating to increase generator output and direct current system emergency DC power support is adjusted model, to improve stability and the economy of system cloud gray model.

Summary of the invention

The object of the invention is to overcome the shortcoming of prior art and deficiency, a kind of frequency stabilization control method of alternating current-direct current combined hybrid system is provided.

Object of the present invention is realized by following technical scheme:

A frequency stabilization control method for alternating current-direct current combined hybrid system, comprises the step of following order:

S1. design the optimization aim of the UFLS scheme of a reflection alternating current-direct current combined hybrid system, described optimization aim takes into account the size of frequency retrieval performance and cutting load amount;

S2. design Emergency electric generation Controlling model and high voltage direct current power support controller, and both are attached in the computation model of UFLS, optimize UFLS scheme;

S3. determine the setting parameter optimizing UFLS scheme, comprise and respectively take turns operating frequency and each round cut load, and using each operating frequency and each round cut load of taking turns as control variables, determine setting program; Determine UFLS scheme relevant parameter simultaneously;

S4. in optimized algorithm, embedded in time-domain simulation program emulate, is criterion with optimization aim, obtains the Optimal tunning value in optimizing feasible zone, obtains the prioritization scheme of UFLS.

In step S1, the described optimization aim taking into account frequency retrieval performance and cutting load amount is as follows:

$F_{\mathrm{obj}}=\min ({\mathrm{\λ}}_1{S}_{f_N,f_k}+{\mathrm{\λ}}_2\mathrm{\ΔP})$

Wherein, F _objfor optimization object function value, for minimum value; Δ P is rated frequency and actual frequency area that track encloses, total cutting load amount under f-t coordinate system respectively; λ ₁, λ ₂be respectively the weights of Δ P, and λ ₁+ λ ₂=1.0; f _nfor rated frequency, f _kfor actual frequency, i.e. the instantaneous value of frequency.

Described computational methods be: wherein, n is the action wheel number of UFLS, t' _kfor the real time of kth wheel action;

The computational methods of Δ P are: wherein, Δ P _lkfor the cutting load amount of kth wheel action.

Described step S2 is specific as follows: in alternating current-direct current combined hybrid system, and the frequency control model increasing self adaptation contragradience sliding mode controller in the i of region is specially:

$\left\{\begin{array}{c}\mathrm{\Δ}{\stackrel{\·}{\mathrm{\ω}}}_{\mathrm{COI},i}=(1/M_{T,i})(\underset{u=1}{\overset{n_i}{\mathrm{\Σ}}}\mathrm{\Δ}{P}_{M,i,u}-(\mathrm{\Δ}{P}_{L,i}+\mathrm{\Δ}{P}_{\mathrm{dc}}+\mathrm{\Δ}{P}_{\mathrm{out},i})-D_{G,i}\mathrm{\Δ}{\mathrm{\ω}}_{\mathrm{COI},i});\\ \mathrm{\Δ}{P}_i=M_{T,i}{\stackrel{\·}{\mathrm{\ω}}}_{\mathrm{COI},i}\end{array}\right.$

Based on WAMS, design a self adaptation contragradience sliding mode controller model considering that Emergency electric generation controls and high voltage direct current power is supported specific as follows:

A, Emergency electric generation Controlling model:

$\left\{\begin{array}{c}\mathrm{\Δ}{\stackrel{\·}{P}}_{M,i,u}=(1/T_{T,i,u})(-\mathrm{\Δ}{P}_{M,i,u}+\mathrm{\Δ}{\mathrm{\μ}}_{i,u})\\ \mathrm{\Δ}{\stackrel{\·}{\mathrm{\μ}}}_{i,u}=(1/T_{G,i,u})(-\mathrm{\Δ}{\mathrm{\ω}}_{\mathrm{COI},i}/R_{i,u}-\mathrm{\Δ}{\mathrm{\μ}}_{i,u}+u_{G,i,u})\end{array}\right.;$

B, high voltage direct current Controlling model:

$\mathrm{\Δ}{\stackrel{\·}{P}}_{\mathrm{dc}}=(1/T_{\mathrm{dc}})(-\mathrm{\Δ}{P}_{\mathrm{dc}}+u_{\mathrm{dc}});$

Wherein, Δ ω _{cOI, i}for the increment of the center of inertia angular frequency in region, for angular acceleration, for the equivalent inertia time constant of region i, M _u.ifor the inertia time constant of the u unit in region, n _ifor the generating set quantity of region i, Δ P _{m, i, u}for the increment of the u unit output of region i, for Δ P _{m, i, u}rate of change, Δ P _l,ifor the increment of the load of region i, Δ P _{out, i}for the increment of the power that the alternating current circuit of region i transfers out, Δ P _dcfor direct current transmission power, for Δ P _dcrate of change; D _g,ifor the equivalent damping of region i, Δ P _ifor the power shortage of region i, T _{t, i, u}for the time constant of the u power generator turbine of region i, Δ μ _i,ufor the u power generator turbine valve opening of region i, for Δ μ _i,urate of change, T _{g, i, u}for the time constant of the u machine unit speed regulating device of region i, R _i,ufor the speed regulator speed droop of the u unit of region i, u _{g, i, u}for the urgent power input of the u unit of region i, T _dcfor direct current power control time constant, u _dcfor the control signal of hvdc transmission line.

In step S3, the described each wheel operating frequency as control variables, each round cut load, limit according to engineering experience:

The operating frequency that the first round takes turns substantially: f ₁<49.5Hz;

Last takes turns the operating frequency of basic wheel: f _last>47.5Hz;

Institute's cutting load amount: $\underset{k=1}{\overset{n}{\mathrm{\&Sigma;}}}\mathrm{\&Delta;}{P}_{\mathrm{lk}}\&le;\mathrm{\&Delta;}{P}_s;$

Wherein, Δ P _sfor the power shortage of system.

Described Δ P _sdetermined by following formula: wherein M _{t, sys}for the inertia time constant of system, for the rate of change of the center of inertia frequency of system.

Described step S4 specifically comprises following steps:

S41. utilize simulated program set up optimization object function and consider the UFLS computation model that Emergency electric generation controls and high voltage direct current power is supported, and carry out initialization;

S42. give certain disturbance to the model set up, produce certain power shortage, carry out simulation calculation;

S43. the power shortage of estimating system, and the measure selecting reply disturbance;

S44. given controling parameters emulating, calculates initial target functional value;

S45. adopt optimized algorithm successively to produce controling parameters, again carry out simulation calculation, calculating target function value, and upgrade optimal solution;

S46. judge the number of times whether iterations reaches default, if meet, stop emulation, current optimal solution is the optimal solution of the program; Otherwise, perform step S45.

In step S4, described optimized algorithm is the cloud APSO algorithm containing dimension mutation, specific as follows:

(1) according to the strategy of Clouds theory, molecular group self-adaptative adjustment, the Changing Pattern of inertia weight ω such as formula:

$\mathrm{\&omega;}=\left\{\begin{array}{cc}{\mathrm{\&omega;}}_1& 0<\mathrm{\&epsiv;}\&le;{\mathrm{\&epsiv;}}_1\\ {\mathrm{\&omega;}}_1-{\mathrm{\&omega;}}_2*e^{\frac{-{(f_i-\mathrm{Ex})}^2}{2{\left(E{n}^{\&prime;}\right)}^2}}& {\mathrm{\&epsiv;}}_1<\mathrm{\&epsiv;}<{\mathrm{\&epsiv;}}_2\\ {\mathrm{\&omega;}}_2& {\mathrm{\&epsiv;}}_2\&le;\mathrm{\&epsiv;}\&le;1\end{array}\right.;$

Wherein, Ex=f _av, En=(f _av-f _min)/m1, He=En/m2, m1, m2 are controling parameters, and in normal cloud model, En determines the steep of normal cloud model, and He determines the dispersion degree of water dust, produces by En '=normrnd (En, He) random number that probability is normal distribution;

ε is the current global optimum in population kth generation and the ratio of the current fitness value of certain particle; f _ifor particle X in kth time iteration _ifitness value, f _avfor the mean value of all particle fitness value summations; f _minfor the fitness value of global optimum's particle;

(2) the optimizing later stage introduce dimension Variation mechanism, its strategy as shown in the formula:

X _{id_min}＝X _id?min+rand×(X _id?max-X _id?min)rand<p _m；

Wherein, id_min representative needs the dimension of variation, the dimension that namely degree of convergence is the highest; X _{id_min}for the position of particle in the i-th d_min dimension; Rand is [0,1] upper equally distributed random number; Aberration rate p _mfor the constant on [0,1], dimension variation makes particle again be evenly distributed on the area of feasible solutions [x of this dimension _{id min}, x _{id max}] on.

Compared with prior art, tool has the following advantages and beneficial effect in the present invention:

1, the present invention sets up the computation model of optimization object function and optimal control by simulated program, and minimum optimization object function is obtained in optimizing feasible zone, to optimize UFLS scheme, intelligent search Optimal tunning scheme, consider frequency retrieval performance, minimize institute's cutting load amount, improve alternating current-direct current combined hybrid system transient performance and steady-state behaviour.

2, this method can adapt to D _g,iuncertain, Δ P _ithe unfavorable factors such as the modeling error of limited change and system, and the system that ensures has Lyapunov stability.

Accompanying drawing explanation

Fig. 1 is the flow chart of the step S4 of the frequency stabilization control method of a kind of alternating current-direct current combined hybrid system of the present invention;

Fig. 2 is the flow chart of the optimized algorithm of the frequency stabilization control method of a kind of alternating current-direct current combined hybrid system of the present invention;

Fig. 3 is after meritorious vacancy appears in alternating current-direct current combined hybrid system, the frequency curve chart of the emulation of UFLS optimization control scheme and traditional scheme dynamic response.

Embodiment

Below in conjunction with embodiment and accompanying drawing, the present invention is described in further detail, but embodiments of the present invention are not limited thereto.

A kind of frequency stabilization control method of alternating current-direct current combined hybrid system, a self adaptation contragradience sliding mode controller based on WAMS is increased in traditional low-frequency off-load (UFLS) scheme, this controller comprises Emergency electric generation and controls (EAGC) and high voltage direct current power support (HVDC), and take into account frequency retrieval performance and reduce cutting load amount, it comprises the following steps:

S1. design the optimization aim of the UFLS scheme of a reflection alternating current-direct current combined hybrid system, described optimization aim takes into account the size of frequency retrieval performance and cutting load amount; Specific as follows:

A, optimization aim have taken into account the size of frequency retrieval performance and cutting load amount, which includes the quantitative indices describing frequency retrieval performance and cutting load amount;

Design optimization object function as follows:

$F_{\mathrm{obj}}=\min ({\mathrm{\&lambda;}}_1{S}_{f_N,f_k}+{\mathrm{\&lambda;}}_2\mathrm{\&Delta;P})$

Wherein, Δ P is rated frequency and actual frequency area that track encloses and total cutting load amount under f-t coordinate system respectively; λ ₁, λ ₂be respectively the weights of Δ P, and λ ₁+ λ ₂=1.0; N is the action wheel number of UFLS, f _nfor rated frequency, f _kfor the instantaneous value of frequency, t' _kfor the real time of kth wheel action;

B, employing sub-goal portray frequency retrieval performance:

$S_{f_N,f_k}=\underset{k=1}{\overset{n}{\mathrm{\&Sigma;}}}{\&Integral;}_{t_{k-1}^{\&prime;}}^{t_k^{\&prime;}}(f_N-f_k)\mathrm{dt}$

After increase self adaptation contragradience sliding mode controller, f in above formula _krelation and some difference of relationships of indices when only considering adjustment effect of load, but find to represent by Segment Index relational expression equally after over-fitting; Therefore in simulation process, obtain the frequency values of each simulation step length in real time, under being used in f-t coordinate system, represent its size with the area that rated frequency and actual frequency track enclose, and with this area of trapezoidal integration numerical computations;

C, employing sub-goal Δ P portray the size of cutting load amount:

Δ P _lkfor the cutting load amount of kth wheel action;

S2. design Emergency electric generation Controlling model and high voltage direct current power support controller, and both are attached in the computation model of UFLS, optimize UFLS scheme; The UFLS model of described optimal control, that includes Emergency electric generation Controlling model and high voltage direct current power supports controller, shown in specific as follows:

A, set up Emergency electric generation Controlling model:

$\left\{\begin{array}{c}\mathrm{\&Delta;}{\stackrel{\&CenterDot;}{P}}_{M,i,u}=(1/T_{T,i,u})(-\mathrm{\&Delta;}{P}_{M,i,u}+\mathrm{\&Delta;}{\mathrm{\&mu;}}_{i,u})\\ \mathrm{\&Delta;}{\stackrel{\&CenterDot;}{\mathrm{\&mu;}}}_{i,u}=(1/T_{G,i,u})(-\mathrm{\&Delta;}{\mathrm{\&omega;}}_{\mathrm{COI},i}/R_{i,u}-\mathrm{\&Delta;}{\mathrm{\&mu;}}_{i,u}+u_{G,i,u})\end{array}\right.$

Wherein, Δ P _{m, i, u}for the increment of the u unit output of region i, T _{t, i, u}for the time constant of the u power generator turbine of region i, Δ μ _i,ufor the u power generator turbine valve opening of region i, T _{g, i, u}for the time constant of the u machine unit speed regulating device of region i, R _i,ufor the speed regulator speed droop of the u unit of region i, u _{g, i, u}for the urgent power input of the u unit of region i;

B, set up high voltage direct current Controlling model:

$\mathrm{\&Delta;}{\stackrel{\&CenterDot;}{P}}_{\mathrm{dc}}=(1/T_{\mathrm{dc}})(-\mathrm{\&Delta;}{P}_{\mathrm{dc}}+u_{\mathrm{dc}})$

Wherein, Δ P _dcfor direct current transmission power, T _dcfor direct current power control time constant, u _dcfor the control signal of hvdc transmission line; In the middle of, u _{g, i, u}and u _dccontrol law asked for by self adaptation contragradience Sliding Mode Control Design Method;

C, set up the computation model of UFLS prioritization scheme:

$\left\{\begin{array}{c}\mathrm{\&Delta;}{\stackrel{\&CenterDot;}{\mathrm{\&omega;}}}_{\mathrm{COI},i}=(1/M_{T,i})(\underset{u=1}{\overset{n_i}{\mathrm{\&Sigma;}}}\mathrm{\&Delta;}{P}_{M,i,u}-(\mathrm{\&Delta;}{P}_{L,i}+\mathrm{\&Delta;}{P}_{\mathrm{dc}}+\mathrm{\&Delta;}{P}_{\mathrm{out},i})-D_{G,i}\mathrm{\&Delta;}{\mathrm{\&omega;}}_{\mathrm{COI},i})\\ \mathrm{\&Delta;}{P}_i=M_{T,i}{\stackrel{\&CenterDot;}{\mathrm{\&omega;}}}_{\mathrm{COI},i}\end{array}\right.$

Wherein, Δ ω _{cOI, i}for the increment of the center of inertia angular frequency of region i, for the equivalent inertia time constant of region i, M _u.ifor the inertia time constant of the u unit of region i, n _ifor the generating set quantity of region i, P _l,ifor the load of region i, P _{out, i}for the power that the alternating current circuit of region i transfers out, D _g,ifor the equivalent damping of region i, Δ P _ifor the power shortage of region i;

S3. determine the setting parameter optimizing UFLS scheme, comprise and respectively take turns operating frequency and each round cut load, and using each operating frequency and each round cut load of taking turns as control variables, determine setting program; Determine UFLS scheme relevant parameter simultaneously;

Operating frequency, cutting load amount, time delay, differential, the steady frequency scope that allows and transient frequency scope that the setting parameter of the UFLS prioritization scheme in this step is taken turns by each form, and the operating frequency of taking turns each, this Two Variables of cutting load amount are as control variables, limit according to engineering experience:

The operating frequency that the first round takes turns substantially: f ₁<49.5Hz;

Last takes turns the operating frequency of basic wheel: f _last>47.5Hz;

Institute's cutting load amount: $\underset{k=1}{\overset{n}{\mathrm{\&Sigma;}}}\mathrm{\&Delta;}{P}_{\mathrm{lk}}\&le;\mathrm{\&Delta;}{P}_s;$

for the power shortage of system, M _{t, sys}for the inertia time constant of system, for the rate of change of the center of inertia frequency of system;

Again its dependent variable is limited:

Time delay: Δ t=0.2s;

Differential: Δ f=0.2 ~ 0.3Hz;

Steady frequency: f _ss>=49.5Hz;

Transient frequency: f _ts>=45Hz;

S4. in optimized algorithm, embedded in time-domain simulation program emulate, is criterion with optimization aim, obtains the Optimal tunning value in optimizing feasible zone, obtains the prioritization scheme of UFLS;

The simulated program of UFLS prioritization scheme, as shown in Figure 1, it specifically comprises its program circuit schematic diagram:

A, utilize MATLAB environment to write simulated program, set up optimization object function and consider that Emergency electric generation controls and the UFLS computation model of high voltage direct current power support, user can the parameter of free setting system and failure condition, and carries out initialization;

B, utilize in computation model in previous step and program to one group of controling parameters, the power shortage of estimating system, and judge the counter-measure that system adopts disturbance; Specific as follows:

(1) for Δ P _s≤ Δ P _p1contingent (Δ P _p1criterion for controller starts), do not start the UFLS program comprising controller, manual load-shedding equipment should be standby;

(2) for Δ P _s> Δ P _p1contingent, start and comprise the UFLS program of controller, as follows in detail:

A. for Δ P _p1< Δ P _s≤ Δ P _p2contingent (Δ P _p2=10%P _las the criterion that UFLS starts, P _lthe load of system, Δ P _p2> Δ P _p1>0), start Emergency electric generation and control and the support of high voltage direct current power, do not start load rejection program, but manual load-shedding equipment should be standby;

B. for Δ P _s> Δ P _p2contingent, simultaneously start-up connector and automatic load rejection program;

C, using the controling parameters of the controling parameters of "current" model as UFLS prioritization scheme, simulation calculation is carried out to system;

D, according to the simulation result calculating target function value in previous step, and to compare with the target function value calculated value of last time, as more excellent in current goal function, then current solution is updated to optimal solution;

Judge the number of times whether simulated program iterations reaches default, if meet, then stop simulated program, otherwise, perform step C;

As Fig. 2, described optimized algorithm is the cloud APSO algorithm containing dimension mutation, specific as follows:

(1) according to the strategy of Clouds theory, molecular group self-adaptative adjustment, the Changing Pattern of inertia weight ω such as formula:

$\mathrm{\&omega;}=\left\{\begin{array}{cc}{\mathrm{\&omega;}}_1& 0<\mathrm{\&epsiv;}\&le;{\mathrm{\&epsiv;}}_1\\ {\mathrm{\&omega;}}_1-{\mathrm{\&omega;}}_2*e^{\frac{-{(f_i-\mathrm{Ex})}^2}{2{\left(E{n}^{\&prime;}\right)}^2}}& {\mathrm{\&epsiv;}}_1<\mathrm{\&epsiv;}<{\mathrm{\&epsiv;}}_2\\ {\mathrm{\&omega;}}_2& {\mathrm{\&epsiv;}}_2\&le;\mathrm{\&epsiv;}\&le;1\end{array}\right.;$

Wherein, Ex=f _av, En=(f _av-f _min)/m1, He=En/m2, m1, m2 are controling parameters, and in normal cloud model, En determines the steep of normal cloud model, and He determines the dispersion degree of water dust, produces by En '=normrnd (En, He) random number that probability is normal distribution;

ε is the current global optimum in population kth generation and the ratio of the current fitness value of certain particle; f _ifor particle X in kth time iteration _ifitness value, f _avfor the mean value of all particle fitness value summations; f _minfor the fitness value of global optimum's particle;

(2) the optimizing later stage introduce dimension Variation mechanism, its strategy as shown in the formula:

X _{id_min}＝X _id?min+rand×(X _id?max-X _id?min)rand<p _m；

Wherein, id_min representative needs the dimension of variation, the dimension that namely degree of convergence is the highest; X _{id_min}for the position of particle in the i-th d_min dimension; Rand is [0,1] upper equally distributed random number; Aberration rate p _mfor the constant on [0,1], dimension variation makes particle again be evenly distributed on the area of feasible solutions [x of this dimension _{id min}, x _{id max}] on;

S5. by emulating the frequency curve of the dynamic response of UFLS scheme and the traditional UFLS scheme being optimized and controlling, as Fig. 3, the feasibility of checking optimum results.

Below by way of example, further supplementary notes are done to the present invention:

Select MATLAB as emulation platform, according to step S1 of the present invention ~ step S5, the optimization alternating current-direct current combined hybrid system of two region-4 machines being carried out to UFLS scheme is adjusted; To system parameter selection be: total generator capacity in region 1 is 1800MW, and load is 876MW.Total generator capacity in region 2 is 1000MW, and load is 1420MW.The transmission rated power of DC line is 200MW, and interregional Power Exchange contracted quantity is 420MW.Failure condition is the power shortage when 1s region 2 No. 14 buses appearance 45%; And choose λ ₁=0.4, λ ₂=0.6;

The UFLS prioritization scheme of gained and the setting parameter of traditional UFLS scheme contrast as following table:

Table 1 is taken turns substantially

Table 2 is special takes turns

As can be seen from table 1, table 2, what the difference of the UFLS setting program after control method optimization of the present invention and traditional UFLS scheme was mainly substantially to take turns adjusts.By table 1, basic wheel first run operating frequency and the cutting load amount of optimizing UFLS scheme are all high than traditional scheme, according to engineering experience, after first run action, the frequency decrease speed of the system of employing optimization UFLS scheme can be less than system when adopting traditional scheme, more be conducive to the recovery of system frequency, and, because each round cut load of prioritization scheme is successively decreased by wheel, can know above several take turns action after, the frequency of system can have better performance than frequency when adopting traditional scheme, therefore, system may have less action wheel number because frequency does not reach the operating frequency of next round than the system of employing traditional scheme, thus reach the effect of few cutting load.

As shown in Figure 3, the curve that the frequency curve that the optimization UFLS scheme adding self adaptation contragradience sliding mode controller obtains obtains than traditional UFLS scheme is more smooth.Can be known by Fig. 3, adopt the system frequency change optimizing UFLS scheme milder, and steady frequency also can be higher, in addition, the recovery time of frequency is also shorter.Consider and can obtain, when adopting optimization UFLS scheme, system has the economy of better frequency retrieval performance and frequency retrieval.In Fig. 3, abscissa representing time, ordinate represents frequency.

Above-described embodiment is the present invention's preferably execution mode; but embodiments of the present invention are not restricted to the described embodiments; change, the modification done under other any does not deviate from Spirit Essence of the present invention and principle, substitute, combine, simplify; all should be the substitute mode of equivalence, be included within protection scope of the present invention.

Claims (8)

1. a frequency stabilization control method for alternating current-direct current combined hybrid system, is characterized in that, comprises the step of following order:

S1. design the optimization aim of the UFLS scheme of a reflection alternating current-direct current combined hybrid system, described optimization aim takes into account the size of frequency retrieval performance and cutting load amount;

S2. design Emergency electric generation Controlling model and high voltage direct current power support controller, and both are attached in the computation model of UFLS, optimize UFLS scheme;

S3. determine the setting parameter optimizing UFLS scheme, comprise and respectively take turns operating frequency and each round cut load, and using each operating frequency and each round cut load of taking turns as control variables, determine setting program; Determine UFLS scheme relevant parameter simultaneously;

S4. in optimized algorithm, embedded in time-domain simulation program emulate, is criterion with optimization aim, obtains the Optimal tunning value in optimizing feasible zone, obtains the prioritization scheme of UFLS.

2. the frequency stabilization control method of alternating current-direct current combined hybrid system according to claim 1, is characterized in that, in step S1, the described optimization aim taking into account frequency retrieval performance and cutting load amount is as follows:

$F_{\mathrm{obj}}=\min ({\mathrm{\&lambda;}}_1{S}_{f_N,f_k}+{\mathrm{\&lambda;}}_2\mathrm{\&Delta;P})$

Wherein, F _objfor optimization object function value, for minimum value; Δ P is rated frequency and actual frequency area that track encloses, total cutting load amount under f-t coordinate system respectively; λ ₁, λ ₂be respectively the weights of Δ P, and λ ₁+ λ ₂=1.0; f _nfor rated frequency, f _kfor actual frequency, i.e. the instantaneous value of frequency.

3. the frequency stabilization control method of alternating current-direct current combined hybrid system according to claim 2, is characterized in that, described computational methods be: wherein, n is the action wheel number of UFLS, t' _kfor the real time of kth wheel action;

The computational methods of Δ P are: wherein, Δ P _lkfor the cutting load amount of kth wheel action.

4. the frequency stabilization control method of alternating current-direct current combined hybrid system according to claim 1, is characterized in that, described step S2 is specific as follows: in alternating current-direct current combined hybrid system, and the frequency control model increasing self adaptation contragradience sliding mode controller in the i of region is specially:

$\left\{\begin{array}{c}\mathrm{\&Delta;}{\stackrel{\&CenterDot;}{\mathrm{\&omega;}}}_{\mathrm{COI},i}=(1/M_{T,i})(\underset{u=1}{\overset{n_i}{\mathrm{\&Sigma;}}}\mathrm{\&Delta;}{P}_{M,i,u}-(\mathrm{\&Delta;}{P}_{L,i}+\mathrm{\&Delta;}{P}_{\mathrm{dc}}+\mathrm{\&Delta;}{P}_{\mathrm{out},i})-D_{G,i}\mathrm{\&Delta;}{\mathrm{\&omega;}}_{\mathrm{COI},i});\\ \mathrm{\&Delta;}{P}_i=M_{T,i}{\stackrel{\&CenterDot;}{\mathrm{\&omega;}}}_{\mathrm{COI},i}\end{array}\right.$

Based on WAMS, devise a self adaptation contragradience sliding mode controller model considering that Emergency electric generation controls and high voltage direct current power is supported specific as follows:

A, Emergency electric generation Controlling model:

$\left\{\begin{array}{c}\mathrm{\&Delta;}{\stackrel{\&CenterDot;}{P}}_{M,i,u}=(1/T_{T,i,u})(-\mathrm{\&Delta;}{P}_{M,i,u}+\mathrm{\&Delta;}{\mathrm{\&mu;}}_{i,u})\\ \mathrm{\&Delta;}{\stackrel{\&CenterDot;}{\mathrm{\&mu;}}}_{i,u}=(1/T_{G,i,u})(-\mathrm{\&Delta;}{\mathrm{\&omega;}}_{\mathrm{COI},i}/R_{i,u}-\mathrm{\&Delta;}{\mathrm{\&mu;}}_{i,u}+u_{G,i,u})\end{array}\right.;$

B, high voltage direct current Controlling model:

$\mathrm{\&Delta;}{\stackrel{\&CenterDot;}{P}}_{\mathrm{dc}}=(1/T_{\mathrm{dc}})(-\mathrm{\&Delta;}{P}_{\mathrm{dc}}+u_{\mathrm{dc}});$

Wherein, Δ ω _{cOI, i}for the increment of the center of inertia angular frequency in region, for angular acceleration, for the equivalent inertia time constant of region i, M _u.ifor the inertia time constant of the u unit in region, n _ifor the generating set quantity of region i, Δ P _{m, i, u}for the increment of the u unit output of region i, for Δ P _{m, i, u}rate of change, Δ P _l,ifor the increment of the load of region i, Δ P _{out, i}for the increment of the power that the alternating current circuit of region i transfers out, Δ P _dcfor direct current transmission power, for Δ P _dcrate of change; D _g,ifor the equivalent damping of region i, Δ P _ifor the power shortage of region i, T _{t, i, u}for the time constant of the u power generator turbine of region i, Δ μ _i,ufor the u power generator turbine valve opening of region i, for Δ μ _i,urate of change, T _{g, i, u}for the time constant of the u machine unit speed regulating device of region i, R _i,ufor the speed regulator speed droop of the u unit of region i, u _{g, i, u}for the urgent power input of the u unit of region i, T _dcfor direct current power control time constant, u _dcfor the control signal of hvdc transmission line.

5. the frequency stabilization control method of alternating current-direct current combined hybrid system according to claim 1, is characterized in that, in step S3, the described each wheel operating frequency as control variables, each round cut load, limit according to engineering experience:

The operating frequency that the first round takes turns substantially: f ₁<49.5Hz;

Last takes turns the operating frequency of basic wheel: f _last>47.5Hz;

Institute's cutting load amount: $\underset{k=1}{\overset{n}{\mathrm{\&Sigma;}}}\mathrm{\&Delta;}{P}_{\mathrm{lk}}\&le;\mathrm{\&Delta;}{P}_s;$

Wherein, Δ P _sfor the power shortage of system.

6. the frequency stabilization control method of alternating current-direct current combined hybrid system according to claim 5, is characterized in that, described Δ P _sdetermined by following formula: wherein M _{t, sys}for the inertia time constant of system, for the rate of change of the center of inertia frequency of system.

7. the frequency stabilization control method of alternating current-direct current combined hybrid system according to claim 1, is characterized in that, described step S4 specifically comprises following steps:

S41. utilize simulated program set up optimization object function and consider the UFLS computation model that Emergency electric generation controls and high voltage direct current power is supported, and carry out initialization;

S42. give certain disturbance to the model set up, produce certain power shortage, carry out simulation calculation;

S43. the power shortage of estimating system, and the measure selecting reply disturbance;

S44. given controling parameters emulating, calculates initial target functional value;

S45. adopt optimized algorithm successively to produce controling parameters, again carry out simulation calculation, calculating target function value, and upgrade optimal solution;

S46. judge the number of times whether iterations reaches default, if meet, stop emulation, current optimal solution is the optimal solution of the program; Otherwise, perform step S45.

8. the frequency stabilization control method of alternating current-direct current combined hybrid system according to claim 1, is characterized in that, in step S4, described optimized algorithm is the cloud APSO algorithm containing dimension mutation, specific as follows:

(1) according to the strategy of Clouds theory, molecular group self-adaptative adjustment, the Changing Pattern of inertia weight ω such as formula:

$\mathrm{\&omega;}=\left\{\begin{array}{cc}{\mathrm{\&omega;}}_1& 0<\mathrm{\&epsiv;}\&le;{\mathrm{\&epsiv;}}_1\\ {\mathrm{\&omega;}}_1-{\mathrm{\&omega;}}_2*e^{\frac{-{(f_i-\mathrm{Ex})}^2}{2{\left(E{n}^{\&prime;}\right)}^2}}& {\mathrm{\&epsiv;}}_1<\mathrm{\&epsiv;}<{\mathrm{\&epsiv;}}_2\\ {\mathrm{\&omega;}}_2& {\mathrm{\&epsiv;}}_2\&le;\mathrm{\&epsiv;}\&le;1\end{array}\right.;$

Wherein, Ex=f _av, En=(f _av-f _min)/m1, He=En/m2, m1, m2 are controling parameters, and in normal cloud model, En determines the steep of normal cloud model, and He determines the dispersion degree of water dust, produces by En '=normrnd (En, He) random number that probability is normal distribution;

ε is the current global optimum in population kth generation and the ratio of the current fitness value of certain particle; f _ifor particle X in kth time iteration _ifitness value, f _avfor the mean value of all particle fitness value summations; f _minfor the fitness value of global optimum's particle;

(2) the optimizing later stage introduce dimension Variation mechanism, its strategy as shown in the formula:

X _{id_min}＝X _id?min+rand×(X _id?max-X _idmin)rand<p _m；

Wherein, id_min representative needs the dimension of variation, the dimension that namely degree of convergence is the highest; X _{id_min}for the position of particle in the i-th d_min dimension; Rand is [0,1] upper equally distributed random number; Aberration rate p _mfor the constant on [0,1], dimension variation makes particle again be evenly distributed on the area of feasible solutions [x of this dimension _{id min}, x _{id max}] on.