CN109586287B - Voltage coordination control method and device based on improved adaptive model prediction control - Google Patents
Voltage coordination control method and device based on improved adaptive model prediction control Download PDFInfo
- Publication number
- CN109586287B CN109586287B CN201811498082.1A CN201811498082A CN109586287B CN 109586287 B CN109586287 B CN 109586287B CN 201811498082 A CN201811498082 A CN 201811498082A CN 109586287 B CN109586287 B CN 109586287B
- Authority
- CN
- China
- Prior art keywords
- control
- voltage
- time domain
- evaluation object
- optimization
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 73
- 230000003044 adaptive effect Effects 0.000 title claims abstract description 22
- 238000011156 evaluation Methods 0.000 claims abstract description 91
- 238000005457 optimization Methods 0.000 claims abstract description 79
- 238000005070 sampling Methods 0.000 claims abstract description 66
- 230000035945 sensitivity Effects 0.000 claims abstract description 28
- 230000033228 biological regulation Effects 0.000 claims abstract description 16
- 238000005259 measurement Methods 0.000 claims abstract description 11
- 238000005096 rolling process Methods 0.000 claims abstract description 10
- 239000011159 matrix material Substances 0.000 claims description 18
- 230000008569 process Effects 0.000 claims description 16
- 230000002123 temporal effect Effects 0.000 claims description 8
- 230000000694 effects Effects 0.000 claims description 6
- 230000009471 action Effects 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 4
- 238000011217 control strategy Methods 0.000 claims description 3
- 238000012937 correction Methods 0.000 claims description 3
- 241000764238 Isis Species 0.000 claims description 2
- 238000010206 sensitivity analysis Methods 0.000 claims description 2
- 230000001960 triggered effect Effects 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 abstract description 9
- 238000011084 recovery Methods 0.000 abstract description 8
- 230000003247 decreasing effect Effects 0.000 abstract description 4
- 238000004088 simulation Methods 0.000 description 6
- 238000013459 approach Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 3
- 239000003990 capacitor Substances 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000013386 optimize process Methods 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Feedback Control In General (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a voltage coordination control method and device based on improved adaptive model predictive control. Determining an evaluation target according to the predicted track of the system, estimating the maximum voltage regulation capacity of the alternative control on the evaluation target according to two conditions of overvoltage and undervoltage by calculating the sensitivity characteristic of the evaluation target relative to the alternative control of the system, constructing a unified time domain parameter applicability judgment index based on the result, and providing an online quantitative selection basis of the time domain parameter. And in each sampling period, updating the time domain parameters according to the adaptability judgment indexes and the real-time measurement data. Under the single fault scene, the control time domain parameters and the prediction time domain parameters are decreased progressively, so that the online calculation time can be obviously reduced, and the voltage recovery speed is accelerated. The application range of the invention is not limited to the voltage control problem of the power system, and can be expanded to the rolling time domain optimization problem based on a linear prediction model.
Description
Technical Field
The invention relates to the technical field of research on safety and stability control of a power system, in particular to a voltage coordination control method based on improved adaptive model predictive control.
Background
Model Predictive Control (MPC) is a rolling time domain multi-step control method, can well deal with uncertainty of a predictive model, and is widely applied to related researches of system voltage coordination control at present. In the existing voltage control method based on MPC, the lengths of the control time domain and the prediction time domain have significant influence on the control effect. The longer the control horizon, the smoother the control trajectory, but the longer the computation time is required. The shorter the control time domain, the fewer the control steps, and the more aggressive the strategy, but it is possible to reduce the ability of the algorithm to cope with system uncertainty. The length of the prediction time domain should be greater than or equal to the length of the control time domain. Considering the on-line calculation amount, the length of the prediction time domain should be equal to the control time domain unless the controller needs to consider the system trajectory after the control time domain.
At present, most of research works based on MPC adopt constant control time domain and prediction time domain parameters in the optimization process, but do not discuss selection basis of relevant parameters, generally rely on experience or enumeration method to determine time domain parameters off-line, and lack corresponding quantitative analysis. And the control time domain is kept constant in the process of rolling time domain optimization and cannot be adaptively adjusted according to the evolution state of the system.
Document "adaptive control time domain parameter-based voltage coordination control" realizes adaptive adjustment of a control time domain parameter in an optimization process, but the method cannot be applied to the situation that the voltage exceeds an upper limit, and has great limitation in practical application, and the method does not consider adaptive adjustment of a prediction time domain parameter.
Disclosure of Invention
The invention provides a voltage coordination control method based on improved adaptive model prediction control. The control penalty is minimized while maintaining the bus voltages within the required range. Compared with the voltage control method based on the traditional MPC, the invention provides an online quantitative selection basis of the time domain parameters, establishes an online index for evaluating the adaptability of the time domain parameters, updates the time domain parameters according to the adaptability index and the real-time measurement data in each sampling period, can obviously improve the control effect and the calculation time, and accelerates the voltage recovery speed after disturbance.
In order to achieve the purpose, the invention adopts the following technical scheme:
a voltage coordination control method based on improved adaptive model prediction control comprises the following steps:
step 1: at the sampling moment, the system voltage track in the current prediction period is obtained based on real-time measurement data, a voltage regulation capacity evaluation object is determined, and a sensitivity matrix of the evaluation object relative to system alternative control is obtained;
step 2: based on the voltage prediction result and the sensitivity matrix information, calculating the maximum evaluation object variation caused by applying control at the sampling point; aiming at two conditions of undervoltage and overvoltage, providing a unified control time domain parameter applicability judgment index based on the maximum evaluation object variable quantity, and determining a self-adaptive adjustment mode of the control time domain parameter and a prediction time domain parameter;
and step 3: establishing a voltage coordination optimization model;
and 4, step 4: solving the optimal control sequence, and applying the obtained first-step control to the system at the next sampling point moment;
and 5: repeating the steps 1-4 at the next sampling point until all the node voltages in the system are recovered to the required range, which is characterized in that,
in step 2, based on the maximum evaluation object variation, the applicability decision index of the unified control time domain parameter is provided as follows:
in the formula Is(i)mWhen the control time domain length is mtsThe suitability determination index of the ith evaluation object, where m is the number of control steps, tsIs the length of the sampling period, VrFor voltage optimization reference values, DB is a control dead band,
for the ith evaluation object under the condition of not applying control at the sampling point tn+tpnPredicted voltage amplitude at time, Va(i)mFor controlling the time domain length to be mtsUnder the condition that the maximum adjusting capacity of the candidate control on the ith evaluation object is controlled, the control strategy is
Is not triggered, so that the applicability determination index of the control time domain parameter does not have the condition that the denominator is 0, tnIs the nth sub-optimal starting time, tpnIs the predicted temporal length used in the nth sub-optimization.
The step 1 specifically comprises:
initializing optimization parameters at the sampling moment, obtaining a system voltage track in the current prediction period based on real-time measurement data, and ending the control flow if the voltage is normal; if the voltage out-of-limit condition exists in the system and the current optimization frequency is 1, determining a voltage regulation capacity evaluation object, and solving a sensitivity matrix of the evaluation object relative to the system alternative control, and if the voltage out-of-limit condition exists in the system and the current optimization frequency is greater than 1, only calculating the sensitivity matrix of the evaluation object relative to the system alternative control.
The step 1 of determining the voltage regulation capability evaluation object specifically comprises the following steps: t is tnTime, the predicted time domain [ t ] is obtainedn,tn+tpn]Inner voltage trace, tpnIs the predicted time domain length used in the nth sub-optimization at the initial control time t1If the predicted voltage amplitude at the bus i satisfies the following constraint, the voltage amplitude at the bus i is determined as a voltage regulation capability evaluation object,
in the formula VrThe reference value is optimized for the voltage in question,
the predicted voltage amplitude of the voltage at the bus i at the current predicted time domain end moment under the condition of no control, DB is a control dead zone, VmaxAnd VminRespectively is the upper limit and the lower limit of the system voltage amplitude, and satisfies Vmax＝Vr+DB/2，Vmin＝Vr-DB/2。
Wherein the evaluation subject is at t in step 1n+tpnThe values of the time instants are expressed with respect to the sensitivity matrix of each step control as:
in the formula Sk(i,j)Representing the trajectory sensitivity, N, of the i-th evaluation object relative to the j-th candidate control variable in the k-th control stepiAnd NjRespectively representing the number of the evaluation objects and the alternative control variables, and the sensitivity is calculated on line or calculated off line and matched on line.
Wherein the step 2 comprises the steps of:
step 2-1: based on the voltage prediction results and the trajectory sensitivity analysis, the maximum evaluation target variation amount caused by applying the control at the sampling point is calculated as follows:
when in use
During optimization, the jth candidate control variable is increased to increase or decrease the voltage at the bus i to be close to the reference value, and under the constraint of single-step control variation and control variables, when the variation of the jth control variable is measured to obtain the maximum positive value, the maximum value of the variation of the ith evaluation object to the reference value direction is as follows:
wherein,
for the ith evaluation object under the condition of not applying control at the sampling point tn+tpnPredicted voltage amplitude, Δ V, at a timek(i,j)max(tn+tpn) Is the ith evaluation object at (t)n+tpn) Maximum value of the variation of the time point towards the reference value, which variation is caused by the action of the jth candidate control variable in the kth step control, DeltaumaxThe maximum value of the variable quantity of the control variable among sampling points is set for processing the deviation of the control effect caused by the mismatching problem of the model in the rolling optimization process, and the delta u(j)maxRepresents the maximum value of variation of the jth candidate controlled variable, umaxVector formed by the maximum values of the alternative control variables, u(j)maxRepresents the maximum value, u, of the jth candidate control variablek(j)Representing the value of the jth candidate control variable in the kth control;
similarly, when
In the optimization, the jth candidate control variable is reduced to reduce or increase the voltage at the bus i to be close to the reference value, and the maximum value of the variation of the ith evaluation object to the reference value is as follows:
in the formula uminVector formed by minimum values of alternative control variables, u(j)minRepresents the minimum value of the jth candidate control variable.
Considering the variation of all the alternative control measures, the maximum variation of the ith evaluation object toward the reference value is:
in the formula, delta Vk(i)max(tn+tpn) In order to consider all the alternative control variation amounts, the maximum variation amount, DeltaV, of the i-th evaluation object toward the reference value thereofk(i,j)max(tn+tpn) Showing that the ith evaluation object is (t) in the case of considering the single-step control variation amount obtained in the step 2-1n+tpn) Maximum value of variation of time to the reference value direction; control sequence (Deltau u) taking into account m-step control applied in the current control time domain1,△u2,…,△um) At sampling point tn+tpnThe maximum adjustment capacity of the evaluation object at the time is estimated as follows:
in the formula Va(i)mFor controlling the time domain length to be mtsMaximum adjustment capability of the alternative control to the i-th evaluation object in the case, tsFor the length of the sampling period to be,
for the ith evaluation object under the condition of not applying control at the sampling point tn+tpnThe predicted voltage amplitude at the time.
The predicted voltage value when no control is applied is less than the voltage-optimized reference value, i.e.
The applicability determination index of the control time domain parameter is expressed as:
the predicted voltage value when no control is applied is greater than the voltage-optimized reference value, i.e.
The applicability determination index of the control time domain parameter is expressed as:
wherein in the step 2, if the suitability is judged by the index IsmIn the presence of an element Is(i)mIf the control step number is less than or equal to 0, the control step number m is continuously increased until IsmAll elements in the composition satisfy Is(i)m>0, and selecting mtsFor the currently optimised control time-domain length and the next optimised prediction time-domain length, i.e. mts＝tcn＝tp(n+1)，tcnFor controlling the length of the time domain in the nth sub-optimization, tp(n+1)Is the predicted temporal length used in the n +1 th sub-optimization.
If Is(i)m>0, indicating that the current control time domain parameter setting has the condition of returning the evaluation object I to the control dead zone, and enabling Ism＝(Is(1)m,Is(2)m,……,Is(Ni)m) M has an initial value of 1, if
M continues to be increased until IsmAll elements in the composition satisfy Is(i)m>0, and selecting mtsFor the currently optimised control time-domain length and the next optimised prediction time-domain length, i.e. mts＝tcn＝tp(n+1)，tcnFor controlling the length of the time domain in the nth sub-optimization, tp(n+1)Is the predicted temporal length used in the n +1 th sub-optimization.
If the alternative control cannot return all the evaluation objects to the control dead zone range, the control step number m is reset to 1, and the alternative control adds a high-cost control measure considering load shedding; it is assumed that the system is equipped with sufficient reactive compensation and voltage regulation devices in the set voltage correction control scenario.
In step 3, the objective is to minimize the voltage offset and the control cost of each bus, and the voltage coordination optimization model is as follows:
the model constraints are:
Np(n+1)＝Ncn＝tcn/ts
umin≤ug≤umax
|△ug|≤△umax
in the formula NpnAnd NcnThe number of sampling points, t, in the predicted time domain period and the control time domain in the nth optimization respectivelycn＝Ncnts,tcnFor the control time domain length, t, used in the nth sub-optimizationsIn order to be the sampling period of time,
is the predicted voltage amplitude, u, at sample point fgFor the value of the control variable at the sampling point g,. DELTA.ugQ, R are punishment weight matrixes of voltage offset and control cost respectively, which are control variable quantities at a sampling point g; u. of0An initial value representing a control variable; vmaxAnd VminRespectively is the upper limit and the lower limit of the system voltage amplitude, and satisfies Vmax＝Vr+DB/2，Vmin＝Vr-DB/2；
The third constraint described above is a voltage constraint in the prediction time domain,
for the current prediction timeThe voltage amplitude at the sampling point of the domain ending time, on the premise that the control variable meets the constraint, the objective function only specifies the voltage constraint of the predicted time domain ending time,
the fourth constraint defines the control variable ugThe maximum value and the minimum value of the input are calculated,
the fifth constraint defines a single-step control variation Δ ugMaximum absolute value of (a).
In the step 4, if the optimal control sequence at the sampling point is obtained as (delta u)1,△u2,…,△um) Then at tn+tsControlling delta u in the first step in the optimal control sequence at any moment1Is applied to the system.
The invention also provides a voltage coordination control device based on the improved adaptive model prediction control, which is used for implementing the voltage coordination control method.
Compared with the prior art, the method mainly has the following advantages:
(1) based on the track change rule of the voltage control process, a self-adaptive adjustment method of the prediction time domain in the rolling optimization process is provided.
(2) And aiming at two conditions of overvoltage and undervoltage, a unified control time domain parameter applicability judgment index is established.
(3) And each suboptimal optimal control time domain and prediction time domain are determined based on the real-time measurement information, so that the calculation speed and the voltage recovery speed can be obviously improved.
(4) The data required for determining the self-adaptive time domain parameters are all derived from the process quantity of a prediction link in the voltage coordination control, only simple algebraic calculation is involved, and extra calculation time is hardly consumed.
Drawings
Fig. 1 is a schematic diagram of a model predictive control principle.
FIG. 2 is a flow chart illustrating a voltage coordination control method based on the improved adaptive model predictive control according to the present invention.
Fig. 3 is a schematic diagram of a topology of an IEEE39 node system.
FIG. 4 illustrates the example system voltage trace of FIG. 3 without control applied.
FIG. 5 is a voltage trace of the exemplary system shown in FIG. 3 after the voltage coordination control method of the present invention is employed.
FIG. 6 is a time domain parameter selection during optimization of the exemplary system of FIG. 3 using the voltage coordination control method of the present invention.
Detailed Description
The invention is further described with reference to the following figures and examples.
The above description is illustrative of the invention and does not limit the scope of the invention in any way, and equivalents or obvious variations of its technical features are intended to fall within the scope of the invention.
The Model Predictive Control (MPC) principle is shown in fig. 1. Where the control input is u and the trajectory without further control applied is
Predicted trajectory is
The actual trajectory after the control is applied is shown in solid lines.
The time domain parameters of the MPC method include a prediction time domain, a control time domain and a sampling period. In the prediction time domain, the optimal control sequence to be applied in the control time domain is solved by minimizing an objective function, and the optimization process is repeated repeatedly with the sampling period as an interval.
The invention selects the time domain parameters of MPC according to the following principle:
sampling period ts
MPC sampling period tsThe length of the sampling period t is required to meet the time requirement of measurement information acquisition, processing and optimization decision, and the requirement of the measurement information acquisition, processing and optimization decision is not greatly changed in the whole optimization process, so that the constant sampling period t is selected by the inventions10s, the same parameters as those in the two-stage voltage control of the actual system are selected.
Controlling the time domain tc
In the case of a constant sampling period, the optimal control time domain length is related to the evolution trajectory of the voltage and the sensitivity characteristics of the voltage to the control measures. The invention selects a linearized voltage prediction model, and if no new disturbance changing the system topology occurs in the optimization process, the sensitivity change is very small, so the obtained optimal control variation is in direct proportion to the monitored real-time voltage deviation.
In the case of large voltage deviations, the length of the control time domain should take into account the voltage recovery speed, the on-line computation time, and the ability to handle system model mismatch and uncertainty. Along with the optimization, the system voltage gradually approaches to the reference value, and the absolute value of the corresponding optimal control variable quantity is correspondingly reduced. In this case, selecting a smaller control time domain length can reduce the number of online optimization and the amount of calculation.
Predicting the time domain tp
The MPC calculates the system response track caused by the control action in the prediction time domain, so that the length of the prediction time domain is larger than or equal to the control time domain. When the control target voltage gradually approaches to the steady-state value, the necessity of obtaining the system trajectory outside the control time domain is reduced. Therefore, the length of the prediction time domain at the initial control time is greater than the control time domain. As the rolling horizon optimization progresses, the length of the prediction horizon gradually approaches the length of the control horizon to reduce unnecessary online computation.
Fig. 2 shows a flow chart of a voltage coordination control method based on improved adaptive model prediction control proposed by the present invention, which comprises the following steps:
step 1: at the sampling moment, initializing optimization parameters, and obtaining a current prediction time domain [ t ] based on real-time measurement datan, tn+tpn]If the voltage is normal, the control flow is ended. And if the voltage out-of-limit condition exists in the system and the current optimization frequency is 1, determining a voltage regulation capacity evaluation object, and solving a sensitivity matrix of the evaluation object relative to the system alternative control. If the voltage out-of-limit condition exists in the system and the current optimization times are more than 1, the evaluation object is not determined again,but the sensitivity matrix of the evaluation object determined by the first optimization with respect to the alternative control is updated.
The determination method of the evaluation object in the step 1 comprises the following steps: t is tnTime, the predicted time domain [ t ] is obtainedn,tn+tpn]Inner voltage trace, tpnIs the predicted temporal length used in the nth sub-optimization. At an initial control time t1If the predicted voltage amplitude at the bus i satisfies the following constraint, the voltage amplitude at the bus i is determined as a voltage regulation capability evaluation object,
in the formula VrThe reference value is optimized for the voltage in question,
the predicted voltage amplitude of the voltage at the bus i at the current predicted time domain end moment under the condition of no control, DB is a control dead zone, VmaxAnd VminRespectively is the upper limit and the lower limit of the system voltage amplitude, and satisfies Vmax＝Vr+DB/2，Vmin＝Vr-DB/2。
The evaluation object is at tn,tn+tpn]The value of the time of day with respect to the sensitivity matrix for each step control can be expressed as:
in the formula Sk(i,j)Representing the trajectory sensitivity, N, of the i-th evaluation object relative to the j-th candidate control variable in the k-th control stepiAnd NjRespectively representing the number of the evaluation object and the alternative control variable, and the sensitivity information can be calculated on line, can be calculated off line and can be matched on line.
Step 2: based on the voltage prediction result and the sensitivity matrix information, calculating the maximum evaluation object variation caused by applying control at the sampling point; needleFor under-voltage and over-voltage conditions, providing a unified control time domain parameter applicability judgment index I based on the maximum evaluation object variable quantitysmIf I issmIn the presence of an element Is(i)mIf the control step number is less than or equal to 0, the control step number m is continuously increased until IsmAll elements in the composition satisfy Is(i)m>0, and selecting mtsFor the currently optimised control time-domain length and the next optimised prediction time-domain length, i.e. mts＝tcn＝tp(n+1)，tcnFor controlling the length of the time domain in the nth sub-optimization, tp(n+1)Is the predicted temporal length used in the n +1 th sub-optimization.
Based on the voltage prediction result and the sensitivity matrix information, the maximum evaluation object variation amount caused by applying control at the sampling point k is calculated as follows:
when in use
When optimization is to increase the jth alternate control variable to be high (when S isk(i,j)Not less than 0 and
when S is decreased) or decreased (when S isk(i,j)Is less than or equal to 0 and
and then) the voltage at the bus i is close to the reference value, and under the constraint of single-step control variation and the control variable, when the variation of the jth candidate control variable is taken to be the maximum positive value, the maximum value of the variation of the ith evaluation object to the reference value direction is as follows:
wherein,
for the ith evaluation object under the condition of not applying control at the sampling point tn+tpnPredicted voltage amplitude, Δ V, at a timek(i,j)max(tn+tpn) For evaluation of subject i in (t)n+tpn) The maximum value of the variation of the time to the reference value direction of the time is caused by the action of the jth candidate control variable at the sampling point k. Delta umaxThe maximum value of the variable quantity of the sampling point control variable is set to process the deviation of the control effect caused by the model mismatching problem in the rolling optimization process. Delta u(j)maxRepresents the maximum value of the variation of the jth candidate control variable. u. ofmaxVector formed by the maximum values of the alternative control variables, u(j)maxRepresents the maximum value, u, of the jth candidate control variablek(j)Represents the value of the jth candidate control variable at sample point k.
Similarly, when
When optimization is to decrease the jth candidate control variable to decrease (when Sk(i,j)>0 and
when S is increased) or increased (when S is increased)k(i,j)<0 and
time) the voltage at the bus i is made to approach the reference value, and the maximum value of the variation of the ith evaluation object to the reference value is as follows:
wherein u is(j)minRepresents the minimum value of the jth candidate control variable.
Considering the variation of all the alternative control measures at the sampling point k, the maximum value of the variation of the ith evaluation object is:
in the formula, delta Vk(i)max(tn+tpn) In order to consider all the candidate control variation amounts, the variation amount of the i-th evaluation target is the maximum value. Delta Vk(i,j)max(tn+tpn) The maximum value of the variation of the i-th evaluation object toward the reference value thereof in consideration of the single control variation obtained in step 2-1 is shown.
Control sequence (Deltau u) taking into account m-step control applied in the current control time domain1,△u2,…,△um) At sampling point tn+tpnThe maximum adjustment capability of the time to the evaluation subject is estimated as follows:
in the formula Va(i)mFor controlling the time domain length to be mtsThe alternative in this case controls the maximum adjustment capability for the i-th evaluation object,
for the ith evaluation object under the condition of not applying control at the sampling point tn+tpnThe voltage amplitude at the moment.
Based on the voltage regulation capability estimation of an evaluation object, the applicability judgment index of the control time domain parameter is provided:
in the formula Is(i)mWhen the control time domain length is mtsApplicability judgment index, V, of the ith evaluation targeta(i)mFor controlling the time domain length to be mtsUnder the condition, the maximum adjusting capacity of the alternative control on the ith evaluation target is obtained; the control strategy is as follows
Since the trigger is not generated, the applicability determination index of the control time domain parameter does not have a denominator of 0.
The predicted voltage value is less than its reference value when no control is implemented, i.e.
The applicability determination index of the control time domain parameter is expressed as:
the predicted voltage value is greater than its reference value when no control is implemented, i.e. when
The applicability determination index of the control time domain parameter is expressed as:
if Is(i)m>0, indicating that the current control time domain parameter setting has the condition of returning the evaluation object I to the control dead zone, and enabling Ism＝(Is(1)m,Is(2)m,……,Is(Ni)m) M has an initial value of 1, if IsmIn
M continues to be increased until IsmAll elements in the composition satisfy Is(i)m>0, and selecting mtsFor the currently optimised control time-domain length and the next optimised prediction time-domain length, i.e. mts＝tcn＝tp(n+1)，tcnThe length of the time domain is controlled for the nth sub-optimization.
If the alternative control cannot return all the evaluation objects to the control dead zone range, resetting m to 1, and adding high-cost control measures such as load shedding and the like into the alternative control; it is assumed that the system is equipped with sufficient reactive compensation and voltage regulation devices in the set voltage correction control scenario.
And step 3: and establishing a voltage coordination optimization model. Aiming at minimizing the voltage deviation of each bus and the control cost, the voltage coordination optimization model is as follows:
s.t.
Np(n+1)＝Ncn＝tcn/ts
umin≤ug≤umax
|△ug|≤△umax
in the formula NpnAnd NcnThe number of sampling points, t, in the predicted time domain period and the control time domain in the nth optimization respectivelycn＝Ncnts,tcnFor the control time domain length, t, used in the nth sub-optimizationsIn order to be the sampling period of time,
is the predicted voltage amplitude, u, at sample point fgFor the value of the control variable at the sampling point g,. DELTA.ugQ, R are punishment weight matrixes of voltage offset and control cost respectively, which are control variable quantities at a sampling point g; u. of0Representing the initial value of the control variable.
The third constraint described above is a voltage constraint in the prediction time domain,
for the voltage amplitude at the sampling point of the current predicted time domain end time, the objective function only specifies the voltage constraint of the predicted time domain end time on the premise that the control variable meets the constraint,
the fourth constraint specifies the maximum and minimum values of the control variable input,
the fifth constraint mentioned above specifies the maximum value of the absolute value of the single-step control variation.
When the voltage out-of-limit condition of the system is monitored, the voltage regulation control measures should be implemented as soon as possible so as to reduce the negative influence on the power grid and the load caused by the voltage out-of-limit and prevent the continuous deterioration of the voltage. Because the problem of mismatching of the prediction model is inevitable in the application of an actual system, the problems that optimization efficiency is reduced and even a generator is over-excited due to over-control and the like can be caused by adopting a single-step control method or adopting an excessively large variation of single-step control, and therefore a single-step control variation constraint (a fifth constraint) is added into the optimization model.
And 4, step 4: solving for an optimal control sequence (Deltau u)1,△u2,…,△um) At tn+tsControlling delta u in the first step in the optimal control sequence at any moment1Is applied to the system.
And 5: and (4) repeating the steps 1-4 at the next sampling point until all the node voltages in the system are recovered to be within the required range.
The invention verifies the effectiveness of the voltage coordination control method provided by the invention through simulation of an IEEE39 node arithmetic system. Wherein the voltage reference value VrThe voltage control dead band DB is 0.1p.u. 1, i.e. the voltage should be kept at 0.95p.u., 1.05p.u.]Within the range. Sampling period ts10 s. Initial prediction time domain t p160 s. The diagonal elements of the penalty weight matrices Q and R for voltage offset and control cost are both set to 1.
IEEE39 node system topology as shown in fig. 3, alternative control measures include a parallel capacitor bank at bus 8, AVR setpoints for all 10 generators, load tap changers OLTCs connecting buses 11, 12, 13 and 19, 20. In each step of control, the absolute value upper limits of the control variation of the parallel capacitor bank, the AVR setpoint, and the OLTCs are 0.1p.u., 0.04p.u., and 1.67%, respectively. The load model in the simulation adopts a dynamic index recovery model [82,83 ]]Model parameter αs＝βs＝0,αt＝βt＝2,Tp＝T q60. The system trajectory prediction and the selection of the adaptive time domain parameters are based on a quasi-steady state system model, and the actual control effect of the proposed method is verified by full-time-domain simulation of a detailed system model.
When t is 10s, the generator trips due to a fault at the bus 32. As shown in fig. 4, the voltage at the bus bars 4, 7, 8, 12 drops below the lower voltage threshold of 0.95p.u. after a fault. The voltage coordination control method based on the improved adaptive model prediction control provided by the invention comprises the following steps:
step 1: the initialization optimization times n is 1, and m is 1. After a fault occurs, at the time when a sampling point t is 10s, a control center collects system real-time data transmitted by a wide area measurement system after disturbance, power grid real-time data of a control initial time is obtained after state estimation, and a predicted voltage track in a [10s,70s ] time period is obtained based on implicit trapezoidal time domain simulation by taking the system real-time data as an initial value point.
And extracting a Jacobian matrix obtained in the time domain simulation based on the implicit trapezoidal method, and calculating to obtain the track sensitivity of the system node voltage relative to the control measures. And (4) taking the track sensitivity value of the predicted time domain ending time at the time of 70S to form a sensitivity matrix S. And (3) shifting the voltage amplitude values at the buses 4, 7, 8 and 12 out of the control dead zone at the time of predicting the time domain end by the first rolling optimization, and determining the voltage amplitude values as an evaluation target.
Step 2: and (3) calculating the maximum evaluation object variation caused by applying control at the sampling point based on the voltage prediction result and the sensitivity matrix information obtained in the step (1). And solving the applicability judgment index of the currently optimized control time domain parameter based on the maximum evaluation object variation.
Parameter applicability determination index I in this examplesmSee table 1, n represents the current optimization times. When the optimization number n is equal to 1, the value of the optimization step number m can be increased to 3, so that I can be enableds3All the elements in the optimization are larger than 0, so that the current optimization control step number is selected to be 3, the control time domain is selected to be 30s, and the prediction time domain when the next optimization n is 2 is selected to be 30 s.
TABLE 1 determination of parameter applicability in optimization Process
And step 3: and (3) establishing a voltage coordination control model based on the control time domain and the prediction time domain parameters determined in the step (2) and based on the voltage coordination control method for improving the adaptive model prediction control.
And 4, step 4: solving for the Current control time Domain [10s,40s ]]Inner optimal control sequence (Deltau u)1,△u2,△u3). Considering communication and calculation delay, the control time domain initial time is calculated as a control sequence to be applied after one sampling period, namely the control is respectively carried out at the 20s, 30s and 40s moments. In this sub-optimization, Δ u is applied only at the time t-20 s1. The control measures for the optimized implementation are shown in table 2. U in watchref,GiAVR reference Voltage set Point, n, representing Generator it(i-j)Representing the transformation ratio of the on-load tap changer connecting the bus i and the bus j.
TABLE 2 implementation of control measures
And 5: at the next control time domain initial time, i.e. t2＝t1+tsAnd repeating the steps until the voltage amplitude of each node in the prediction time domain is restored to be within the required range at the moment of 20 s. After the control method provided by the invention is adopted, the voltage track of the system is shown in figure 5, the suboptimal parameter applicability judgment indexes are shown in table 1, the concrete implemented control measures are shown in table 2, and after three rolling controls, the voltages of all nodes of the system are completely restored to the required range. The time domain selection in the optimization process is shown in fig. 6.
To further verify the effectiveness of the proposed method, the present section will adopt a constant time domain parameter (t) based on the improved adaptive model predictive control voltage coordination control method and the traditional methodc＝30s,t p60s) MPC voltage coordination control methodThe methods were compared. Defining the optimized process voltage deviation index as follows:
in the formula NlThe number of nodes is loaded for the system. Integration time Δ t 300s, Δ VoffsetThe smaller the value of (b), the faster the voltage recovery rate is represented, and the better the control performance is.
The selection of the simulation fault scene, the system parameters and other controller parameters is not changed. The comparison result of the control performance based on the constant time domain parameter MPC and the method of the present invention is shown in Table 3.
TABLE 3 optimized Performance comparison
From the comparison results, it can be seen that the present invention has outstanding advantages in terms of control time and voltage recovery speed compared to the conventional MPC. The present invention employs a decreasing prediction time domain and a control time domain. When n is 1, the same time domain parameters are used in both methods, and the optimization time is basically consistent. With the increase of n, the time consumption of single optimization of the invention is obviously reduced, and the main reason is that the time length of numerical integration is shortened due to the shortening of the prediction time domain, and the control steps contained in the solution control sequence are reduced. Meanwhile, the voltage recovery speed is accelerated, the control times required in the whole optimization process are reduced, and the online calculation time can be obviously reduced.
Finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that changes may be made in the embodiments and/or equivalents thereof without departing from the spirit and scope of the invention. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims (14)
1. A voltage coordination control method based on improved adaptive model prediction control comprises the following steps:
step 1: at the sampling moment, the system voltage track in the current prediction period is obtained based on real-time measurement data, a voltage regulation capacity evaluation object is determined, and a sensitivity matrix of the evaluation object relative to system alternative control is obtained;
step 2: based on the voltage prediction result and the sensitivity matrix information, calculating the maximum evaluation object variation caused by applying control at the sampling point; aiming at two conditions of undervoltage and overvoltage, providing a unified control time domain parameter applicability judgment index based on the maximum evaluation object variable quantity, and determining a self-adaptive adjustment mode of the control time domain parameter and a prediction time domain parameter;
and step 3: establishing a voltage coordination optimization model;
and 4, step 4: solving the optimal control sequence, and applying the obtained first-step control to the system at the next sampling point moment;
and 5: repeating the steps 1-4 at the next sampling point until all the node voltages in the system are recovered to the required range, which is characterized in that,
in step 2, based on the maximum evaluation object variation, the applicability decision index of the unified control time domain parameter is provided as follows:
in the formula Is(i)mWhen the control time domain length is mtsThe suitability determination index of the ith evaluation object, where m is the number of control steps, tsIs the length of the sampling period, VrFor voltage optimization reference values, DB is a control dead band,
for the ith evaluation without control appliedObject at sampling point tn+tpnPredicted voltage amplitude at time, Va(i)mFor controlling the time domain length to be mtsUnder the condition that the maximum adjusting capacity of the candidate control on the ith evaluation object is controlled, the control strategy is
Is not triggered, so that the applicability determination index of the control time domain parameter does not have the condition that the denominator is 0, tnIs the nth sub-optimal starting time, tpnIs the predicted temporal length used in the nth sub-optimization.
2. The method of claim 1, wherein the method comprises:
the step 1 specifically comprises:
initializing optimization parameters at the sampling moment, obtaining a system voltage track in the current prediction period based on real-time measurement data, and ending the control flow if the voltage is normal; if the voltage out-of-limit condition exists in the system and the current optimization frequency is 1, determining a voltage regulation capacity evaluation object, and solving a sensitivity matrix of the evaluation object relative to the system alternative control, and if the voltage out-of-limit condition exists in the system and the current optimization frequency is greater than 1, only calculating the sensitivity matrix of the evaluation object relative to the system alternative control.
3. The method of claim 2, wherein the method comprises:
the step 1 of determining the voltage regulation capability evaluation object specifically comprises the following steps: t is tnTime, the predicted time domain [ t ] is obtainedn,tn+tpn]Inner voltage trace, tpnIs the predicted time domain length used in the nth sub-optimization at the initial control time t1If the predicted voltage amplitude at the bus i satisfies the following constraint, the voltage amplitude at the bus i is determined as a voltage regulation capability evaluation object,
in the formula VrThe reference value is optimized for the voltage in question,
the predicted voltage amplitude of the voltage at the bus i at the current predicted time domain end moment under the condition of no control, DB is a control dead zone, VmaxAnd VminRespectively is the upper limit and the lower limit of the system voltage amplitude, and satisfies Vmax＝Vr+DB/2，Vmin＝Vr-DB/2。
4. The method of claim 3, wherein the method comprises:
wherein the evaluation subject is at t in step 1n+tpnThe values of the time instants are expressed with respect to the sensitivity matrix of each step control as:
in the formula Sk(i,j)Representing the trajectory sensitivity, N, of the i-th evaluation object relative to the j-th candidate control variable in the k-th control stepiAnd NjRespectively representing the number of the evaluation objects and the alternative control variables, and the sensitivity is calculated on line or calculated off line and matched on line.
5. The method of claim 4, wherein the method comprises:
wherein the step 2 comprises the steps of:
step 2-1: based on the voltage prediction results and the trajectory sensitivity analysis, the maximum evaluation target variation amount caused by applying the control at the sampling point is calculated as follows:
when in use
During optimization, the jth candidate control variable is increased to increase or decrease the voltage at the bus i to be close to the reference value, and under the constraint of single-step control variation and control variables, when the variation of the jth control variable is measured to obtain the maximum positive value, the maximum value of the variation of the ith evaluation object to the reference value direction is as follows:
wherein,
for the ith evaluation object under the condition of not applying control at the sampling point tn+tpnPredicted voltage amplitude at time, Δ Vk(i,j)max(tn+tpn) Is the ith evaluation object at (t)n+tpn) The maximum value of the variation of the time to the reference value direction, which is caused by the action of the jth candidate control variable in the kth step control, ismaxThe maximum value of the variable quantity of the control variable among sampling points is set for processing the deviation of the control effect caused by the mismatching problem of the model in the rolling optimization process, and the maximum value is delta u(j)maxRepresents the maximum value of variation of the jth candidate controlled variable, umaxVector formed by the maximum values of the alternative control variables, u(j)maxRepresents the maximum value, u, of the jth candidate control variablek(j)Representing the value of the jth candidate control variable in the kth control;
similarly, when
In the optimization, the jth candidate control variable is reduced to reduce or increase the voltage at the bus i to be close to the reference value, and the maximum value of the variation of the ith evaluation object to the reference value is as follows:
in the formula uminVector formed by minimum values of alternative control variables, u(j)minRepresents the minimum value of the jth candidate control variable.
6. The method of claim 5, wherein the method comprises:
considering the variation of all the alternative control measures, the maximum variation of the ith evaluation object toward the reference value is:
in the formula,. DELTA.Vk(i)max(tn+tpn) In order to consider all the alternative control variation amounts, the maximum variation amount, Δ V, of the i-th evaluation object toward its reference valuek(i,j)max(tn+tpn) Showing that the ith evaluation object is (t) in the case of considering the single-step control variation amount obtained in the step 2-1n+tpn) Maximum value of variation of time to the reference value direction; control sequence (Δ u) taking into account the m-step control applied in the current control time domain1,Δu2,…,Δum) At sampling point tn+tpnThe maximum adjustment capacity of the evaluation object at the time is estimated as follows:
in the formula Va(i)mFor controlling the time domain length to be mtsMaximum adjustment capability of the alternative control to the i-th evaluation object in the case, tsFor the length of the sampling period to be,
for the ith evaluation object under the condition of not applying control at the sampling point tn+tpnThe predicted voltage amplitude at the time.
7. The method of claim 1, wherein the method comprises:
the predicted voltage value when no control is applied is less than the voltage-optimized reference value, i.e.
The applicability determination index of the control time domain parameter is expressed as:
8. the method of claim 1, wherein the method comprises:
the predicted voltage value when no control is applied is greater than the voltage-optimized reference value, i.e.
The applicability determination index of the control time domain parameter is expressed as:
9. the method of claim 1, wherein the method comprises:
wherein in the step 2, if the suitability is judged by the index IsmIn the presence of an element Is(i)mIf the control step number is less than or equal to 0, the control step number m is continuously increased until IsmAll elements in the composition satisfy Is(i)m>0 and selectMt selecting devicesFor the currently optimised control time-domain length and the next optimised prediction time-domain length, i.e. mts＝tcn＝tp(n+1)，tcnFor controlling the length of the time domain in the nth sub-optimization, tp(n+1)Is the predicted temporal length used in the n +1 th sub-optimization.
10. The voltage coordination control method based on the improved adaptive model predictive control according to any one of claims 7 to 9, characterized in that:
if Is(i)m>0, indicating that the current control time domain parameter setting has the condition of returning the evaluation object I to the control dead zone, and enabling Ism＝(Is(1)m,Is(2)m,……,Is(Ni)m) M has an initial value of 1, if
M continues to be increased until IsmAll elements in the composition satisfy Is(i)m>0, and selecting mtsFor the currently optimised control time-domain length and the next optimised prediction time-domain length, i.e. mts＝tcn＝tp(n+1)，tcnFor controlling the length of the time domain in the nth sub-optimization, tp(n+1)Is the predicted temporal length used in the n +1 th sub-optimization.
11. The method of claim 10 for voltage coordinated control based on improved adaptive model predictive control, wherein:
if the alternative control cannot return all the evaluation objects to the control dead zone range, the control step number m is reset to 1, and the alternative control adds a high-cost control measure considering load shedding; it is assumed that the system is equipped with sufficient reactive compensation and voltage regulation devices in the set voltage correction control scenario.
12. The method for voltage coordination control based on improved adaptive model predictive control according to any one of claims 1-9, characterized by:
in step 3, the objective is to minimize the voltage offset and the control cost of each bus, and the voltage coordination optimization model is as follows:
the model constraints are:
Np(n+1)＝Ncn＝tcn/ts
umin≤ug≤umax
|Δug|≤Δumax
in the formula NpnAnd NcnThe number of sampling points, t, in the predicted time domain period and the control time domain in the nth optimization respectivelycn＝Ncnts,tcnFor the control time domain length, t, used in the nth sub-optimizationsIn order to be the sampling period of time,
is the predicted voltage amplitude, u, at sample point fgFor the value of the control variable at the sampling point g, Δ ugQ, R are punishment weight matrixes of voltage offset and control cost respectively, which are control variable quantities at a sampling point g; u. of0An initial value representing a control variable; vmaxAnd VminRespectively is the upper limit and the lower limit of the system voltage amplitude, and satisfies Vmax＝Vr+DB/2，Vmin＝Vr-DB/2；
The third constraint described above is a voltage constraint in the prediction time domain,
for the voltage amplitude at the sampling point of the current predicted time domain end time, the objective function only specifies the voltage constraint of the predicted time domain end time on the premise that the control variable meets the constraint,
the fourth constraint defines the control variable ugThe maximum value and the minimum value of the input are calculated,
the fifth constraint mentioned above specifies the single-step control variation Δ ugMaximum absolute value of (a).
13. The method of claim 12, wherein the adaptive model predictive control is based on a voltage coordination control method comprising:
in the step 4, if the optimal control sequence at the sampling point is obtained as (Δ u)1,Δu2,…,Δum) Then at tn+tsControlling the first step in the optimal control sequence by delta u1Is applied to the system.
14. A voltage coordination control device based on improved adaptive model predictive control, for implementing the method of any one of claims 1-13.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811498082.1A CN109586287B (en) | 2018-12-07 | 2018-12-07 | Voltage coordination control method and device based on improved adaptive model prediction control |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811498082.1A CN109586287B (en) | 2018-12-07 | 2018-12-07 | Voltage coordination control method and device based on improved adaptive model prediction control |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109586287A CN109586287A (en) | 2019-04-05 |
| CN109586287B true CN109586287B (en) | 2020-12-01 |
Family
ID=65927911
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811498082.1A Active CN109586287B (en) | 2018-12-07 | 2018-12-07 | Voltage coordination control method and device based on improved adaptive model prediction control |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109586287B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111509724B (en) * | 2020-05-06 | 2021-09-07 | 国网山东省电力公司电力科学研究院 | Hierarchical power distribution network voltage control method combining decentralized time sequence and centralized model prediction |
| CN111682792B (en) * | 2020-07-02 | 2023-07-11 | 湖南师范大学 | Converter model prediction control method based on multi-step prediction |
| CN115758784B (en) * | 2022-11-30 | 2023-12-12 | 南方电网数字电网研究院有限公司 | Large jacobian matrix low-time-consumption iteration method for supporting power system time domain simulation |
-
2018
- 2018-12-07 CN CN201811498082.1A patent/CN109586287B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109586287A (en) | 2019-04-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109586287B (en) | Voltage coordination control method and device based on improved adaptive model prediction control | |
| US11326579B2 (en) | Adaptive dynamic planning control method and system for energy storage station, and storage medium | |
| CN106655202B (en) | A kind of multiple target overload aid decision-making method based on interruptible load | |
| CN106959618B (en) | Voltage coordination control method based on step optimization weight | |
| CN114142498B (en) | Data-driven distributed energy storage self-adaptive prediction control voltage regulation method | |
| CN110165667B (en) | Transmission and distribution cooperative reactive power optimization method and system considering static voltage safety constraint | |
| KR20140075617A (en) | Method for estimating smart energy consumption | |
| CN105337290B (en) | A kind of idle method of adjustment suitable for low-frequency oscillation of electric power system aid decision | |
| CN105262084A (en) | Voltage emergency control method based on adaptive model prediction control | |
| CN109546677B (en) | Safety control strategy solving method for large-scale offshore wind power flexible sending-out system | |
| KR20140075614A (en) | Method for making smart energy consumption indicator | |
| CN109167368B (en) | User voltage optimization and regulation method and system with distributed photovoltaic | |
| CN105989544A (en) | Adaptive short-term load prediction method for power distribution transformer area | |
| CN104756031A (en) | Power control device and power control method | |
| CN112526882A (en) | Supercritical unit coordination control method based on hierarchical model predictive control algorithm | |
| CN105068422A (en) | MPC method based on triangular interval constraints | |
| CN111651878B (en) | Large power grid static voltage stability optimization decision method and system considering situation assessment | |
| CN107968400B (en) | Enumeration combination and linear programming based transformer transformation ratio optimization method for monthly operation mode | |
| CN104155876B (en) | The separation of a kind of PID controller realizes method | |
| CN111798060A (en) | Power instruction optimal distribution method based on unit climbing rate estimation | |
| CN115664924A (en) | Endogenous security perception resource management method for social asset participation power grid interaction | |
| CN102158164B (en) | Trapezoidal variable weight regression control method for magnetically controlled shunt reactor | |
| CN114757548A (en) | Wind power energy storage equipment adjusting performance evaluation method adopting scene construction | |
| CN112615364A (en) | Novel wide-area intelligent cooperative control method for power grid stability control device | |
| CN108599198B (en) | Power system frequency and voltage coordination correction method and system based on model predictive control |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |